GEOTECHNOLOGIEN Science Report - gfz-potsdam.degfzpublic.gfz-potsdam.de/pubman/item/escidoc:42034:3/component/... · GEOTECHNOLOGIEN Science Report No. 16 Mineral Surfaces – From

Mineral Surfaces – From Atomic Processes to Industrial Application

Status Seminar26-27 October 2010Johannes-Gutenberg-UniversitätMainz

Programme & Abstracts

GEOTECHNOLOGIENScience Report

No. 16


ISSN: 1619-7399

In the frame of the R&D-programme GEOTECHNOLOGIEN thirteen joint interdiscipli-nary projects have been launched in April 2008. The object of research is the betterunderstanding of the multiple physical and chemical reactions at mineral surfaces.The abstract volume summarizes the scientific results presented during the statusseminar at the Johannes Gutenberg University in Mainz, Germany in October 2010.

Research activities focus on different applications of geomaterials:

– reactions on external and internal interfaces (e. g., immobilization of toxic ele-ments/compounds and radioactive irradiated phases) and bio-mineralization suchas biomimetic functional materials as well as bone implants.

– development of high technology and its applications for the exploration of theEarth's interior on the one hand and the synthesis of new materials (e. g. defor-mation hardening, ultra hard phases) on the other.

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for

Education and Research (BMBF) and the German Research Council (DFG)

No.

16

Min

eral

Sur

face

s –

From

Ato

mic

Proc

esse

s to

Indu

stria

l App

licat

ion

GEO

TECH

NO

LOG

IEN

Scie

nce

Repo

rt Umschlag_SR16.qxd:Umschlag_Infosyst.II.qxd 30.09.10 13:44 Seite 1

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite 1



Status Seminar26-27 October 2010Johannes-Gutenberg-Universität, Mainz


No. 16

Number 1

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite I

Impressum

Editorship / SchriftleitungDr. Ute Münch, Werner Dransch

© Koordinierungsbüro GEOTECHNOLOGIEN, Potsdam 2010ISSN 1619-7399

The Editors and the Publisher can not be held responsible for the opinions expressed and the statements made in the articles published, such responsibility resting with the author.

Die Deutsche Bibliothek – CIP Einheitsaufnahme

GEOTECHNOLOGIENMineral Surfaces – From Atomic Processes to Industrial ApplicationStatus Seminar26-27 October 2010Johannes Gutenberg-Universität MainzProgramme & Abstracts –Potsdam: Koordinierungsbüro GEOTECHNOLOGIEN, 2010(GEOTECHNOLOGIEN Science Report No. 16)ISSN 1619-7399

Distribution / BezugKoordinierungsbüro GEOTECHNOLOGIENTelegrafenberg14473 Potsdam, GermanyFon +49 (0)331-288 10 71Fax +49 (0)331-288 10 [email protected]

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite II

Processes of the mineral surfaces take place at atomic and molecular length scales. Elementaryreactions such as dissolution and the growth and mineral transformations, adsorption and ionexchange are important for a variety of geological processes. Even with technical proceduresweathering reactions are taken advantage of: e.g. in fixation of pollutants on mineral surfaces,in corrosion processes, or in the activation and making of industrial minerals more functional.The aim of the research is to understand and quantitatively assess mineral surfaces in their fullcomplexity in terms of new processes and products to alter chemical, physical and biologicalfunctionalization. Clay minerals are of particular interest due to their specific charge distributions.Also biotechnological processing procedures and the development of bioactive mineral surfacesare investigated.

The understanding of all these different processes and reactions is expected to find new economicpossibilities to clean drinking water in polluted areas, improving ceramics and other consumergoods as well as to discover new bone replacement materials. Since April 2008 thirteen collaborative projects on the topic of »mineral surfaces« have beenfunded by the Federal Ministry of Education and Research (BMBF) with a volume of nearly 8million Euros.

The main objective of the status seminar is to bring scientists and industrial partners together todiscuss new ideas and find new possibilities for cooperation and synergies.

Ute MünchHartmut Fueß

Preface

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite III

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite IV

Table of Contents

Scientific Programme Status Seminar »Mineral Surfaces«. . . . . . . . . . . . . . . . . . . . . . . 1

HYDROPHOBINS – Using hydrophobins to prevent microbial biofilm growth on mineral surfacesRieder A., Schwartz T., Obst U., Bollschweiler C., Gutt B., Zoller J., Fischer R. . . . . . . . . . . . . . . . . . . . . . . . . . 3

BioMin – Functionalized mineral surfaces: Sorption mechanisms of growth-stimulating proteins on surfaces of bone substitutes based on calcium phosphatesLindner M., Koczur K., Kirsten A., Oliveira A., Seifert G., Gemming S., Zurlinden K.,

Meißner M., Jennissen H.P., Müller-Mai C., Fischer H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

HydraSmec – Understanding processes at the hot smectite-water interface for tailoring industrial bentonite applicationsStanjek H., Meister D., Steinkemper U., Wolff H., Diedel R., Habäck M., Grefhorst C., Böhnke S.,

Schmidt E., Latief O.,Schmahl W., Jordan G., Eulenkamp C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

The impact of mineral and rock surface topography on colloid retention Fischer C., Darbha G.K., Michler A., Schäfer T., Lüttge A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Surface and Wetting Properties of Diagenetic Minerals and Sedimentary Grains in Reservoir Rocks (NanoPorO)Altermann W., Drobek T., Frei M., Heckl W.M., Kantioler M., Phuong K.L., Stark R.W., Strobel C. . . . . . . . . . . . . . . . 65

Reactivity of Calcite/Water-Interfaces (RECAWA): Molecular level process un derstanding for technical applicationsStelling J., Neumann T., Kramar U., Schäfer T., Heberling F., Winkler B., Vinograd V., Arbeck D., Müller H.S.,

Haist M., Glowacky J., Vucak M., Fischer U., Pust C., Huber J., Bosbach D. . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Identification and modification of the surface properties of calcite fillers as a basis for new, highly filled adhesives Diedel R., Dörr H., Geiß P. L., Presser M., Roth E., Wittwer W.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

InProTunnel: Interfacial Processes between Mineral and Tool Surfaces – Causes, Problems and Solutions in Mechanical Tunnel DrivingFeinendegen M., Spagnoli G., Stanjek H., Neher H.P., Ernst R., Weh M.,

Fernández-Steeger T. M., Ziegler, M., Azzam, R. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

SIMSAN – Simulation-supported development of process-stable raw material components and suspensions for the production of ceramic sanitary ware on the basis of modified mineral surfaces Engels M., Agné T., Bayard E., Diedel R., Emmerich H., Emmerich K., Haas S., Latief O.,

Peuker M., Steudel A., Vuin A., Yang H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite V

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite VI

Control mechanisms of clays and their specific surface area in growing media – assessment of clay properties and their parametrization for the optimization of plant quality Dultz S., Schellhorn M., Schmilewski G., Schenk M.K., Dombrowski, I., Schmidt, E., Walsch, J., Below, M. . . . . . . . . . . 140

SURFTRAP – Development and Optimisation of a Process to Biosynthesize Reactive Iron Mineral Surfaces for Water Treatment PurposesPeiffer S., Paikaray S., Damian C., Janneck E., Ehinger S., Martin M., Schlömann M., Wiacek C.,

Kipry J., Schmahl W., Pentcheva R., Otte K., Götz A., Wang Z., Hsieh K., Meyer J., Schöne G.,

Koch T., Ziegler A., Burghardt D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

MicroActiv – Optimization of water purification technology for arsenic and antimony scavenging by microbially-activated Fe-oxide mineralsKersten M., Bahr C., Daus B., Driehaus W., Kappler A., Karabacheva S., Kolbe F., Posth N., Reich T.Y.,

Schurk K., Stanjek H., Wennrich R. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

Microstructural Controls on Monosulfide Weathering and Heavy Metal Release (MIMOS)Pollok K., Harries D., Hopf J., Etzel K., Chust T., Hochella M.F., Jr., Hellige K., Peiffer S., Langenhorst F. . . . . . . . . . . . . 182

Authors’s Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199

GEOTECHNOLOGIEN Science Reports – Already published/Editions . . . . . . . . . . . . . . . 203

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite VII

SR16_Mineraloberfl_V8.qxp:Abstract_InfosysII.qxd 08.10.10 13:24 Seite VIII

Scientific Program Status Seminar »Mineral Surfaces« 26–27 October 2010, Universität Mainz

26. October 2010

13.00 – 13.15 Welcome13.15 – 14.00 HYDROPHOBINS – Using hydrophobins to prevent microbial biofilm growth on

mineral surfaces14.00 – 14.45 BioMin – Functionalized mineral surfaces: Sorption mechanisms of growth-sti-

mulating proteins on surfaces of bone substitutes based on calcium phosphates14.45 – 15.30 HydraSmec – Understanding processes at the hot smectite-water interface for

tailoring industrial bentonite applications

15.30 – 16.00 Coffee and Poster

16.00 – 16.45 The impact of mineral and rock surface topography on colloid retention 16.45 – 17.30 Surface and Wetting Properties of Diagenetic Minerals and Sedimentary Grains

in Reservoir Rocks (NanoPorO)17.30 – 18.15 Reactivity of Calcite/Water-Interfaces (RECAWA):Molecular level process un der -

standing for technical applications18.15 – 19.00 Identification and modification of the surface properties of calcite fillers as a

basis for new, highly filled adhesives

about 19.30 Dinner

27. October 2010

08.30 – 09.15 InProTunnel: Interfacial Processes between Mineral and Tool Surfaces – Causes,Problems and Solutions in Mechanical Tunnel Driving

09.15 – 10.00 SIMSAN – Simulation-supported development of process-stable raw materialcomponents and suspensions for the production of ceramic sanitary ware onthe basis of modified mineral surfaces

10.00 – 10.45 Control mechanisms of clays and their specific surface area in growing media– assessment of clay properties and their parametrization for the optimizationof plant quality

10.45 – 11.15 Coffee Break

11.15 – 12.00 SURFTRAP – Development and Optimisation of a Process to BiosynthesizeReactive Iron Mineral Surfaces for Water Treatment Purposes

12.00 – 12.45 MicroActiv – Optimization of water purification technology for arsenic andantimony scavenging by microbially-activated Fe-oxide minerals

12.45 – 13.30 Microstructural Controls on Monosulfide Weathering and Heavy Metal Release(MIMOS)

13.30 – 13.45 Final Discussion

1



3

IntroductionMicrobial biofilms are an extremely successfulway of life. Bacteria and fungi benefit in thissymbiotic life form of metabolic exchange, pro-tection and genetic flexibility. They produce amatrix of organic molecules in which they areembedded and which offers new habitats toother organisms, such as other bacteria orfungi. Biofilms cannot be avoided to colonizesurfaces in unsterile habitats. So, they can befound everywhere in nature and in technicalsystems, but they play an ambivalent role. Onthe one hand biofilms are essential to degradeand transform water contaminations, but onthe other hand they can diminish product qual-ities and damage capital equipment. Biofilmscan cover medical equipment such as cathetersand pathogenic bacteria, which may be living inthe biofilms, are a continuous source of infec -tion of the patients. In addition, the metabolismof the biofilm microorganisms may change thecomposition of the fluids or contaminate themwith their products. As biofilms are all-round,the understanding of the biofilm formation andits manipulation are of prime importance inmicrobiology and material sciences.The choice of a material and the correspond-ing surface properties like mechanical proper-ties, structure, polarity, and chemistry influencethe binding of various molecules and cells. Thesurface properties affect the biocompatibility of

a material and consequently also bacterialadhesion, and biofilm growth. In this projecthydrophobins are used as a novel modificationof surfaces to change surface properties likehydrophobicity and thus might have an effecton biofi lm formation. Hydrophobins are fungal proteins, which self-assemble on hydrophobic as well as hydrophilicsurfaces into extremely stable monolayers. Re -combinant hydrophobins provide the opportu-nity to use these highly surface-active proteinsfor large-scale surface coatings. Hydrophobinsare non-toxic and can be used for surfacemodification and functionalization (with e.g.enzymes) of industrial re levant materials likesteel, plastics, and ceramics.

In this project hydrophobin coated surfacesand their properties are studied with respect tobacterial cell adhesion, cell differentiation, andcellular growth with the aim to influencebiofilm formation.In the first part of the project recombinanthydrophobins were produced and purified.Different surfaces were coated with hydropho-bins and characterized, since the coating effi-ciency is the basis for subsequent biofilm for-mation studies. Biofilms were grown on natu-ral as well as hydrophobin coated surfaces anddifferent methods were established to analysebiofilm formation.

HYDROPHOBINS – Using hydrophobins to prevent microbial biofilm growth on mineral surfaces

Rieder A. (1), Schwartz T. (1), Obst U. (1), Bollschweiler C. (2), Gutt B. (3), Zoller J. (3), Fischer R. (3)*

(1) Department of Microbiology of Natural and Technical Surfaces, Institute of Functional Interfaces (IFG), Karlsruhe

Institute of Technology (KIT), e-mail: [email protected]

(2) BASF SE, e-mail: [email protected]

(3) Department of Microbiology, Institute for Applied Biosciences, Karlsruhe Institute of Technology (KIT)

e-mail: [email protected]

*Coordinator of the project


4

Since the hydrophobin coated surfaces did notreduce microbial growth, we designed modi-fied fusion hydrophobins and attached cation-ic antimicrobial peptides (AMPs) to the hydro -phobins.

1. HydrophobinsHydrophobins are small proteins with about100 amino acids, which are produced by fila-mentous fungi. In nature they coat the surfaceof fungal spores to ease the growth out of theground and to protect the spore itself. It isrecently known that hydrophobins also preventthe human immune recognition of airbornefungal spores. There are two classes of hydrophobins, whichonly differ in the arrangement of eight highlyconserved cysteine residues. Self-assembledmonolayers formed by class I hydrophobins aremore stable than those formed by class IIhydrophobins. For our research we are usingDewA, which is produced by Aspergillus nidu-lans and belongs to the class I hydrophobins.Since hydrophobins are highly surface-activethey can easily be used for surface modificationand functionalization. Until recently, the use ofhydrophobins for large-scale surface coatingshas been prevented by the fact that they hadto be isolated from fungal cultures in a lengthyprocess. Recently the BASF SE achieved abreak through and established an expressionsystem for the large-scale production of hydro -phobins.

1.1 Fusion hydrophobinsBased on the large-scale industrial production,it is now possible to obtain large amounts ofhydrophobins in a comparatively short time, butdue to their characteristics, the hydro phobins

had to be produced as so called fusion hydro -phobins. The BASF SE provides two fusionhydrophobins for surface coating. The availablefusion hydrophobins are composed of the classI hydrophobin DewA of Aspergillus nidulansand a fusion partner. The fusion partner is thesynthase yaaD of Bacillus subtilis in completerespectively shortened form (Figure 1). The syn-thase is used to lead the hy dro phobin into theinclusion bodies in Escherichia coli for thepurification. H*Protein B has the shortest ver-sion of yaaD which still works as a leader. Thefusion hydrophobins have a mass of 47 kDarespectively 19 kDa, whereas the hydrophobinpart itself has a mass of only 10 kDa. In fusionhydrophobin H*Protein A the hydrophobin partforms 20%, in H*Protein B 50% of the totalprotein mass.

1.2 Modification of the fusion hydropho-binsBecause of the lack of being antimicrobial byitself, we modified the fusion hydrophobins.We used cationic antimicrobial peptides (AMPs),which are only 9 to 50 amino acids in size andare proven to be active against bacteria andeven against yeasts and filamentous fungi.They are an alternative to antibiotics and donot affect human cells. Until now there is noresistance known. The AMPs were fused to theN-terminus of the long and short version of theyaaD in the expression vector pQE60 (Figure 2).The modified plasmids (pHHC-011, pHHC-012,pF1-011 and pF1-012) were transformed intoE. coli BL21. The induction of the expressionwas performed with 0.5mM IPTG (Isopropyl β-D-1-thiogalactopyranoside) at OD600 0.6 to 1.0;we let the culture grow over night. The purifi-cation of the modified fusion hydrophobins

Figure 1: The fusion hydrophobins consist of the synthase yaaD in complete (H*Protein A) respectively shortened(H*Protein B) form, the class I hydrophobin DewA and a C-terminal His-tag.


5

was done by an approved protocol of the BASFSE, where a pH-shift from 12 to 9 dissolves themodified fusion hydrophobins out of the IBs.After centrifugation the dissolved proteins re -main in the supernatant, which was lyo phi lizedafterwards. The molecular masses of the modi-fied fusion proteins are only a little larger, be -cause of the small size of the AMPs. The mod-ified fusion hydrophobins have a mass of about48 kDa and 20 kDa, respectively (Figure 3).

1.2.1. Testing the functionality of themodified fusion hydrophobinsWe used different organisms, gram-positiveand gram-negative bacteria and eukaryotes (S.pneumoniae, B. subtilis, E. coli, P. putida and A.nidulans) to show the functionality of the mo -dified fusion hydrophobins. We performed kil-ling curves (A. nidulans, B. subtilis, S. pneumo-niae) and peptide activity tests (A. nidulans, B.subtilis, S. pneumoniae, P. putida) to show this.First we tested if only the fusion hydrophobins

have an effect on the growth of microorga-nisms. We used the strain B. subtilis TMB488,which is a reporter strain for the two-compo-nent system LiaFSR and is inducible by cellenvelope stress by antibiotics, which have aneffect on the lipid II cycle (Jordan et al., 2007).Using different concentrations of the fusionhydrophobin we could not detect a growthinhibition (Figure 4). We found the same resultswith the other organisms. Killing curves wereperformed using the modified fusion hydro-phobins, but there was no difference in thegrowth recognizable. Because of these results we decided to performpeptide activity assays in microtiter plates.Peptide activity assays were performed earlierto check the antimicrobial activity of the AMPsF1 and HHC in A. nidulans (Mania et al., 2010).We found MICs (Minimal Inhibitory Con cen tra -tions) of about 1µg/ml respectively about5µg/ml, so we decided to use these AMPs formodification.

Figure 2: Fusion hydrophobins with cationic antimicrobial peptides. We fused the AMP to the N-terminal site of theyaaD. After biochemical calculations the AMPs are on the top of the coated surfaces and the contact point for biofilms.


6

In each column of the plate, one peptide isdiluted from top to bottom to roughly identifythe MIC. This is done in Minimal Medium(MM) with glucose as sole carbon source andresazurin as indicator of viability. Resazurin is ablue dye that turns pink when it is reduced, forexample by respiration of living microorga-nisms in the medium. Each well is inoculatedwith a defined a mount of spores (105 spores)of an A. nidulans wild type strain (RMS011).First we tested all the AMPs we have in ourstock (F1, HHC, #15, #29, Bac2a, C6, D5, E6,Indolicidin and Kai13) and then we tested themodified fusion hydro pho bins. There was nogrowth inhibition of the F1 and HHC fusionhydrophobin in any tested or ganism (Figure 5,and data not shown), although the peptidesalone were highly active. This suggested aninhibitory effect of the fused hydro phobin.

1.2.2. Modification of fusion hydrophobinswith a surface-tethered peptideMeanwhile Hilpert et al. (2009) found someAMPs, which work better while they are tethe-red to surfaces. One of these peptides, KaiH,consists of 12 amino acids (WIVVIWRRKRRR)and has a mass of 1.7 kDa. We attached thispeptide to the N-terminus of the shortenedand full-length version of the yaaD and clonedit also in the pQE60 vector (Figure 6). As anegative control we chose a peptide slightlychanged peptide, hW (WIVVIWRAKRRR), andattached it also to the shortened and full-length versions of the yaaD. The over expres-sion and purification will be done according tothe same protocol as before.

Figure 3: Over expression analysis. A) Western blot: The modified fusion hydrophobin F1-012 serves as an example fora western blot. The modified fusion hydrophobins were transformed with E. coli BL21 an expression strain. With induc-tion by 0.5mM IPTG the modified fusion proteins were expressed and could be detected by the 6x-His tag. There is aclear band at about 20kDa, which corresponds to the mass of the modified fusion hydrophobin. B) SDS-Gels: By overexpression the modified fusion proteins were enriched and applied on 15-15% SDS Gels. Of the lyophilized proteinpowder we applied 10µg/ml in each column. As comparison serve the fusion hydrophobins we got from the BASF SE.


7

Figure 4: Growth curve of B. subtilis TMB488. The strain grew in LB-Media (lysogenic broth) with 5µg/ml chlorampheni-col. A 100ml culture was inoculated at OD600 0.1 with a preculture of B. subtilis TMB488. Then let it grow until OD6000.6 when it was splited into five 10ml cultures. To each of these cultures we gave a different concentration of fusionhydrophobins (0.1µg/ml, 1µg|ml, 10µg/ml and 100µg/ml). One of these grew without any add-on.

Figure 5: Peptide activity assays with P. putida. A) Testing 10 different AMPs for antimicrobial activity. As control servesaqua dest. In each column there is another peptide with decreasing concentration. B) In a second step the modifiedfusion proteins (5-10) were tested. In each column there is another modified fusion protein with decreasing concentra-tion. As control serve the AMP (F1 or HHC), aqua dest. and the fusion hydrophobins from the BASF SE.


8

2. Characterization of hydrophobin coatedsurfaces In this project hydrophobin coated surfacesand their properties are studied with respectto bacterial cell adhesion, cell differentiation,and cellular growth with the aim to influencebiofilm formation.In a first step unmodified recombinant fusionhydrophobins were used and a coating protocolwas developed. The hydrophobin coated sur -faces were characterized under various aspects.

2.1 Surface coating with hydrophobinsGlass surfaces were used as reference materialfor coating experiments. Glass is a veryhydrophilic material and thus comparable toceramic materials. The glass surfaces werecoated with the fusion hydrophobins H*Pro-tein A and H*Protein B. The used concentra-tion of the hydrophobin solution was 10µM.The glass surfaces were coated according toJanssen et al. (2004). The hydrophobins weresolved in buffer (50mM Tris-HCl, 1mM CaCl2,pH 8). Subsequently, the surfaces were cleanedwith ethanol and incubated for 1-16 hours at23-80 °C in the hydrophobin solution. Duringthe incubation the hydrophobins self-assem-bled on the surface in α-helical conformation.The surfaces were rinsed with distilled waterand dried at room temperature. Half of thesurfaces was further treated for the inductionof a β-sheet shift. The β-sheet conformation ofthe hydrophobin coating is said to be more

stable than the α-helical conformation. For thisthe surfaces were incubated for 10 minutes at80 °C in 2% SDS-solution, subsequently rinsedwith distilled water and dried. Different parameters like incubation tempera-ture and incubation time might influence theself-assembly behavior and the postulated mo -nolayer of hydrophobins. These parameterswere changed to analyze the fundamental cha -racteristics of fusion hydrophobins on surfaces.

2.2 Characterization of hydrophobin coa-ted surfacesDifferent surface analysis methods were ap p -lied for the characterization of hydro pho bin-coated surfaces and the clarification of the in -fluence of the variable coating parameters.The hydrophobin-coated surfaces were ana-lyzed in matters of change of surface hydro -pho bicity and homogeneity of the coating. Inaddition also the adsorption characteristics ofhydrophobins were studied.

Change of surface hydrophobicityContact angle measurement (CA) is a simple-to-adopt method for surface analysis. It is usedto detect the presence of films or coatings andto determine their properties. The contactangle describes the shape of a liquid dropletresting on a solid surface and thus gives infor-mation about the wettability of the surface.The more a liquid droplet spreads, the smalleris the contact angle and the more hydrophilic

Figure 6: Modified fusion hydrophobins with the cationic antimicrobial peptide KaiH. Fusion of the AMP to the N-termi-nal side of yaaD and cloning in the pQE60 vector.Figure 6: Modified fusion hydrophobins with the cationic antimicrobial peptide KaiH. Fusion of the AMP to the N-termi-nal side of yaaD and cloning in the pQE60 vector.


9

Figure 7: Contact anglesand immunofluorescencemicroscopy are applied tomonitor the change of sur-face hydrophobicity andcoating homogeneity ofhydrophobin coated sur-face. The fusion hydro pho -bins H*Protein A andH*Protein B adhere in atemperature and time de -pendent manner. The long -er the incubation time themore homogenous is thecoating in alpha-helicaland beta-sheet conforma-tion. Less surface hydro -pho bicity is lost with theinduction of the beta-sheetshift at long incubationtimes.

is the surface. Since hydrophobins are amphi-philic proteins, which self-assemble on surfa-ces the contact angel of uncoated vs. coated,surfaces is changed and used as a parameterfor the efficiency of the coating.For contact angel measurements the ContactAngle System OCA20 and the software SCA20(Dataphysics, Filderstadt, Germany) were used.The contact angle of a 5 µl water droplet (ses-sile drop technique) was calculated with theLaplace-Young equation. The contact anglewas determined on at least five different spotson the surface. The hydrophobin coating changed the wettabi-lity of the glass surface. Glass is a hydrophilicmaterial (contact angle 11° ± 2°). With the self-assembly of the hydrophobins on the glass sur-face it became more hydrophobic. The changeof the surface wettability was dependent onthe incubation time, the incubation temperatu-re and the hydrophobin conformation. H*Protein A and H*Protein B changed thewettability of the glass surface in a very similarway. Longer incubation times and an increasedtemperature resulted in significant hydropho-bicity changes of about 60° (Figure 7). Theinduction of the b-sheet shift caused losses of

up to 20° of primarily achieved surface hydro-phobicity. Here a big influence of incubationtemperature and incubation time was noticed.Coatings prepared at room temperature losemuch more hydrophobicity compared to coa-tings at 80 °C. The longer the incubation timethe more stable was the coating and the lesswas lost with the induction of the b-sheet shift.

These results indicated that the longer the incu-bation time and the higher the incubation tem-perature the more efficient and stable was thehydrophobin coating. A long incubation time(16h) and high incubation temperature(80°C) increased the surface hydrophobicityof hydrophobin coated surfaces significantly.

In addition to the surface hydrophobicity alsothe stability of hydrophobin coatings wasdetermined with contact angel measurements.The surface hydrophobicity of freshly coatedsurfaces was compared to the surface hydro-phobicity of surfaces stored at room tempera-ture for up to five weeks. The coating was sta-ble and no change was detected in surfacehydrophobicity over time.


10

Homogeneity of hydrophobin coatingTo determine the homogeneity of the hydro-phobin coating immunofluorescence micro-scopy was applied. Immunofluorescence mic -roscopy is often used in biology and medicineto visualize proteins and their distribution.The proteins act as antigens and can bedetected with fluorescent labeled antibodies.They are studied using a fluorescence or con-focal microscope.For the detection of the hydrophobin layer onthe surface a primary Anti-his antibody wasused which specifically binds to the His-tag ofthe proteins. A secondary fluorescent labeledantibody binds to the Fc region of the primaryantibody and could be detected with the fluo-rescence microscope. The characteristics of H*Protein A and H*Pro -tein B on glass surfaces were very similar. Sur -face coatings prepared at room temperature ina-helical protein conformation were homoge-nous with just a few holes and cracks. The lon-ger the incubation time the more homogenouswas the coating, but the differences wereinsignificant. After the induction of the b-sheetshift the surface coating was very disordered.On glass surfaces coated for one respectivelysix hours at room temperature no hydrophobincoating was detectable with immunofluore-

scence microscopy after the b-sheet shift.These results were in accord with the contactangle data, which showed that a lot of hydro-phobicity was lost on these surfaces. The surfaces which were coated at 80 °C weremuch more homogenous compared to the sur-faces coated at room temperature. There wereslight differences between the various incuba-tion times, but all coatings in a-helical confor-mation were homogenous (Figure 7). After theinduction of the b-sheet shift big differencesoccurred. The longer the incubation time themore homogenous was the coating. Theseresults confirmed the contact angle data. Thelonger the incubation time and the higher theincubation temperature the more homoge-nous was the coating in the a-helical and theb-sheet conformation. It was essential to incu-bate the materials for 16 hours at 80 °C in theprotein-solution to form a homogenoushydrophobin-layer in a-helical and b-sheetconformation.

To confirm the results of the immunoflures-cence microscopy AFM (atomic force micro-scope) measurements were applied. The mea-surement were performed in the measurementcell of an MFP-3D BioAFM (Asylum, Mann -heim) having a commercial Si3N4 cantilever of

Figure 8: AFM height images of glasssurfaces coated with H*Protein Aand H*Protein B in alpha helical con-formation for 16 hours at 80 °C. Theproteins self assemble on the sur-faces in ordered globular structures.


11

a normal spring constant of 0,56 N/m (nanoand more) in air. The microscope was operat-ing in an AC-mode, where the tip was scannedback and forth at 0° along the horizontal linein a scan range of 10 µm. The AFM was usedwith a 5 µm z-range and 150 µm x- and y-range scanner (type J Digital instruments). To -po graphic images were evaluated with theScan ning Probe Image Processor (SPIP). On glass surfaces hydrophobins adhered inordered protein layers. Surface coatings perfor -med at 80 °C with H*Protein A and H*ProteinB in alpha-helical and beta-sheet conformationshowed a globular protein adsorption pattern(Figure 8). These results were in accordancewith the immunfluorescence microscopy. Al -so at the nano level hydrophobin coatingsperformed for 16 hours at 80 °C were ho -mogenous.

Adsorption characteristics of hydrophobinsThe adsorption characteristics of fusionhydrophobins were monitored with QCM-D(quartz crystal microbalance with dissipation)measurements. QCM is a very sensitive tool todetect changes in weight and thus a helpfulmethod to sense adsorption processes on sur-faces. The quartz crystal microbalance deter-mines a mass per unit area by measuring thechange in frequency of a quartz crystal reso-nator. The resonance is disturbed by the addi-tion or removal of a small mass due to, forexample, protein deposition at the surface ofthe acoustic resonator. It can be used undervacuum, in gas phase and in liquid environ-ments. In addition to measuring the frequen-cy, the dissipation is measured. The dissipa-tion is a parameter quantifying the dampingin the system, and is related to the sample'sviscoelastic properties. The adsorption of H*Protein A and H*ProteinB (10µM in 50mM Tris-HCl, 1mM CaCl2, pH8) were analyzed on a SiO2 coated quartz cry-stal. The protein layer thickness and theabsorbed mass were calculated with the Voigtvicoelasitc model. The protein adsorption wasmonitored for up to 16 hours at 20 °C and aflow rate of 50 µl/min. The stability of the for-

med protein layer was determined by rinsingwith 2% SDS solution. H*Protein A formed a layer of 17 ± 3 nm andH*Protein B of 14 ± 2 nm on the SiO2 surfa-ce. After 1 hour of incubation the maximumlayer thickness was already reached. No morechanges were observed during further incu-bation. These results were in accordance withthe data of the contact angle measurementsand the immunofluorescence microscopy. Toachieve a homogenous hydrophobin coatingand a significant change of the surface hydro-phobicity short incubation times were suffi-cient. The formed coating thickness of 17 ± 3nm respectively 14 ± 2 nm matched theexpected thickness of hydrated monolayers offusion hydrophobins. The fact that even witha longer incubation time no changes of thelarger thickness were monitored, pleads forthe formation of a stable monolayer.The adhered protein mass was 17 ± 3 mg/m2

for H*Protein A and 14 ± 2 mg/m2 for H* Pro -tein B (Figure 9). Unfortunately the formedprotein layer was not stable under these con-ditions. After rinsing with 2% SDS solutionthe hydrophobin layer was detached comple-tely. As already mentioned, the temperatureplays an important role in the formation of astable hydrophobin layer. Contact angle mea-surement and immunofluorescence microsco-py have shown that higher temperatures (80 °C) were necessary to form stable mono-layers. The QCM experiments were perfor-med at 20 °C. This might result in this non-stable hydrophobin layer.

Various surface analysis methods were establis-hed and applied for the characterization ofhydrophobin coated surfaces. At a later stageof this project these techniques will also beused for the characterization of surface coa-tings with modified fusion hydrophobins. Theinfluence of the modification on the surfaceperformance of the fusion hydrophobins willbe determined.

3. Analysis of microbial biofilm formationon different surfaces Biofilms were grown on natural as well as


12

hydrophobin coated surfaces with singlespecies and natural mixed populations.Different methods were applied and estab-lished to analyze the bacterial and fungal pop-ulation in terms of population composition,microbial density and spatial distribution.Additionally the interaction of bacteria andfungi in biofilms was monitored.

3.1 Bacterial biofilm formationMicrobial biofilms represent a special, very suc-cessful life form of bacteria. The microorgan-isms benefit in this symbiotic life form frommetabolic exchange, genetic flexibility and pro-tection, which a biofilm offers in comparisonto planktonic growth. Nearly all bacterial spe -cies are able to exhibit this type of growthfirm ly adhering to a substrate, in addition tothe planktonic freely floating growth variety. A special characteri stic of all biofilms, besidesfirm adhesion to one site, is their highly struc-tured character (Costerton et al. 1999, Coster-ton et al. 2001.) The development of a biofilmproceeds in various chronological steps. Theseare dependent on the involved bacterial spe -cies external factors and also surface charac-teristics. The first step, the initial adhesion isreversible. With the production of a matrix oforganic molecules, the extracellular matrix, theadhesion gets irreversible. The bacteria are

embedded in the extracellular matrix, whichalso offers new habitats to other organismsuch as other bacteria or fungi. The biofilmstarts to mature and to form three-dimension-al structures. A mature biofilm releases singlecells, which again adhere to the substrate andstart the biofilm life cycle again. In order to characterize the influence of hy dro -phobin coated surfaces on bacterial biofilmformation the primary adhesion as well as ma -ture biofilms grown on uncoated and coatedsurfaces were analyzed.

Primary adhesionThe primary adhesion is the first and conse-quently the crucial step at the change fromplanktonic to sessile growth.

Primary adhesion of Escherichia coli on hydro-phobin coated and uncoated glass slidesTo analyze the primary bacterial adhesion aGFP-tagged E. coli strain (E. coli BW3110,pJOE 4056.2 His e-GFP) was used. After induc-tion of the promoter with 0.2% Rhamnose the»green fluorescent protein« (GFP) is stableexpressed intracellular. The bacteria can bedetected with the epifluorescence microscopewithout further labeling steps. E. coli was grown in biofilm reactor on hydro-phobin coated and uncoated glass slides and

Figure 9: Adsorption of hydrophobinH*Protein B (10µM) in Tris buffer(50mM Tris, 1mM CaCl2, pH 8.0) onSiO2 surface monitored with QCM-D.The adsorped mass was quantifiedwith Voigt viscoelastic model.


13

the bacterial adhesion was monitored at diffe-rent time points for up to 24 hours. The biofilmwas washed in 0.89 % NaCl to remove looselyattached microorganisms and subsequentlyanalyzed with epifluorescence microscopy.More bacteria adhered on hydrophobin coatedglass slides compared to uncoated glass slides.

Quantification of biofilm growthFor the quantification of the first steps of thebiofilm formation a crystal violet assay wasapplied. Individual cavities of a 96-well micro-titer plate (polystyrol) were coated with H*Pro -tein A and H*Protein B according to the esta-blished protocol. The GFP-tagged E. coli (E.coli BW3110, pJOE 4056.2 His e-GFP) wasgrown overnight and diluted in sterile growthmedium (OD600nm 0.25). 100 µl bacterial sus -pen sion were portioned in each cavity andincubated at 37 °C for up to 6.5 hours. Theadhered bacteria were stained with crystal vio-let and the biofilm formation could be quanti-fied with the determination of the optical den-sity at 590nm. At most time points a higher adhesion of E. coliwas detected on hydrophobin coated surfacescompared to uncoated surfaces. After just 0.5hours a significant higher bacterial adhesion onthe various hydrophobin coated surfaces wasmonitored. This effect diminished and eveninverted after 6.5 hours. Now more bacteriaadhered on hydrophobin coated surfaces com-pared uncoated polystyrol (Figure 10).

Mature biofilmThe mature bacterial biofilms were grown inwastewater effluent on hydrophobin coatedand uncoated glass slides for up to fourweeks and subsequently analyzed in terms ofpopulation composition, microbial density andspatial distribution.

Population analysesMolecular biological population analyses likePCR-DGGE (polymerase chain reaction withsubsequent denaturing gradient gel electro-phoresis) allow studying the diverse populationof bacteria in wastewater biofilms. Populationanalyses were used to detect variations in thecomposition of the bacterial population onhydrophobin coated and uncoated surfaces.For the analysis of the bacterial population thebiofilm DNA was isolated and amplified in aPCR-reaction with the primer set 27f and 517rtargeting the eubacterial 16S rDNA. The 16S rDNA of prokaryotes is the most conser-ved (least variable) gene and for this reasonused to identify bacteria. The initial PCR-reac-tion resulted in a mixture of PCR ampliconswhich all have the same length (526 bp). Se -quen ce variations like differences in GC contentand distribution were used to separate theseamplicons in a denaturing gradient gel electro-phoresis. Here each band represents one bac-terial species. DGGE banding patterns can beused to visualize variations in microbial di ver sityand provide a rough estimate of the richness

Figure 10: The primary adhesion of E. coli was monitored at differenttime points and quantified with crys-tal violet staining. The differences in the biofilm growthon hydrophobin coated and uncoat-ed surfaces are shown. At early timepoints (0.5 h) significantly more bac-teria adhere on hydrophobin coatedsurfaces.


14

and abundance of predominant microbial com-munity members. The bacterial population wasanalyzed with PCR-DGGE (Figure 11). Theplanktonic bacterial population (left) differedclearly from the biofilm population (right).No significant differences were visible in thebanding pattern of biofilms grown on diffe-rent surfaces. The bacteria adhered with thesame preference to natural and hydrophobincoated surfaces.

Microbial density A further characteristic of biofilm growth isbesides the population composition the num-ber of bacteria growing in a biofilm. Colonyforming units (CFU) were used to estimate thegrowth density. The biofilm was washed in0,89% NaCl to remove loosely attached mic ro -organisms, scraped of the surface with a sterilecell scraper, resuspended in sterile PBS and seri-al diluted. Suitable dilution steps were plated intriplicates on DEV-agar. They were incubatedfor up to 7 days at 37 °C. The bacterial co lonieswere counted and the number of bacteria persquare cm was calculated. A lot of bacteriagrew on the uncoated and hydrophobin coatedsurfaces, but there was no significant differen-ce in the cell numbers on various materials.

Spatial distribution of biofilm growth Biofi lms were stained directly on the surfacewith the DNA intercalating fluorescent dyesSyto9 to scan the spatial distribution of biofilmgrowth. Prior to fluorescent staining the bio-film was washed in 0.89 % NaCl to removeloosely attached microorganisms. Syto9 solu-tion was dropped on the biofilm sample tostain all present microorganisms. The samplewas incubated for 20 minutes in the dark atroom temperature and subsequently washedcarefully with water. The slides were preparedwith the mounting media citifluor and acoverslip for fluorescence microscopy. No significant differences were observed in thespatial distribution of the biofilm growth onuncoated and hydrophobin coated surfaces.

Different steps of biofilm development weremonitored on hydrophobin coated and uncoa-ted surfaces. Hydrophobin coated surfacesinfluence the primary bacterial adhesion. Thebacteria are attracted by the hydrophobin coa-ted surfaces and adhere in higher numbers ason uncoated surfaces. But the impact ofhydrophobin coated surfaces on biofilm for-mation seems to be time dependent. In matu-re biofilms (4 weeks) no differences in biofilms

Figure 11: DGGE profiles of amplifiedpartial 16S rRNA genes of planktonicwastewater bacteria (right) andwaste water biofilms (left). The bandsobtained from different pure culturesof hygienically relevant bacteria serveas reference marker (left):Campylobacter jejunii, Staphy lococ -cus aureus, Enterococcus faecium,Pseudomonas aeruginosa, E. coli,Salmonella enterica, Mycobacteriumparatuberculosis (top to bottom). The banding pattern of wastewaterbiofilms on uncoated and coatedmaterials showed no significant dif-ferences.


15

established on natural and hydrophobin coa-ted surfaces could be detected.The established techniques for the biofilm cha-racterization will further be applied to charact-erize the influence of modified hydrophobinson biofilm formation. The hydrophobins will bemodified with antimicrobial peptides to enhan-ce their effect on biofilm formation.

3.2 Fungal biofilm formationIn addition to bacteria there are also fungi insi-de of a biofilm. When the mature biofilm is for -med, fungi joined the biofilm. To analyze thefun g al populations in biofilms we use nativebiofilms. A native biofilm is a biofilm, which canbe found in the natural environment and har-bors more than one microorganism species. In anative biofilm, protozoan, algae, bacteria andfungi can be found.

Design of a clone libraryTo analyze and identify the fungal populationof a biofilm a molecular biological approachwas used and a clone library was designed. Aclone library offers the advantage that alsofungi, which are not able to grow on agarplates, can be detected and identified. With aclone library it is possible to estimate the abun-dance of single fungal species in biofilms. Forthe generation of the clone library the com-plete biofilm DNA was isolated and amplifiedin a PCR-reaction with the primer set ITS1 andITS4. These primers are specific for the ITS-region of eukaryotes. The ITS-region is wellsuited for the identification of fungi since ithas a higher variability than other regions ofthe rDNA. The PCR products were cloned inEscherichia coli and tested with a colony PCR(primer set M13/T7) and the restriction enzyme

Table 1: Isolated fungi and protozoa from native biofilms. A clone library was designed for the isolation of the fungi with the amplified ITS-region in E. coli. The sequences wereanalyzed by a M13/T7 PCR, a digestion and by the online-tool BLAST. Plant 1 and 2 are the different places in thepurification plant of Karlsruhe/Neureut. Plant 1 is the place in the mechanical treatment direct after the rack and thesecond place was in the biological treatment of the plant. Influent biofilm is a biofilm, which was incubated in influentwastewater for one month. The isolates from the biofilm of the lake are also listed in the table.


16

MspI. Clones, which showed different bandingpatterns on an agarose gel after the restrictiondigest, were sequenced.

The isolated and identified fungi and protozoafrom wastewater biofilms and biofilm from alake in Karlsruhe-Leopoldshafen are listed intable 1. In addition to these isolates, 111 clo-nes were sequenced which were not describedso far. All isolates of the biofilm from the plantin autumn, the Alb, the effluent and the Rheinwere species, which are not described so far.Most of the undescribed clones belong to thegenus Candida and Trichosporon, which canbe shown in a genealogical tree. Until now it isin general not much known about fungi in bio-films. These clones are very interesting sincethey might be fungi preferably growing in bio-films (Table 1). There was no significant diffe-rence in the fungal population on uncoatedand hydrophobin coated materials.

Spatial distribution of biofilm growthBiofilms were stained with Calcofluor directlyon the surface of glass slides. Calcofluor bindsspecifically to chitin and cellulose and stainsfungi and algae. Prior to fluorescent stainingthe biofilm was washed in 0.85% NaCl toremove loosely attached microorganisms. Forthe staining of fungi and algae a small volumeof 0.1% Calcofluor and 15 % KOH were drop-ped on the biofilm and subsequently washedwith 70 % ethanol and water. The sampleswere analyzed with the fluorescence microsco-pe (Figure 12).

3.3 Interaction of bacteria and fungi inbiofilmsFluorescence in situ hybridization (FISH) wasapplied to analyze the distribution and interac-tion of bacteria and fungi in biofilms. FISH is atechnique used to detect and localize the pre-sence of specific DNA sequences in situ. FISHuses specific fluorescent-labelled probes thatbind to only those parts of the genome with

Figure 12: Spatial distribution of biofilms on uncoatedsurfaces stained with Calcofluor. The glass slide was incubated for four weeks in the influ-ent of the purification plant in Karlsruhe/Neureut. Afterthe incubation the slide was washed with 0.85% NaCland stained with 0.1% Calcofluor and 15% KOH andwashed again with 70% ethanol and water.


17

which they show a high degree of sequencesimilarity. With this technique different mic ro -organism groups can be stained in situ withdiverse fluorescent dyes.

To analyze the interaction of bacteria and fungitwo different probes were applied. EuUni waslabeled with Fluorescein and bound to a high-ly conserved region of the eukaryotic 18SrRNA. The prokaryotic 16S rRNA was detectedwith the Cy3 labeled probe Eub338. The FISHwas done as described Baschien et al. 2008.The biofilms were fixed, dehydrated and after-wards hybridized with the probes. The sampleswere analyzed with the fluorescence microsco-pe. For our investigations we used glass slides,which were directly incubated in the influentof the plant Karlsruhe/Neureut and in the Alb(effluent of the plant). Figure 13 shows a FISHof a biofilm, which was incubated in the

influent of the plant. The slide was incubatedfor four weeks, washed with 0,85 % NaCl toremove loosely attached microorganisms andhybridized with both probes.

ReferencesBaschien, C., Manz, W., Neu, T.R., Marvanová,L., Szewzyk, U. (2008). In situ detection offresh water fungi in an alpine stream by newtaxon-specific fluorescence in situ hybridiza-tion probes. AEM, Oct. 2008, p. 6427-6436

Costerton, J.W., Stewart, P.S., Greenberg, E.P.(1999). Bacterial Biofilms: A Common Causeof Persistent Infections. Science 284, 1318 –1322.

Costerton, J. W., Damgaard, H. N., K. -J. Cheng(2001). Cell Envelope Morphology of RumenBacteria. J Bacteriol. 118, 1132-1143.

Figure 13: Fluorescence in situ hybridization with a biofilm of the purification plant. The glass slide was incubated for four weeks in the influent of the purification plant Karlsruhe/Neureut. AfterwardsFISH was carried out with the specific probes for fungi and bacteria. The biofilms were fixed, dehydrated and hy bri -dized with the probes. Shown in red is the bacteria specific probe (Cy3), in green the fungi specific probe (FAM).


18

Hilpert, K., Elliott, M., Jenssen, H., Kindrachuk,J., Fjell, C.D., Körner, J., Winkler, D.F., Weaver,L.L., Henklein, P., Ulrich, A.S., Chiang, S.H.,Farmer, S.W., Pante, N., Volkmer, R., Hancock,R.E. (2009). Screening and characterization ofsurface-tethered cationic peptides for antimi-crobial activity. Chem Biol. 2009 Jan30;16(1):58-69.

Janssen, M.I., van Leeuwen, M.B., van Kooten,T.G., de Vries, J., Dijkhuizen, L., Wosten, H.A.(2004). Promotion of fibroblast activity by coa-ting with hydrophobins in the beta-sheet endstate. Biomaterials 25 (14), 2731-2739

Jordan, S., Rietkötter, E., Strauch, M.A., Kala -morz. F., Butcher, B.G., Helmann, J.D., Ma -scher, T. (2007). LiaRS-dependent gene expres-sion is embedded in transition regulation inBacillus subtilis. Microbiology 153, 2530-2540

Mania, D., Hilpert, K., Ruden, S., Fischer, R.,Takeshita, N. (2010). Screening for antifungalpeptides and their modes of action in As per -gillus nidulans. Not yet published.


19

BioMin – Functionalized mineral surfaces:Sorption mechanisms of growth-stimulatingproteins on surfaces of bone substitutes basedon calcium phosphates: Second status report

AbstractThe manufactured samples of amorphous andrecrystallized bioactive glass 45S5 were sterili-zed for further investigations especially for thein vivo testing. Two common sterilization pro-cedures for solid bodies, steam sterilizationand hot air sterilization, were tested. After thesteam sterilization the surface topography andthe chemical content at the surface of thesamples changed. This effect could not beobserved after hot air sterilization. Hence, hotair sterilization procedure was chosento sterili-ze the samples. We developed a manufactu-ring process to produce porous b-tricalciumphosphate specimens. NH4HCO3 was succes-fully used to create pores inside the samples.The advantage of this material is that it decom-poses at low temperature (60 °C) and so onlycan affect the samples a short temperaturerange from room temperature up to 60 °C.Hence, no cracks where detected at the surfa-ce and the core of the samples. Different kindof pore sizes and amount of porosity was cre-

ated in the ceramic parts by using this porosi-fying agent. The bounding ability of the BMP-2 (Bone Morphogenic Protein) of amorphousand recrystallized bioactive glass was charact-erized by two different concentrations of theBMP-2 in the immersion fluid. The results showthat a doubling of the concentration of theBMP-2 in the immersion fluid resulted in thetwo-fold amount of BMP-2 at the surface ofthe amorphous and the recrystallized bioactiveglass. However, the release of the BMP-2 sho-wed a difference with respect to these twomaterials. Both materials show an initial burstphase in the release of the bone morphogene-tic protein from the surface of about two daysfollowed by a sustained release. During theinitial burst phase more BMP-2 was releasedfrom the recrystallized surface compared toamorphous surface of the bioactive glass. Asexperimental results on the BMP-2 activity onmineral surfaces indicate that a non-covalentattachment to an unpolar surface functionali-zation yields the best coverage, the relevant

Lindner M. (1), Koczur K. (1), Kirsten A. (1), Oliveira A. (2), Seifert G. (2), Gemming S. (3), Zurlinden K. (4),

Meißner M. (4), Jennissen H.P. (4), Müller-Mai C. (5), Fischer H. (1)*

(1) Department of Dental Materials and Biomaterials Research, RWTH Aachen University Hospital,


(2) Department of Chemistry, Theoretical Chemistry, TU Dresden, e-mail: [email protected]

(3) Institute of Ion Beam Physics and Materials Research, Research Center Dresden-Rossendorf,


(4) Institute of Physiological Chemistry, Biochemical Endocrinology, University of Duisburg-Essen,


(5) Clinic for Surgery, Department of Trauma Surgery and Orthopaedics, Knappschaftskrankenhaus Bochum-Langendreer,


* Coordinator of the project


20

structural characteristics of BMP-2 under oste-ogenic conditions were studied by moleculardynamics simulations with a biological forcefield. An analysis of the local flexibility of BMPmonomers and dimers in solution and on anon-polar functionalization revealed majorstructure changes at the N-terminus, close tothe a-helix, and around the disulfide bridge inthe dimer. As all secondary structure elementsremained intact in the bonded dimer, theresults indicate that the flexible areas in-bet-ween may facilitate the docking of BMP-2 tosurfaces, but are not involved in sterically dri-ven osteogenic activity. Electronic structure cal-culations suggest alkylphosphonic acids andnucleotides as suitable surface functionaliza-tions, which bind to the mineral surface withthe polar part and expose an unpolar part toattach. The amorphous and recrystallized bio-active glass coated and not coated with BMP,respectively, were implanted in New Zealandwhite rabbits for four different implantationtimes (7, 28, 84, 168 days). 6 implants were us -ed for one kind of material and implantationtime, i. e. overall 96 cylindrical implants (3.96 mmdiameter, 8.1 mm height) have been inserted.At present, the evaluations of specimens priorimplantation and at 7 days were completed.Due to the implantation procedure a gap hea-ling was observed, leading to a delay in boneformation in comparison to former studies inthe same animal model with bioactiveimplants. Up to 7 days after implantation nobone-bonding was observed in any of the spe-cimens. All of the materials (amorphous, cry-stalline each with and without BMP-coating)displayed no obvious changes in surface densi-ty and structure up to now indicating a lowersurface reactivity as known from 45S5 bioacti-ve glass. At 7 days the highest expression of anearly marker, which indicates that the bonematrix is activated, was observed at the amor-phous coated implants. Therefore, stimulationin bone formation around the BMP-coatedimplants is possible. Subsequently, we willhistologically analyze the samples of longerimplantation times to confirm this observation.

IntroductionOur research joint project focuses on the inter-face between implant material and biologicalenvironment. More precisely we want to com-pare the interaction of an amorphous material(bioactive glass) und a crystalline material (re -crystallized bioactive glass) with bone tissue.The bioactivity and subsequently the boneremodelling process weakens especially whengreater bone defects have to be restored bythis class of material. It is known that thegrowth of bone tissue can significantly be sti-mulated by so called Bone MorphogeneticProteins (BMPs). In our project we want tocompare the coupling ability and the desorp-tion kinetics of the BMPs on the amorphousand the crystalline surface. Tests in vitro, in vivoand additionally computational simulation ofthe coupling process will create further know-ledge of the interaction of the BMPs with themineral surface and the biological environ-ment. This knowledge is important for thedevelopment and manufacturing of tailoredbone substitute implants, so that degradationof the substitute material and build-up of newbone tissue can go hand-in-hand in vivo. In thefollowing chapters the results are presentedthat have been achieved within the first fifteenmonth of our joint project.

Sterilization of amorphous and recrystal-lized bioactive glassThe manufactured and analyzed samples ofamorphous and recrystallized bioactive glass(see science report 2009) must be sterilizedbefore further experiments like the immobiliza-tion of BMP-2 on the surface and the in vivotesting can be carried out. Therefore a stan-dard sterilization procedure for solid statebodies of a steam autoclave (120 °C, 1h) wasused. The surface of the samples was exami-ned before and after the sterilization processusing a scanning electron microscope (SEM)and energy dispersive X-ray analysis (EDX).After the sterilization process the surface topo-graphy of the samples changed compared tothe surface of the samples which were not ste-rilized with hot steam (Fig. 1, Fig. 2). Becausethe surface topography changed after the


21

steam sterilisation, another suitable standardi-zed sterilization process had to be found whichhas no or only a marginal effect to our speci-mens. One adequate sterilization process pro-ved successfully was the hot air sterilizationmethod (180 °C, 2h). After the hot air sterili-zation process the surface topography did not

change compared to the surface of the sam-ples which were not sterilized (Fig. 1, Fig. 2).

Furthermore the method of steam sterilizationdid not only influence the surface topographyof the samples. This method changed also thechemical composition at the surface. The EDX-

Figure 1: Analysis of the microstructure of a recrystallizedbioactive glass surface after grinding without subsequentsterilization.

Figure 2: Analysis of the microstructure of a recrystallizedbioactive glass surface after grinding and subsequentlysterilized with steam.

Figure 3: Analysis of the microstructure of a recrystallizedbioactive glass surface after grinding and subsequentlysterilized with hot air.

Figure 4: EDX-analysis of a recrystallized bioactive glasssurface after grinding without subsequent sterilization.

Figure 5: EDX-analysis of a recrystallized bioactive glasssurface after grinding and subsequently sterilized withsteam. An enrichment of Ca and P could be detected.

Figure 6: EDX-analysis of a recrystallized bioactive glasssurface after grinding and subsequently sterilized with hotair. No enrichment of Ca and P could be detected.


22

analysis showed an enrichment of phosphorand calcium at the surface of the sampleswhich were sterilized by the steam sterilizationmethod compared to the specimens whichwere not sterilized (Fig. 4, Fig. 5). The samplestreated using the method of hot air sterilizationdid not show an enrichment of phosphor andcalcium at the surface and provided the samechemical content on the surface like the sam-ples which were not sterilized (Fig. 4, Fig. 6).

Now that there was found a sterilizationprocedure which did not influence the sur-face of the specimen, 124 cylindrical samples(3.96±0.03 mm diameter, 8.1±0.1 height)were manufactured for the in vivo experi-ments. 96 specimens were implanted, 16 bak-kup samples were fabricated and 12 cylinderswere used for checking the amount of theadsorbed BMP-2.

Fabrication and characterization ofporous b-tricalcium phosphate (b-TCP)specimensb-tricalcium phosphate (b-Ca3(PO4)2) is one ofthe most investigated biodegradable ceramicmaterials for medical applications to supportbony defects. The pore size and the amount ofporosity are one of the most important para-meters for using this material for bone repla-cement. Different procedures have been deve-loped to synthesize porous ceramic parts.Nearly all techniques to establish porosity in aceramic specimen use a porosifying agentwhich burns out during the sintering processand creates the desired porosity in the manu-factured part. Therefore, one main property ofthe porosifying agent is, that it completelydisappears during firing process. For that rea-son organic substances like oil or carbon basedmaterials are common materials to createporosity in ceramic specimen. As first attemptsto create porosity, it was used a polyurethanefoam. One disadvantage using polyurethane isthe higher thermal expansion of the polyur-ethane compared to the ceramic. This resultedin cracks in the part before the sintering processcould start. A sample was heated up to only200 °C instead up to the sintering temperature

of 1100 °C. The foam did not burn out becau-se of the low temperatures and large crackswere visible at the surface of the samples.Because of the described problems with thepolyurethane foam we decided to test othermaterials for creating the porosity in the sam-ples. One of these materials was NH4HCO3.The advantage of this material was that itdecomposes at low temperatures (60 °C) toNH3, H2O and CO2. Hence there was only asmall temperature range (room temperatureup to 60 °C) when this material could interactwith the ceramic material. This material wastested by mixing powder of b-TCP andNH4HCO3 together with different weightratios. Tablets were pressed of this materialand the sintering processes was performed. Apressing device with a diameter of 12 mm wasused to press cylindrical specimen of a heightof about 1 mm and 18 mm. Using NH4HCO3 asa porosifying agent, the samples showed nocracks at the surface after the sintering pro-cess. The manufactured parts were analyzed incomparison to specimens which were made ofb-TCP only (Tab. 1).

The results in Table 1 show, that the amount ofporosity of the sample is about 38 vol.-%when no NH4HCO3 is used. This porosity iscaused by the sintering process of the b-TCPparticles which results in pores smaller than 10 µm (Fig. 7 and Fig. 8).

The SEM micrographs of the sintered specimenwith NH4HCO3 showed clearly the differencein the microstructure of the manufactured spe-cimen compared to the parts made of b-TCPonly. The fracture surface of a ceramic partmade of 50 % b-TCP, 50 % NH4HCO3 < 50µmshows the pores which are caused by the poro-sifying agent (Fig. 9 and Fig. 10)Using more NH4HCO3 results in specimenswith over 80 vol.-% porosity. The mechanicalstrength of these parts was very low. Evenhandling these parts for further investigationsresulted in small pieces which cracked from thesamples.


23

Table 1: Resulting porosity for different particle size and different amounts of NH4HCO3.

Figure 7: Overview of a fracture surface a sintered speci-men made of b-TCP. No pores can be detected at thismagnification.

Figure 9: Overview of a fracture surface a sintered samplemade of 50 % b-TCP, 50 % NH4HCO3 < 50µm. In com-parison to Fig. 10 pores of 50 µm can be detected.

Figure 8: Microstructure of a fracture surface a sinteredspecimen made of b-TCP. Only small pores (< 10 µm) canbe detected.

Figure 10: Microstructure of a fracture surface a sinteredsample made of 50 % b-TCP, 50 % NH4HCO3 < 50 µm.In comparison to Fig. 8 the pores in the size of 50 µm canbe detected.

Now that there was found a suitable porosify-ing agent the next goal was to create differentdefined pore sizes in the samples. Therefore,the NH4HCO3 was grinded and the particle sizedistribution was analyzed before, during andafter the milling process (Tab 2). The materialwas milled dry with ZrO2 milling balls on a rol-ling platform.The meaning of the value d10 is, that 10 vol.-%of the particles are smaller (or equal) than thegiven value; analog for d50 and d90. The resultsshow a proper milling process by the time of 9hours. When now NH4HCO3 of different mil-

ling times is used to create the porosity in theparts, different sizes of pores are the result.Also we wanted to obtain a material with onlya short range in pore size distribution.Therefore the NH4HCO3 can be sieved withtwo sieves at the same time. For example asieve of 63 µm mesh is stuck on top of a sieveof a 45 µm mesh and the NH4HCO3 is sieved.The particles which do not pass the 45 µm andremain on the top of this sieve have a particlesize between 45 µm and 63 µm. Hence theover all range of this particle size distribution isonly 18 µm. If this material now is used as a


24

porosifying agent, it will lead to same smalldistribution in pore size of the manufacturedceramic part.

Immobilization of BMP-2 on cylindricalbioactive glass 45S5 samplesFirst experiments for investigating the immobi-lization and the controlled release of BMP-2from bioactive glass were carried out by usingminiplates containing of amorphous and cry-

stalline bioactive glass respectively. Results ofthese experiments were shown in our lastreport.In contrast for animal experiments cylindricalsamples were used. So it was necessary toinvestigate the immobilization and the control-led release of BMP-2 from these cylindricalsamples. First investigations show unsteadyresults which could explained by microscopicinvestigations. REM measurements illustrate

Table 2: Specific values of the particle size distribution depending on the milling time (NH4HCO3).

Table 3: Immobilization of BMP-2 on cylindrical samples of bioactive glass.

Table 4: Amount of bound BMP-2 on cylindrical samples used for animal experiments

Figure 11:Release of BMP-2 from cylindricalbioactive glasssamples.


25

gaps and holes on the surface as a consequen-ce of the steam sterilisation process.After changing to the hot air sterilisation pro-cess cylindrical samples with an undamagedsurface could be used for pre-experiments andanimal experiments.Depending on BMP-2 concentration in the in -cubation solution 1.1 - 2.2 µg BMP-2 per cm²glass surface were bound (tab. 3).

Furthermore the release of BMP-2 from thesesurfaces was examined (see Figure 11). A greatpart of immobilized protein is released fromthe surface during first two days.

In a »press-fit-experiment« cylindrical sampleswere put into rabbit bones comparable to theoperation process. We could show, that onlyca. 10 % of immobilized protein were lost du -ring this procedure.

Unfortunately the amount of bound protein onsamples used for animal experiments differsfrom the results of the pre-experiments andalso among themselves although all terms andconditions were constant. Table 4 shows thesedata for all parts of the animal experiment.

Adsorption and desorption of BMP-2 onbioactive glass 45S5 and silica glass mini-platesThe fabricated and sterilized bioactive glass45S5 was used to the adsorption of BMP-2 fromthe mineral surface. It is the aim of this work tobestow osteoinductivity by immobilizing BMP-2on a glass surface. The binding of BMP-2 to animplantable bioactive glass surface (45S5 speci-fication) by a radiotracer method and to amodel silica glass surface by evanescent wavetechnology will be described here.Amorphousbioactive glass (45S5 specification, Ra ~ 0.9 µm)was prepared in polished miniplate form (5 x 10x 1 mm) for BMP-2 binding measurements byradiotracer technology. Highly polished silicaglass (Suprasil I, Ra ~ 1-3 nm) in form of roundplates (∅ = 36 mm, 0.9 mm thick). Ultra hydro-philic surfaces (θ 0-10°) were prepared withchromosulfuric acid (CSA). 125I-BMP-2 (0.1 mg/ml) was adsorbed on bioactive glass for 17hours according to and unlabelled BMP-2 (0.1mg/ml, 3.8 µM) on Suprasil for 5 min in a TIRF-rheometer (Trp fluorescence) by air-bubble tech-nology. Adsor p tion was performed in 20 mMNa acetate buffer pH 4.5 (buffer A) in bothcases. After binding to bioactive glass for 17 h125I-BMP-2 was desorbed for 11 days in phos-phate buffered saline pH 7.4 (buffer B, PBS). OnSuprasil after obtaining equilibrium in ca. 5 minBMP-2 was desorbed in buffer A for 20 h.

Figure 12: Release of BMP-2 from the surface of polished amorphous bioactive glass (bio-active glass). The data was fitted to a two-phase exponential decay function (r2 = 0.99)


26

As shown in Fig. 12, BMP-2 is adsorbed on theultra-hydrophilic (~8 µg/cm2). The release ofBMP-2 occurs in the form of a two phase expo-nential first order decay reaction. The firstexponential corresponds to a burst phase witha half-life (t1/2) of ~ 0.3 days (k-1 = 2.7 x 10-5 s-1).The second phase corresponds to a sustainedrelease phase with a half life of ~ 42 days (k-2

= 1.9 x 10-7 s-1). For biological activity the sam-ples loaded with BMP-2 were allowed to releaseBMP-2 for three days into the cell culturemedium. From the dose-response curves theK0.5 values for released BMP-2 were determinedto 11-20 nM (control values 5-10 nM) indicatinga high biological activity of released BMP-2.

Fig. 13A shows the adsorption kinetics ofBMP-2 on silica glass in the TIRF-rheometer.

Adsorption occurs as a single exponential (kobs = 0.199 s-1) without a mass transport limi-tation. Desorption after 5 minutes (Fig. 13B)proceeds as atwo phase exponential decay (k-1 = 4.3 x 10-4 s-1;t1/2,1 = 0.45 h and k-2 = 1.6x 10-5 s-1;t1/2,2 = 12.2 h). Thus for k-2 the des-orption rate is 42 times higher on Suprasil thanon bioactive glass.

Figure 14 shows the kinetic constants of initialBMP-2 binding (encounter complex) to Sup ra -sil: k+1 = 7.98 x 104 M-1s-1, k-1 = 0.071 s-1 anda binding constant of K = 1.1 x 106 M-1. From the k+1 = 7.98 x 104 M-1s-1 of BMP-2 bin-ding on Suprasil (Fig. 14) and the k-2 = 1.9 x10-7 s-1 of desorption on bioactive glass (Fig.12) an apparent binding constant K´ = 4.2 x1011 M-1 can be calculated. This high apparent

Figure 13: Adsorption (A) and Desorption (B) of BMP-2 on suprasil silica glass measured by TIRF-rheometry. Desorptionwas measured after 5 min of adsorption.

Figure 14: kobs-Plot of the kobs valuesvs the BMP-2 concentration for the cal-culation of the initial kinetic constants.


27

binding constant for sustained release, compa-red to the encounter complex, may be due toconformational changes during the 17 h incu-bation step.

Simulation of solvent and surface effectson BMP-2 monomer and dimerBone morphogenetic proteins such as BMP-2actively participate in the bone remodelingprocess by stimulating osteoblast differentia-tion and by regulating osteoblast and oste-oclast activities. Experimental evidence sug-gests that this activity relies on specific structu-ral characteristics, because also fragments ofBMPs exhibit osteogenic activity [D1]. Resultsobtained by the BIOMIN project partners onthe BMP-2 activity on bioactive glass and bio-ceramic surfaces with polar and non-polarfunctionalized surfaces indicate that a non-covalent attachment of BMP-2 to an unpolarsurface functionalization yields the best cove-rage. These findings were the starting point forsystematic molecular dynamics simulationswhich aim at determining the relevant structu-ral characteristics of BMP-2 under physiologi-cal, osteogenic conditions. Such detailedknowledge of the functional BMP-2 substruc-ture is essential for devising biologically tailo-red implants with better remodeling propertiesand yields a perspective for a better, moredirect medicamentation of bone cancer. The present molecular dynamics simulationswere carried out with the established, biologi-cally oriented force field CHARMM27 and theNAMD simulation package, which allow tosystematically incorporate the conditions at thebone remodeling site in a step-wise manner. Incollaboration with the experimentally workingBIOMIN partners we have determined threefocus areas, which lead to subsequently morecomplex model structures and modeling condi-tions and thus allow to systematically expandour understanding of the biological activity ofBMP-2 during regrowth: first is to determinethe structural stability of the BMP dimer andthe monomers which form BMP-2 in aqueoussolution and at body temperature, second is torationalize the osteogenically most favorableinteraction of BMP-2 with the functionalized

mineral surface of the bone substitute cera-mics, and third is to identify the specific activestructural motifs within the BMP-2 geometryand their steric, polar and non-polar features.Results from the first step were reported alrea-dy last year, thus we summarize only the majorfindings: The optimum BMP-2 structure, com-posed of two equal strands of 114 amino acidseach and bonded via a disulfide bridge of thecystein residue number 78 was determinedand compared with a non-bonded dimer andthe monomer, because the latter two structu-res would be more easily accessible by synthe-tic methods than the complete BMP-2 protein.Further findings from those calculations werethat major structural rearrangements occur onthe time scale of several nanoseconds, whichposes the necessity for adequately long simu-lation times and that water stabilizes all struc-tures and most pronouncedly the dimericforms of BMP-2. The outcome of this task wit-hin the BIOMIN collaboration has been docu-mented in several contributions to scientificconferences and journals [D2-D6]. During the second project year, which is repor-ted here, first the study of the structural stabi-lity was concluded with more detailed data onflexible and rigid parts of the BMP, and majorprogress was achieved in focus area two,which concerns the interaction of BMP-2 withdifferently functionalized mineral surfaces.

To finish the structural study of BMP in mono-meric and dimeric forms, a detailed, residue-resolved analysis of the local flexibility was per-formed, which revealed that major structurechanges occur around the disulfide bridge andthe flexible N-terminus of the amino acid chainunder physiological conditions. Fig. 15 showsthe time-averaged root mean square deviation(rmsd) of the residue positions as function ofthe residue number. The rmsd values of themonomer, the non-bonded dimer and thedisulfide-bridged dimer in aqueous surroun-ding, T = 37 °C, are displayed from top to bot-tom. The spatially and time-averaged rmsdvalues of the three structures are indicated asdashed lines. With a value of 2 Å, the bondeddimer exhibits a lower overall flexibility than


28

the monomer and the non-bonded dimer(both about 3 Å). In the spatially resolved cur-ves, residue 114 corresponds to the rather rigidC terminus, whereas residues 1 to 10 belongto the highly flexible part at the N-terminuswhich reaches rmsd values of up to 14 Å(monomer), 18 Å (non-bonded dimer), and 12Å (dimer). Those rsmd values are beyond thedisplayed range, which was adapted to theoverall smaller variations along the polypeptidechain. This result indicates that the N-terminuscan not be involved in an osteogenic activity,which relies on purely steric effects. However,the flexibility may facilitate the docking ofBMP-2 to surfaces or receptors via the N-ter-minus, hence its flexibility may be of seconda-ry, chemically driven physiological relevance. Other flexible areas with rmsd values of up to6 Å comprise the residues around the disulphi-de bridge at amino acid number 78 (region II)and two non-polar regions between residues40 to 60 (region I) and residues 90 to 100(region III). To analyze the inherent steric chan-ges, the secondary structure of the system wasdetermined by applying the standard routines,here program STRIDE [D7], which is part of theMD simulation and visualization package. The

original experimental assignment comprisesnine separate beta-sheet units, b1 to b9, andone beta-sheet, b5a, attached to the alphahelix a3. These rigid subunits are joined by fle-xible parts, which are several amino acids wide[D8]. If the experimental secondary structure isderived by means of the MD analysis routines,the rigid subunits are determined in closeagreement with the experimental assignment.Differences concern the neighboring sheets b6and b7 and also b8 and b9, which are found tocombine into two longer sheets, and b5a,which is missing in the MD-derived analysis.

Fig. 16 depicts the secondary structure sche-matically for the original experimental assign-ment (exp.), the re-analysis of that data by theMD routines (*exp.), the monomer, and eachstrand of the non-bonded and bonded dimers,separately. Triangles denote amino acids thatparticipate in a beta sheet, circles indicate theposition of residues which belong to an alphahelix. For all model systems the secondary fea-tures agree excellently with the ones of theexperimental data. As sole difference only thebonded dimer matches the reference length ofb2, whereas b2 is too long in the other struc-

Figure 15: Time-averaged root meansquare deviation (rmsd) of the residueposition as function of the residue num-ber along the amino acid chain of BMP-2. The curves show the rmsd of themonomer, the non-bonded dimer andthe bonded dimer from top to bottom(solid lines) along with the spatially ave-raged values (dashed lines). Residue 1 isthe N-terminus.


29

tures. This analysis allows correlating the peaksin Fig. 15 to residues which do not participatein secondary structure elements. Regions I andII brace the alpha helix, thus this analysis impliesthat the rigid helix moves as a whole relative tothe rest of the protein. This mechanism mayfacilitate BMP-2 bonding to non-polar surfaces.The peaks in region III imply that all four betasheets b6 to b9 can move relative to one anot-her, hence the flexibility of the area is reprodu-ced well by the simulation despite the assign-ment as two larger beta sheets. For the second step, studying the interactionof BMP monomer and dimers with differentlyfunctionalized mineral surfaces, adsorption onmodel surfaces was investigated by moleculardynamics. As simple model for a mineral surfa-ce with non-polar aprotic functionalization aregular, self-assembled bilayer of POPC (POPC= 1-palmitoyl-2-oleoyl-phoshatidyl-choline) wasequilibrated at body temperature. As POPCwould attach to a mineral surface via the polarphosphatidyl-choline moiety and expose thenon-polar palmitoyl and oleoyl chains towardsthe liquid phase, a bilayer which exposes thealkyl chains was chosen as first model. In con-tact with an aqueous solution of BMP-2 theprotein then employs its non-polar fragmentsto anchor on the alkyl-functionalized surfacevia van-der-Waals interactions. Initial cursorysimulations pursued during the first reportingperiod have been extended to simulations ofup to 30.000 ps simulation time and confir-med the previous findings. After an initial peri-od of a few nanoseconds a non-polar region ofthe protein starts interacting with the surface.The final BMP-2 adsorption site is an upright

position coordinated via the alpha helix area ofone of the monomers, and still exposing theother one. Under the given conditions BMP-2is form-stable, i.e. it maintains its active ternaryand quaternary structure. The monomer, in con-trast, looses its ternary structure and spreads onthe surface. For the dimer the disulfide bridgeand additional non-polar interactions betweenthe two strands occur, which stabilize the bon-ded dimer and corroborate the trend to lowerrmsd values obtained for the dimer in solution.

Simulation of hydrophobic functionaliza-tion on bioceramicsInsight into the atomistic processes at the bio-functionalized mineral surface can be achievedby molecular dynamics simulations, whereasand a quantum-mechanical treatment is requi-red for specific details of the structure and thestructure-property relation. In line with the MDsimulations of BMP-2 on non-polar surfaces wehave investigated the local structural and elec-tronic changes at oxide surfaces with aproticand weakly polar to non-polar functionalization.The first system addresses the modification ofa titanium dioxide surface by adsorption of thenucleotide cytidine. Calculations with the den-sity-functional-based tight-binding (DFTB)method show that on the rutile (110) surfacecytidine favors anchoring with two oxygenatoms of its phosphate part. Adsorption occurspreferentially at two neighboring five-foldcoordinated Ti atoms along the [001] direction,thus opening a pathway to an ordered adsorp-tion of unpolar substances along [001]. Theelectronic densities of state show that the aro-

Figure 16: Secondary structure of BMP-2.The curves show beta sheet areas (trian-gles) and alpha helix areas (circles) of theexperimental structure (exp.) in its origi-nal assignment, along with the subunitsfound by the automatized numericalprocedure for the experimental structure(*exp.), the monomer (mon.) and bothstrands of the non-bonded (ndi’ andndi”) and of the bonded dimer (dim.’and dim.”) from top to bottom. Thenum bering of beta sheet and alpha helixareas is indicated above the graph; resi-due 1 is the N-terminus.


30

matic part of the cytidine residue remains un -changed upon adsorption on rutile. This impliesthat no significant changes occur in the nanotu-be binding capacity by p-stacking of the aroma-tic part, hence, nucleotide-functionalized oxidesurfaces are ideal substrates for the ordered,stable and electronically and chemically inertimmobilization of larger unpolar moieties [D9].As second system self-assembled monolayersof different alkylphosphonic acids on corun-dum a-Al2O3 (0001), bayerite b-Al(OH)3 (001)and boehmite γ-AlOOH (010) surfaces werestudied by DFTB calculations. Mono-, bi-, andtridentate adsorption modes were considered.In addition, the organization of single adsor-bed molecules was compared to the organiza-tion at full surface coverage. The height (thick -ness) of the self-assembled monolayers is al -ways shorter than the length of the phospho-nic acid molecules due to tilting of the alkylchains. Tilt angles at full surface coverage arevery similar to the tilt angle of a single adsor-bed molecule, which indicates that the densityof the self-assembled monolayers is limited bythe density of adsorption sites. The lateral inter-actions between alkyl chains are evidenced bysmall torsions of the adsorbed molecules,which may serve to minimize the repulsion for-ces between interchain hydrogen atoms.Similar tilt angles were obtained for mono-, bi-,and tridentate adsorptions. Hence, the coordi-nation mode is not related to the molecule til-ting, which implies that dense unpolar functio-nal films can be formed on arbitrarily shapedmineral surfaces [D10].

Classical modeling of structure formationin aqueous solutionsA reaction-diffusion-type formulation of thebone remodelling process in terms of Turingstructures had been set up during the firstreporting period. This model has been success-fully tested for structure formation in ionic solu-tions and describes brine channel formation inaqueous salty solutions close to the phase tran-sition temperature in very good agreement withexperimental data [D11].

In vivo investigationsAfter production and BMP-coating of bioactiveimplants animal experiments were planned toevaluate the tissue response to the implants. Acomparison between coated and non-coatedimplants was performed. Different qualitativeand quantitative histological methods wereused to describe the tissue response to theimplants and the material response afterimplantation. The main question to be answe-red is whether the BMP-coatings do stimulatebone growth around two different types ofbioactive implant material of similar chemicalcomposition or not. Two different types ofimplant materials were used in this project.One consisted of an amorphous glas, whichwas close to 45S5 bioactive glass and thesecond material was a partially crystallineglass-ceramic of the same composition. Eachgroup of material was either used as producedor after coating with BMP (Bone Morpho ge -netic Protein). The diameter of the implantswas 3.96 mm und the length was 8 mm.Theimplants were intended to fit exactly to thedrill hole to minimize loss of the coating duringinsertion. The burr displayed an outer diame-ter of 3.95 mm.Pathogen free New Zealandwhite rabbits were used for the experiments(Charles River).They were held separately ineither plastic or wire cages according to theGerman Tierhaltungsverordnung BGBl. llNr.485/2004 idF BGBl. ll Nr. 530/2006.Theygot a standard 12 hour night and day rhythm,and their diet consisted of pressed dry pellets(Smith®) to which 25 mg/kg salinomycinso-dium were added.Prior to the operation the implants were steri-lized by using dry heat. One half of theimplants was coated with BMP. The animalswere anesthetized with a Ketamin (0.35 ml perkg) and Xylazin (0.17 ml per kg), which wasgiven intra muscularly. Whilst performing thesurgery an additional Isoflurane anesthesia wasgiven through a respiration mask. 0.5 mlGentamycin per Rabbit was injected to preventInfection, as well as 0.3 ml Rimadyl for painmanagement. After shaving the legs up to thehip joint and disinfecting the area with


31

Braunoderm, an incision was made medially tothe knee joint. The patella was luxated lateral-ly after cutting through the medial retinacu-lum. A hole was then drilled in a sagittal direc-tion into the trabecular bone of the distal epi-physis of the femur (diamond coated cylindrichollow drill, outer diameter of 3.95 mm). Thecooling of the device was assured throughNaCl-solution which was sprayed onto the drillwith a syringe. After implantation, the Patellawas forced back in its normal position. For thesuture of the retinaculum a Dacron 0 Vicryl®

thread was used, and to close the skin thesuture material was 3-0 Mersilene®. The skinwas closed with a continuous suture techniqueand a securing single suture. The animals hadsurgery on both knees to make sure the stressis the same on both knees, and the animaldoes not fall into a relieving posture.The animals were put in a deep anesthesiawith 2.5 ml Ketamin, 1 ml Xylazin and they got1.3 ml of Fentanyl. additionally to that the rab-bits were given Isoflurane. The skin was cutwith a scissor, then the Knee joint was laid bareand the ligaments were cut through. Thepatella was lifted up and the distal femur wasseparated from the femoral diaphysis. Diamondcoated cutting disks were used to cut the distalpiece of the femur right underneath theimplant. Then the condyles on both sides of thebone were cut off to open the intertrabecularspaces for immersion fixation. The last step wascutting off the trimmed piece from the rest ofthe femur. The cooling was also done with NaCl

through a syringe. The Ani mals were then putto death by injecting 2 ml of T61 (Embutramid)intra venously whilst still in deep anesthesia.Data of operations, used implants and sacrificeof the animals are given in Table 5.For histological characterization of the hostresponse to the implants and the concomitantmaterial response after insertion differentmethods were applied. The complete evalua-tion of each specimen comprised conventionallight microscopy (LM) after staining withGiemsa and von Kossa-Fuchsin as well as histo-morphometry, immunohistochemistry (IHC),scannning- (SEM) and transmisssion electronmicroscopy (TEM).

The tissue blocks were fixated in Histochoiceand embedded in plastic. Then they were gluedupon slices and cut with a saw (Leica sp 1600).The cuts are about 30 µm thick and wereground and polished. Each specimen for con-ventional LM was cut with the surrounding tra-becular bone from ventral to dorsal. The first 6got staining for IHC (Osteocalcin, Osteopontin,Osteonectin, Bone Sialoprotein, Alcaline Phos -phatase and Collagen1), one was left for SEM-BSE; two got the Giemsa stain, two the vonKossa Fuchsin reaction and another one wasleft as a reserve. The procedure of cutting theslices is demonstrated in Table 6.

For IHC, slices 1-6 were deacrylated and incu-bated with specific antibodies (OsteonectinAON-1-s, Bone Sialoprotein WVID1 (9C5)-s,

Table 5: Date of implantation/explantation according to the four groups of materials. Per implant material and timepoint 3 animals were operated on and 6 implants have been inserted. Thus, using 4 implant materials (amorphous,crystalline, amorphous BMP-coated, crystalline BMP-coated) a total of 48 animals and 96 implants were used.


32

Osteopontin MPIIIB10 (1)-s, Alkaline Phos pha -tase B4-78-s, Collagen 1 M-38-s from the Uni -versity of Iowa and Osteocalcin [OCG3],ab13420-50 from abcam. The slices were stai-ned with AEC (liquid3-amino-9-ethylcarbazole),counterstained (Mayer´s haematoxylin) andmounted (Kaisers glyceringelantine, coverslips).The samples are fixated with Glutaraldehydeand Cacodylate buffer, drained in an ascen-ding alcoholic line and dried with HMDS (Hexa -methyldisilazane). To discover alterations of im -plant surfaces one implant of each type andimplantation time as well as one implant ofeach type prior and after coating was used forSEM analysis. The TEM-samples were fixated asthe SEM-specimens and embedded in Spurr.Cutting and evaluation start in October 2010.At 7 days all of the cut specimens showed a gapbetween the implant and the surroundingbone, although it was intended to create an inti-mate contact between the preexisting tissueand the implants. The gap between the oldpreexisting bone and the implant surface wasup to 600 µm wide. Those spaces are filled withsoft tissue in some cases and in others they wereempty due to shrinking artefacts. New bone tra-beculae were not observed in the drill holewhich is attributed to the short time period.Bone-implant contacts were observed onlyexceptionally were old bone trabeculae reached

the surface. Obvious differences in the tissueresponse between amorphous and crystallineimplants and between coated and non-coatedimplants as well were not observed. Looking atthe cross-cut implants the crystalline specimensappeared to be denser than the amorphousones. In both materials, coated or non-coated,parts of the brittle material were lost due to thesawing process. The interface between boneand tissue was in most of the cases intact, poin-ting to an intimate contact between implantsurface and tissue. Obvious differences in thedensity of all of the implants between the outer-most surface and the centre were not observed.At 7 days only slight differences in antibodyreactions could be seen between the 4 implantmaterials. The only marker that showed signi-ficant colouring was Osteocalcin (Fig. 17, 18).There were only slight differences betweenamorphous and crystalline implants. But withboth implant types, there was a stronger reac-tion to Osteocalcin (OC) with coated amorphousas well as crystalline implants. In all groups themineralized bone matrix was coloured redwhich indicates a high level of activity of theOC in this area. The doted coloration rangedbetween mild and severe and it was alwaysgeneralized. OC is an early marker, which indi-cates that the bone matrix is activated. In oneof the cuts also osteocytes were coloured

Table 6: Sequence of the slices produced be sawing and grinding. The numbers star form ventrally (1)and stop at the dorsal aspect of the specimens (12).

Figure 17: Amorphous implant, 7 days, BMP-coated.Sur rounding bone with generalized and strong col-oration for OC.

Figure 18: Amorphous implant, 7 days, no coating. Thecoloration for OC is generalized but moderate.


33

moderate and generalized in the mineralizedbone matrix which implicates their activity ofbuilding the bone.

With Alkaline Phosphatase only one (crystalli-ne, no coating) cut showed a mild localizedexpression in the fibrinous matrix. Collagen 1could not be detected in any of the slices. BSP(Bone sialoprotein) showed in one cut a locali-zed mild expression (crystalline, BMP-coating),another one a localized moderate expression inthe fbrinous matrix. For ON (osteonectin) inonly one cut a localized mild expression in thefibrinous matrix was observed. For OP (oste-opontin) no reddish coloration was visible.Implants prior to implantation: The surfacestructure has been characterized before andafter coating. Generally, the amorphous im -plants appeared slightly smoother than the cry-stalline ones. Implants at 7 days: All the 4 im -plant types showed similar tissue componentson their surfaces. These comprised fibrinousstructures between bone and the implant ma -terial. In further enlarged photomicrographsone could find erythocytes in this fibrinous net.Changes in the surface structure of the im -plants were only observed in amorphous typeswith and without coatings as cracks on thesurface.Procession of the TEM-specimens willstart in October.At present, evaluation of specimens prior toimplantation and at 7 days using LM, IHC andSEM as well were completed. The followingobservations have been made:1. Due to the implantation procedure a gap

healing was observed, leading to a delay inbone formation in comparison to formerstudies in the same animal model with bio-active implants. Up to 7 days after implan-tation no bone-bonding was observed inany of the specimens.

2. All of the materials (amorphous, crystallineeach with and without BMP-coating)displayed no obvious changes in surfacedensity and structure up to now indicatinga lower surface reactivity as known from45S5 bioactive glass.

3. At 7 days the highest expression of OC wasobserved at the amorphous coated im plants.

Therefore, stimulation in bone formationaround the BMP-coated implants is possible.Further evaluation of later time-points isnecessary, to assure this observation.

References [D1] Senta H, Park H, Bergeron E, Drevelle O,Fong D, Leblanc E, Cabana F, Roux S, GrenierG, Faucheux N, Cyt. Growth Fact. 20 (2009)213-222.

[D2] Fischer H, Koczur K, Lindner M, JennissenHP, Zurlinden K, Meißner M, Müller-Mai C,Seifert G, Oliveira A, Morawetz K, Gemming S,GeoDresden 2009, 30.09.-02.10.2009, Dres den.

[D3] Oliveira A, Seifert G, Gemming S, Proc.12th Intern. Symp. Biomater. Biomech, 17.-19.03.2010, Essen.

[D4] Seifert G, Oliveira A, Gemming S, Proc.12th Intern. Symp. Biomater. Biomech, 17.-19.03.2010, Essen.

[D5] Oliveira AF, Gemming S, Seifert G,BIOmaterialien (2010) submitted.

[D6] Oliveira AF, Gemming S, Seifert G, J. Phys.Chem. B, (2010) submitted.

[D7] Philips JC et al., J. Comp. Chem. 26(2005) 1781-1802; McKerell AD, et al., J. Phys.Chem. B 102 (1998) 3586-3616; Frishman D,Argos P, Prot. Struct. Funct. Genet. 23 (1995)566-579.

[D8] Scheufler C, Sebald W, Hülsmeyer M, J.Mol.Biol. 287 (1999) 103-115.

[D9] Gemming S, Enyashin AN, Frenzel, J,Seifert G, Int. J. Mater. Sci. 101 (2010) 758-64[D10] Luschtinetz R, Oliveira AF, Duarte HA,Seifert G, Z. Anorg. Allg. Chem. 636 (2010)1506-1512.

[D11] Kutschan B, Morawetz K, Gemming S,Phys. Rev. E 81 (2010) 036106.


34

HydraSmec – Understanding processes at thehot smectite-water interface for tailoring indu-strial bentonite applications

IntroductionThe adhesive properties of smectites are widelyused in many industrial applications. These pro-perties are mainly due to the reversible expan-sion and contraction of their interlayer spacesas a function of water activity. Being part ofmoulding sands, smectites are responsible forthe required mechanical strength of moulds.Due to the need of industry for cas tings ofincreasing complexity but decreasing weight, itbecomes essential to tailor moulding-sand mix-tures beyond their current abilities. So far,however, it has not been discovered in detailwhy changes occurring during the casting pro-cess are partially reversible in the laboratory, butnot in the circuit of the moulding sands. Ingeneral at temperatures below 300 °C and inthe laboratory, de- and rehydration are reversi-ble processes. It is therefore important to un -der stand the complex kinetics of de- and rehy-dration of smectites and their influence on themechanical behaviour of the moulding sands,before improvements can be achieved. Smectites are also widely used as adsorbents,

e.g., for water. Lower water adsorption capa-cities and reduced adsorption rates of industri-ally-dried bentonites compared with bentoni-tes dried to the same water content in thelaboratory show that the kinetics of dryingapparently influences interface processes.Smectites exposed to hot water vapour do notfully rehydrate in contrast to dry-heated smec-tites, which is another important aspect. Dueto contact angle measurements, increasinghydrophobicity could be observed after vapourtreatment, which, however, might depend onphysical changes of aggregation (pore volu-me). Nevertheless, other parameters such asCEC or X-ray diffraction patterns of the 00l-reflexes did not show any conspicuous chan-ges. Initial investigations led to the assumptionthat coordination of Al3+-Ions can be maderesponsible for these processes. However thishas to be confirmed. Not only the hydrationenergy of the cations, but also size and chargeas well as the water to smectite ratio, achieve-ment/attainment of dispersion, speed of dryingand further variables have an influence on the

Stanjek H. (1)*, Meister D. (1), Steinkemper U. (2), Wolff H. (2), Diedel R. (3), Habäck M. (3), Grefhorst C. (4),

Böhnke S. (4), Schmidt E. (5), Latief O. (5), Schmahl W. (6), Jordan G. (6), Eulenkamp C. (6)

(1) Clay and Interface Mineralogy, RWTH Aachen University (CIM), e-mail: [email protected]

(2) Institut für Gießereitechnik GmbH, Düsseldorf (IfG), e-mail: [email protected]

(3) Forschungsinstitut für Anorganische Werkstoffe- Glas/Keramik-GmbH, Höhr-Grenzhausen (FGK),

e-mail: [email protected], [email protected]

(4) S & B Industrial Minerals GmbH, Marl (S&B), e-mail: [email protected], [email protected],

[email protected]

(5) Stephan Schmidt KG (SSKG), Dornburg-Langendernbach, e-mail: [email protected],

[email protected]

(6) Department für Geo- und Umweltwissenschaften, Sektion Kristallographie, LMU München (LMU),

email: [email protected], [email protected],



35

dynamics of dehydration (and possibly rehy-dration).The objective of the current project is thedetailed examination of the aforementionedbasic mechanisms from atomic to industrialscale in order to understand them as well as tooptimize the casting processes. The followingis a summary of the experiments and first inter-pretations of the results carried out by the pro-ject partners.

Bentonite characterisationThe rather initial X-ray characterization of thebentonite samples done in 2008 has now beenadvanced for the raw bentonites (Table 1).

For testing purposes we included the 001 ofsmecite in the fitting procedure, whereas Uferet al. (2008) omitted this first peak. The ironcontents of the octahedral sheets were refinedunconstrained.

All bentonites have montmorillonite contents> 0.8 g/g. Mica and kaolinite were present insmall amounts except for bentonite C, wherethe kaolinite contents varied within two char-ges from 0.14 to 0.2 g/g. Calcite was detectedin bentonites D, E and W, respectively. The pre-sence of carbonates was also evident in theSTA/TG measurements.

The XRD contents of montmorillonite correlatewith the contents determined with methyleneblue sorption (Fig. 1), but at least one of themethods gives systematic deviations.

The mineralogical composition was checkedagainst the chemical analyses (Fig. 2). For allsamples Fe2O3 was overestimated by the mine-ral composition by a factor of two to three,whereas Al2O3 and SiO2 were correspondinglyunderestimated.

Obviously, the phase contents seem to be bia-sed especially for the iron contents. Therefore,

Table 1: Quantitative phase contents in g/g of all raw bentonites and some size fractions. Additional minor phas-es such as cristobalite, brookite, feldspars and others were omitted in this table.


36

Figure 1: Smectite contents determined by X-raydiffraction versus methylene blue adsorption.

Figure 2: Contents of Al2O3, SiO2 and Fe2O3 cal-culated from the Rietveld analyses and plottedagainst the contents determined by X-ray fluores-cence spectroscopy.

Figure 3: Plot of Fe contents constrained to twovariants of an equation (blue and green box) ver-sus Fe contents determined by freely refining theoctahedral site occupancies. Note that bentoniteC is a non-activated Ca-bentonite marked withCa in the figure (only green equation). The otherbentonites have Na in their interlayer space.


37

the fits were repeated by constraining theoctahedral iron contents to the cell edgelength b using the equation Fe = ( b – 8.972)/0.1174 (Köster et al., 1999). A significant cor-relation was found (Fig. 3), but the interceptseems not reasonable.

A more reasonable trend was obtained byincreasing slightly the constant 8.927 (greenbox in Fig. 3). Recently, MD calculations confir-med that the cell edge length b responds tothe kind of interlayer cation and the hydrationstate (Berghout et al., 2010). Their b values,however, where with 9.04 – 9.08 Å for thehydrated states even larger than the refinedones of our samples. Therefore, we artificallyselected 9.00 Å as the constant. The generaldiscrepancies between mineralogical and che-mical composition require more and detailedinvestigations.

Cast experimentsIn the first year of the project, optimization ofthe cast experiments were done (see Report2008). In 2009, three cast experiments wereperformed at the IfG. To get sand sampleswith the same »thermal history«, the roundconfiguration of the moulds have been chan-ged into a rectangular configuration. To impro-ve the sampling of the moulding sand, a splitflask was designed (Fig 4). This flask can beopened after the casting has finished (Fig. 5).In this way the samples can be taken across thewhole mould. Moulding sand, cast iron and casting tempera-ture have been defined previously: mouldingmaterial is quartz sand, type F32, mixed withone of five kinds of bentonite (~ 0.08 g/g) andwith water (~0.035 g/g). For the neutron dif-fraction experiments, pure bentonite was pla-ced adjacent to the hollow space filled later onwith the metal melt (Fig. 6).

Figure 4: Split flask

Figure 6: Cast model with wooden insert(removed before closing the form) and holesfor placing pure bentonite dedicated for theneutron diffraction experiments.

Figure 5: Detail of split flask


38

The five sampled zones ranged from 0-2, 2-3,3-4.5, 4.5-7, 7-10, and 10.5-17.5 cm, respec-tively (Fig. 7). The temperature profiles are vi su - alized in Fig. 8. Possible changes in the hydra-tion pattern were minimized by freezing thesamples in liquid nitrogen and keeping them ina freezer at -18 °C till further analyses.

Cation exchange capacitiesThe cation exchange capacities (CEC) weremeasured in a time series with time steps of15, 30, 45, 60, 120 and 180 minutes on theraw bentonite and on two moulding sand-mix-tures after the cast experiment. The objectiveof this CEC analysis was to find out, whether

heating causes clay interlayer contraction,which would result not only in decreased CECvalues at short exchange times, but also in adifferent exchange kinetic. It was alreadyshown in 2009 that the Cu-Trien exchangereached its final value already after 15 minutesexchange time. The zones, however, varied inthe absolute amounts of the CEC. Meanwhile, the smectite contents of the sandmixtures (FR32 and bentonite D) were determi-ned by XRD. Due to the low contents of smec-tite in the sand (0.08-0.10 g/g) the accuracy ofthe smectite content was checked on a purequartz sample. The Rietveld refinement of thispure quartz resulted in approximately 0.02 g/g

Figure 7: Casting boxshowing the iron castingand the bentonite holes.Note the zoning aroundthe cast iron. Sampleswere taken starting adja-cent to the cast iron (zone 1)and proceeding to wardsthe outer range (zone 5).

Figure 8: Temperature pro-files of casting experimentfrom 19.8.2009. The dataof zone 1 are not shownbecause the temperaturesensor failed shortly afterthe beginning of the castprocess. Note that zones 2to 5 remained for a certaintime at 100 °C and thenexperienced higher tempe -ratures, whereas zone 6and higher did not exceed100 °C.


39

smectite. This offset is mainly due to the over-lapping of smectite reflections with strongquartz reflections, where minor misfits in thepeak profile results in an overestimation ofsmectite. Therefore, the smectite contents ofthe sand mixtures were corrected by this bak-kground value.

Figure 9 shows that the smectite contentsdecrease with decreasing distance to the castiron except for the first zone, which differs alsoin the CEC value from the second zone. Sincethe smectite contents vary less than the CEC,the CEC normalized to the smectite contentshow a pronounced increase with increasingdistance (Figure 9).

Two aspects deserve attention: Although themost distant zone had a smectite content simi-lar to the initial material, the original CEC wasnot measured within an exchange time of onehour. It is also striking that no 10 Å phase wasdetected by XRD, although the diminishedCEC values could result from irreversible layercontraction. The current explanation for thisdiscrepancy is the possible influence of thekinetics of rehydration of possibly contractedinterlayers. The CEC values were determinedon material, which was defrosted immediatelybefore the analysis. The XRD analyses, howe-ver, required grinding in ethanol (with somewater), which had to evaporate afterwards.

This treatment could have provided time forre-expansion of contracted interlayers. Thishypothesis can be checked by CEC measure-ments with exchange times much longer thanone hour.For all samples, the loss on ignition (> 1100 °C)was determined by STA. The results correlatewell with the findings from residual moisturedetection. The mass change is decreasing withdecreasing distance to the cast body and it cor-relates significantly with the smectite contentsfrom XRD (Fig. 10). The regression line, howe-ver, does not intercept next to zero. A theore-tical calculation shows that the first two sam-ples could have released only structural OHgroups, whereas the samples of zones 3 to 6and the initial sample contained additionalinterlayer water (see the dividing green line inFig. 10). Furthermore, about one third of thesmectite has obviously transformed into aphase, which escaped quantification by XRD. Itis tempting to assign this amount to the cis-trans character of bentonite H: The current fit-ting resulted in 0.7 cis- and 0.3 transvacantlayers. This agrees with the maximum tempe-ratures reached in the different zones (Fig. 8).In zone 1 a partial dehydroxylation was likelybecause of Tmax > 600 °C. In zone 2, whichreached 600 °C, the remaining smectite retai-ned its theoretical OH content. The next zoneshad T < 350 °C, which seems too low for adehydroxylation, but it should be noted that

Figure 9: The cation exchangecapacities are normalized to thesmectite content from XRD. Notethat even the outermost zone doesnot reach the value of the initialmixture (blue line). Bentonite H,experiment 19.08.2009.


40

for kinetic reactions a too low temperaturecan be balanced (at least, in parts) by longerreaction times. This hypothesis remains to bechecked.

Neutron radiographyTo investigate the dehydration of the mouldingsand (pore water, smectite interlayer water) indetail with neutron radiography, we designed aspecial casting simulation experiment to in-situvisualize the loss of water. Instead of pouring ahot metal melt into a casting mould, the experi-mental set up consists of a casting mould (Fig. 11) which is dropped on a hot copper block(Fig. 12). The design of the sand box with a Cu-bottom plate ensures an ideal heat transfer cor-responding to the thermal shock-like heatinduction during a real casting process. Thermo -coup les were placed within the moulding sandto provide simultaneously information on thetemporal temperature gradients. They were setin direct contact and in one, two, three, fourand five centimeter distance to the Cu-bottomplate (Fig. 13). The experiments have been per-formed in the research reactor FRM II inGarching with the instrument ANTARES.

In a first set of neutron radiography experimentswe tested the functionality of the simulationexperiment, the sand box charging and moul-ding material rehydration conditions. Also possi-ble sand thicknesses and different bentonitecontents were checked to achieve the best con-

trast in the radiographs.The experiments allowed us to successfully si -mulate the shock-heating of the mould ma -terial in an industrial casting process. We wereable to visualize fluxes of water during the de -hydration process with high temporal and spa-tial resolution. But the experiments revealed al -so that the moulding material was neither ho -mo geneously loaded nor homogeneously re -hy drated. Zones of different density and withdifferent water contents were visible in theradiographs.For a better data evaluation a second set ofexperiments was necessary, where we imple-mented the following modifications:– optimization of the homogeneity of the

mould material (an improved compactionmechanism/procedure while loading the sandmould was essential to avoid zones of diffe-rent density within the moulding material).

– improved rehydration process (right afterthe addition of the water, which was lost bydehydration, the sands were homogenizedfor 3 minutes with a blender).

– utilization of thermocouples with higherresolution at fixed reference positions in allexperiments (allowed a precise comparisonof the thermal parameters in the differentexperimental runs).

We performed five evaluable experiments(compare table 2). The mould thickness of 7cmand a bentonite content of 12 wt% were keptconstant in all experiments. The casting mould

Figure 10: Plot of mass changesdetermined with SimultaneousThermal Analysis (STA) versus smec-tite contents of the moulding sandsdetermined by XRD. The green linedemarks the theoretical mass loss ofa fully dehydrated Na smectite.


41

was dropped onto the hot Cu plate when~650 °C were reached. The charging of themould and compaction of the mould materialwere carried out in the same way to ensure thesame homogeneous density conditions for allexperiments. The experiments were focusedon the following aspects:– effect of multiple cycles of previous dehy-

dration/rehydration on the performance ofthe moulding sand

– effect of 2wt% carbonaceous material as afunctional additive to the moulding sand

– performance of different types of bentonite(W: natural Na-bentonite, D: soda activatedCa-bentonite)

The radiographs (Fig. 14) show in general theremoval of water from the mould. The processinitiates at the heat source at the bottom. Thegrey scale depends on the water amount inthe moulding sand. The dehydrated materialappears light-grey while the hydrated material isdark-grey. The experiments revealed a progres-sive movement of water in the sand and resol-ved a broad transitional zone from the pristinehydration state of the sand to a fully dehydratedstate. At this transitional zone positions can bedetermined which on one hand relate to theonset of pore water dehydration and on theother hand relate to the completion of interlay-er dehydration.

The evaluation of radiographic experimentsrevealed the following results:The optimization of the loading procedure andrehydration process was successful and repro-ducibly yielded a homogeneous and equallydistributed moulding material in the sand box.The temperature-time data at fixed points ofreference in the different experiments showedthat the thermal history of all experiments mat-ched (within the remaining data error) and thatthe thermal parameters (conductivity, heatflow) in different sands were constant (Fig. 15).A constant compaction and homogeneity ofthe different sand moulds in the different expe-rimental runs can, thus, be inferred and thekinetic dehydration parameters of differentexperimental runs were directly comparable.

Figure 11: Top view of sand box

Figure 13: Thermocouples placed in different depths

Figure 12: Casting simulation experiment


42

Table 2: Performed experiments and used moulding sands.Exp.: experiment numbera: bentonite type.b: additive (2 wt% of carbona-ceous material).c: condition of moulding sand prior to radiographic experiment.* Experimental data were not evaluable.

Figure 14: Two Neutron radiographs showing the dehydration of moulding sand.

Figure 15: Comparison of temperature developments of all experiments at ~1 cm distance to the bottomCu-plate. The thermal history of all experiments matches within the remaining data error.


43

The evaluation of the amount of water in thedifferent moulds (represented by image greyscales as a measure of the intensity of neutronscattering) versus distance from the hotplate atthe bottom, time, and temperature revealed:– In all runs, the general behavior was that the

dehydration started with a vaporization ofpore water at about 100 °C. On top of thisrange, an enrichment of water in the areas<100 °C could be observed. This wateraccumulation may be caused by steampushing pore water upwards and conden-sing in the areas <100 °C.

– Dehydration was not completed at 100 °C.A decrease of neutron scattering intensitycould be observed up to about 170 °C. Thisdecrease very likely correlated with therelease of water from the clay interlayers viathe pore system (Fig. 16).

– The comparison of raw and recycled sandsrevealed that from recycled sands morewater was released around 100 °C while thetotal amount of water was almost constantin both types of experiments. As a conse-quence the recycled sands were dryingslightly faster.

– Sands with carbon containing additivesscat ter the neutron beam much stronger attemperatures >100 °C than additive free

sands – although scattering <100 °C isalmost identical. These differences may in -dicate interaction of C with water or hydro-gen >100 °C which retard the H-release.

The amount of material obtained from theradiography experiment was only sufficient formineralogical analyses, but not for the mecha-nical tests such as compactibility. Therefore, anew experimental setup was designed, whichwill allow to mimick the radiography experi-ment on a much larger scale. It consists of fire-clay bricks and a copper plate, which is heatedto 600 °C in a furnace. Then the bricks are pla-ced in a frame. On top of them the copperplate rests. Both layers are kept at the desiredtemperature by blow torches placed below(Fig. 17). The flask filled with moulding sand isthen placed on the copper plate. The tempera-ture evolution in the mould is monitored bythermocouples. After obtaining the desiredtemperatures in the moulding sand, the flaskcan be removed quickly from the copper plateand cooled down for subsequent sampling.

Neutron powder diffractionThe raw and the recycled moulding sands withbentonite W and WC show d (001) values ofabout 19 Å (Fig. 19), but subtle differences in

Fig. 16: Image grey scale, representing the amount of water, versus temperature measured at ~1cm distance to Cu-bottom plate. The recycled samples show the tendency to dry slightly fasteras the corresponding raw material, more water is released at 100 °C.


44

peak shape indicate the presence of higher d-values in recycled smectites. These differen-ces might be related to the different waterrelease visible in the neutron radiographs.Dehydrated smectites, directly measured afterthe casting simulation, reveal d values of about12.3 Å. The differences in the hydration states

of bentonite D containing moulding sand issignificant (Fig. 20). Again the recycled sandwith ad (001) maximum at about 18.8 Å showshigher d values than the raw material.

Figure 17: Scheme of the heating system Figure 18: Pilot test of the heating system

Fig. 19: Neutron powder diffraction pattern of moulding sand with 12%wt Bentonite D (mea-sured at DMC, PSI).


45

Industrial dryingFollowing the initial characterization of severalbentonites, two raw materials were selectedfor further testing. Quantities of about 100 tonseach were activated with 2.5 % soda and sto-red on a stockpile. The homogeneity of theraw material mixture was documented by sam-pling and analysing chemical and mineralogicalcontents. Samples taken before milling consi-sted of large chunks of clay, which had a pro-nounced variability in their mineralogical com-position (Table 3).

Milling of these clays was done in a MPS 100,a cylinder ball mill. Establishing constant mil-ling condition required the processing of about

20 tons of material. The mill inlet was fed withpre-mixed material, which was pre-dried toabout 30 %-wt. water content. Coarse mate-rial such as rock fragments or too large aggre-gates were removed during the mil passage.Monitoring of the size distributions and thetemperatures at various times showed that theoutgoing material had residual water contentsof less than 7 %-wt. Moreover, temperaturesreached 110 °C, which is above the recom-mended maximum temperature of 90 °C fordrying bentonites. The different particle sizeclasses 04 and 07 differed substantially in theirwater uptake as measured by Enslin-Neff.Furthermore, the swelling volumes of class 04is significantly higher than that of class 07.

Figure 20: Neutron powder diffraction pattern of moulding sand WC with 12%wt bentonite and 2 wt% C (measured at SPODI, FRM II) show different hydration states.

Table 3: Mineralogical analyses of three clay chunks (S.D.: Standard deviation, CV: coefficient of variation).


46

SummaryIn the research project HYDRASMEC small-scale industrial experiments provided samples,which were used to identify the potential pro-cesses, which lead to differences between hotsteam interaction in industrial applications ver-sus laboratory procedures. X-ray diffractionmeasurements in combination with time-resol-ved CEC measurements revealed that thesmectites dehydrate depending on the tempe-rature and water pressure regime, but showalso rehydration dynamics, which have notbeen fully investigated yet. The kinetics of thedehydration process has been assessed withneutron radiography, which allowed to esta-blish a two-dimensional time-position charact-erization of the water redistribution process ina mould sand. Conditions for posible steaminteraction have also been identified in theindustrial drying process.

ReferencesBerghout, A., Tunega, D. and Zaoui, A. (2010)Density functional theory (DFT) of Na/Mg/Ca/Sr/Ba-exchanged montmorillonites. Clays andClay Minerals 58, 174-187.

Bickmore, B.R., Hochella, M. E. Jr., Bosbach D.,and Charlet L. (1999). Methods for performingatomic force microscopy imaging of clay mine-rals in aqueous solutions. Clays and Clay Mi ne -rals, 47, 573- 581.

Köster, H. M., Ehrlicher, U., Gilg, H. A., Jordan,R., Murad, E. and Omnich, K. (1999) Mine ra -logical and chemical characteristics of five non-tronites and Fe-rich smectites. Clay Minerals34, 579-599.

Ufer, K., Stanjek, H., Roth, G., Dohrmann, R.,Klee berg, R. and Kaufhold, S. (2008) Quan ti -ta tive phase analysis of bentonites by theRiet veld method. Clays and Clay Minerals 56,272-282.


47

The impact of mineral and rock surface topo-graphy on colloid retention

AbstractOverall goal of this GEOTECHNOLOGIEN pro-ject called »The impact of mineral and rocksurface topography on colloid retention« is asystematic approach to determine, characteri-ze, and quantify the interaction betweenrough mineral surfaces and colloidal particles.A previous systematic study using pitted calci-te single crystal surfaces yielded results aboutthe general correlation between surfaceroughness parameters and the efficiency ofcolloidal retention under electrostically unfavo-rable conditions (Darbha et al. 2010). Here wepresent results about (i) the controlled varia-tion of colloidal retention at well-defined, arti-ficial surface structures, (ii) the characterizationand quantification of sources and sinks ofnatural particles in a quasi-closed system thatis used as a test site for field experiments, and(iii) the quantitative experimental results aboutthe retention of particles on natural rock surfa-ces that allows for a predictive approach tonatural systems. In detail, the following resultsare discussed in this report:(i) As an analog to rough mineral surfaces

with halfpores in the submicron size, thedeposition behavior of latex colloids wasstudied on a regular pit pattern (pit diame-ter = 400 nm, pit spacing = 400 nm, pitdepth = 100 nm). Effects of hydrodynamicsand colloidal interactions in transport and

deposition dynamics of a colloidal suspen-sion were investigated in a parallel plateflow chamber. The experiments were con-ducted at pH ~6.6 under both favorableand unfavorable conditions (in terms ofelectrostatic forces) using carboxylate func-tionalized colloids to study the impact ofsurface topography on particle retention.The influence of particle diameter variation(0.3-2 µm) on retention of monodisperse aswell as polydisperse suspensions as a func-tion of flow velocity over a wide range wasstudied. The impact of surface topographydeviations was found to be more significantfor smaller colloids (0.3 and 0.43 µm).Larger colloids (1 and 2 µm) beyond a criti-cal velocity of 7 x 10-5 and 3 x 10-6 m/s tendto detach from the surface irrespective ofthe impact of roughness since drag forcesexceed adhesion forces. For polydispersesuspensions, an increase in both polydi-spersity and fluid flow velocity resulted inthe decrease of colloid deposition efficiencydue to enhanced double layer repulsions.Hematite colloids of quasi-spherical shapewith diameters of about 950 nm showed ahigher deposition flux compared to spheri-cal latex colloids of equivalent size. Theseexperimental results provide quantitativeconstraints for the prediction of particleretention as a function of fluid-flow

Fischer C. (1,2)*, Darbha G.K. (1), Michler A. (1), Schäfer T. (3), Lüttge A. (2)

(1) Georg-August-Universität Göttingen, Abt. Sedimentologie/Umweltgeologie, Goldschmidtstr. 3, D-37077

Göttingen, Email: [email protected]

(2) Rice University, Dept. of Earth Science & Dept. of Chemistry, 6100 Main Street, Houston, Texas 77251-1892,

U.S.A., Email: [email protected]

(3) Institut für Nukleare Entsorgung,Forschungszentrum Karlsruhe, Postfach 3640, 76021 Karlsruhe,

Email: [email protected]



48

velocity, polydispersity, and particle shapeand size.

(ii) The investigation of the source (iron sulfide-bearing rocks), the processes of particlegeneration (oxidative weathering, acidicmine water conditions, mine water mixing),the particle transport, and the deposition ofparticles at water-rock interfaces within thewell-defined system of an abandoned alumslate mine enables a first-order approach forbalancing of particle retention at rough sur-faces in nature. The quantification of particleconcentration in the mine outflow as well asthe quantification of the deposited materialenables an estimation of the kinetics of par-ticles retention within such a system. Forcomparison, experiments were conducted toget laboratory results about the kinetics ofparticle retention using substrates with simi-lar ranges of topography variations.

(iii) Colloid adsorption experiments at naturalrock samples were performed using blackslate and micrite limestone as substrates.Micrite samples were etched to obtain vari-ances in surface roughness and topogra-phy. Generally, surface steps at grain boun-daries are the major cause for the overallroughness variations. A positive correlationbetween surface roughness of micrite sur-faces and the surface-normalized density ofadsorbed particles was found. For surfacesections < 20 µm, a remarkable dispersionof the surface-normalized particle densitywas found. We conclude that the variancein surface reactivity responsible for therecognized variance in particle deposition iscaused by the density of grain boundariesof the rock material. An analog experimentusing a grid pattern with varying grid linedensity confirmed this conclusion: A strongcorrelation between line density and adsor-bed particle density was found. The resultspresented here show the quantitativeimpact of surface topography variations oncolloid retention under electrostically unfa-vorable conditions. An important applica-tion of these results is their utilization as aninput parameter for predictive approachesto the fate of colloids in the environment.

1. IntroductionUnderstanding the colloidal particle depositionmechanism at fluid-solid interfaces is impor-tant to address a variety of environmental andindustrial processes. Important examples arethe protection of particle adsorption in semi-conductor manufacturing (HEROUX et al.,1996), the control of colloidal deposition (suchas polypeptides) in bio-engineering (ZHENG etal., 2004), the transport of pathogenic micro -organisms (BALES et al., 1991) and heavy me -tals (BAUMANN et al., 2006) in the aquaticsystem, as well as the prediction of colloid for-mation and deposition in nuclear waste reposi-tories (SCHAFER et al., 2004; SCHAFER et al.,2009). Numerous studies have been conduc-ted to investigate the significant factors gover-ning the particle retention mechanisms in thesubsurface environments focusing on fluidphase composition (PRESCOTT et al., 2002;SCHALDACH et al., 2006), the properties ofcolloids and media (ZHUANG et al., 2005), andphysical and chemical conditions of the flow(JOHNSON et al., 2007; JOHNSON and TONG,2006; TONG and JOHNSON, 2006; TORKZA-BAN et al., 2007).The transport and deposition of colloidal parti-cles in porous media such as sediments or sedi-mentary rocks under unfavorable conditions isof great importance to understand the captureof particles in the environment (BRADFORDand TORKZABAN, 2008; CHANG and CHAN,2008). Although there are various hypothesisproposed in the literature to explain the devia-tion from the traditional classical filtration the-ory (BRADFORD et al., 2003; BRADFORD andTORKZABAN, 2008), until now the fate of col-loid transport and retention is still poorlyunderstood. As a consequence, the quantitati-ve prediction of particle retention in nature (atmineral, rock and soil surfaces) is still almostimpossible. In many natural processes, thedeposition of colloids on surfaces is explainedby chemical and charge heterogeneity of sur-faces, impurities etc. (FILBY et al., 2008). Un -der unfavorable conditions, however, whererepulsive electrostatic forces between mineralsurface and particles prevail, an increase of theimportance of roughness for deposition of col-


49

loids was concluded and the quantitativeimpact of surface roughness has to be inclu-ded into predictive considerations (ALONSO etal., 2009; DARBHA et al., 2010). This is alsounderlined by the fact that a multitude ofminerals (both sedimentary grains and mineralcolloids) carry a negative surface charge underneutral pH conditions, that results in the com-mon situation of electrostatically unfavorableconditions (KOSMULSKI, 2002; KOSMULSKI,2004; KOSMULSKI, 2006; KOSMULSKI, 2009).Although theoretical and experimental investi-gations suggest an enhanced colloid deposi-tion with increasing surface roughness, it isoften disregarded. Recent experimental workconducted at field sites (GECKEIS et al., 2004;MORI et al., 2003) and in the laboratory (DELOSet al., 2008) in the context of colloid mediatedradio nuclide migration depicted the deposi-tion of colloids at rough rock and mineral sur-face under favorable conditions. More detai-led, batch sorption experiments were perfor-med to study the interaction of gold nanopar-ticles with rough granite interfaces (ALONSOet al., 2009; ALONSO et al., 2007). The expe-rimental comparison of favorable vs. unfavora-ble electrostatic conditions showed non-negli-gible colloidal retention at the crystalline rockunder unfavorable electrostatic conditionsinfluenced by surface roughness. Results ofexperiments performed by Das and co-authorsdemonstrated that surface roughness providesa large enough restraining torque in predictingthe hydrodynamic detachment of particles(DAS et al., 1994). In the experiments carriedout for colloid straining and filtration in satu-rated porous media, the surface roughness ofthe grains caused an increased colloidal reten-tion due to enhanced collision efficiency by afactor of 2 to 3 (AUSET and KELLER, 2006).Calculations of DLVO interaction energy bet-ween a sphere and simulated membrane sur-faces predict the significant reduction in theenergy barrier with increasing surface rough-ness (HOEK and AGARWAL, 2006; HOEK etal., 2003). Most of the theoretical work isbased on interfacial study between positiveprotrusions and spherical colloids where theparticle is several orders of magnitude larger in

size compared to the asperity (BHATTACHAR-JEE et al., 1998; MARTINES et al., 2008). Theimportance of interaction between a surfacewith negative asperities (pits/holes/etch pits)was investigated in context to the recent collo-id probe technique experiments on a roughmembrane composed of peaks and valleys(BOWEN and DONEVA, 2000). While resultspredict that peaks have a decreased magnitu-de of EDL repulsions compared to valleys, theadhesion rate was found to be higher for val-leys compared to peaks (BHATTACHARJEE etal., 1998; HOEK et al., 2003; MARTINES et al.,2008). Furthermore, experimental results showthat the van der Waals (vdW) contact interac-tions of a Polystyrene latex sphere and a roughsilicon substrate reduced the interaction ener-gy by 90% because of the effective larger sepa-ration distance between them (COOPER et al.,2000). This is according to a recent colloiddeposition study on a rough mineral surface(etched calcite) were etch pit walls were the pre-ferred adsorption sites (DARBHA et al., 2010).Moreover, the deposition rate was shown to bea function of etch pit depth and density.In the light of these observations and calcula-tions, deeper insight into the interaction bet-ween particles and rough surfaces is required.The questions how surface roughness varia-tions of the substrate can influence the colloi-dal deposition, what fundamental processesare related to the removal of fine particles fromrough surfaces in a hydrodynamic flow, as wellas what are the preferred sites for particleadsorption related to substrate surface rough-ness or charge variations are still unanswered.More specifically, the interplay between (i)applied hydrodynamics, (ii) the variation of col-loid size and polydispersity, and (iii) substratesurface topography variation (roughness)during the complex deposition process requiresdeep quantitative insight.

Here we report the results of our investigationsduring the second year of the GEOTECHNO-LOGIEN project: »The impact of mineral androck surface topography on colloid retention«.In detail, this overview provides informationabout the following investigations:


50

(i) Influence of varying surface charge, hydro-dynamic conditions, and polydispersity onparticle retention on well-defined syntheti-cally surfaces (silicon wafer surfaces, APTESand oxidized) that mimic natural mineraland rock surfaces

(ii) Mineral vs. latex particle adsorption (iii) Natural colloids : Shape, size and aggrega-

tion as constraints to particle adsorptionexperiments

(iv) Natural rocks as adsorbents: The exampleof micrite limestone

(v) Influence of pit wall density variation onparticle retentions: An analog study

(vi) Outlook: The evolution of fluid-rock inter-faces and the formation of grain coatings

2. Methods and Material

2.1. Colloidal ParticlesCarboxylate polystyrene latex colloids (d= 0.3,0.43, 1, 2 µm) were purchased from PostnovaAnalytics. The hematite (a-Fe2O3) colloids weresynthesized by the forced hydrolysis of FeCl3(MATIJEVIC and SCHEINER, 1978; RAMING etal., 2002). For preparation, an aqueous sus pen -sion of 0.0315 M Fe2Cl3 · 6H2O in 0.005 MHCl was placed in a hot oven at 100°C for 1week. The resulting sample was allowed tocool and was centrifuged several times byresuspending the colloids to avoid any excessFeCl3 or other impurities. The size of the hema-tite colloids was 955 ± 96 nm.

2.2. Silicon wafer samplesSilicon wafers with (100) orientation were pur-chased from AMO GmbH. Wafers were clean-ed using acetone and isopropanol followed bypiranha acid. After a thorough rinsing in waterthe wafers left in air for complete oxidation for5 to 7 days. For surface functionalization withAPTES (Aldrich, USA), the substrates were cle-aned with piranha acid and rinsed with water.Subsequently, they were dried under a flow ofpure nitrogen. The dry substrates were thenimmersed in a 10 % APTES in methanol solu-tion for 60 min followed by thorough rinsing inmethanol and again dried under flow of nitro-

gen (VAN LOENHOUT et al., 2009). Layer thik-kness is ~0.9 nm and ~0.6 nm for SiO2 andAPTES, respectively (MITCHELL et al., 1994;VAN LOENHOUT et al., 2009). Typical dimen-sion of the silicon sample was 4 mm x 4 mm x0.68 mm.

2.3. Rock samplesSlate rock samples were cleaved and cut toobtain subsamples with dimensions of 4 mm x4 mm x 1 mm. Micrite samples were cut to asize of 10mm x 5mm x 5mm. The surface waspolished. Subsamples were etched for 60’ and120’. Resulting surface root-mean-square rough -ness (Rq) was 70 nm, 85 nm, and 100 nm,respectively.

2.4. Zeta potential and photon correlationspec troscopyZeta potentials and photon correlation spec-troscopy (PCS) measurements of the colloidswere performed using a ZetaPlus (Zeta Po ten -tial Analyzer, Brookhaven Instruments). TheSmoluchowski equation was applied to con-vert electrophoretic mobility measurements ofthe colloids to zeta potentials. The ζ-potentialmeasurements were performed at pH = 6.6 in10-2M NaCl. The respective zeta potentialvalues for 0.3, 0.43, 1, 2 µm were -52, -63, -45, -47 mV, respectibely. The constant averagediameters for the polydisperse colloid mixtu-re(s) from the PCS measurements showed thatsuspensions were stable during the duration ofthe experiments. The zeta potential of thehematite colloids was -25 ± 5 mV. The ζ-potential of both oxidized and APTESfunctionalized silicon wafer sample surfaces wasdetermined by a streaming potential analyzer(Anton Paar Surpass Electrokinetic Ana lyzer)with plane-parallel channel cell method at ionicstrength of 10-2 M NaCl solution. The point ofzero charge (PZC) for oxidized silicon surfaceswas found to be at pH ~ 3-4 and the PZC of theAPTES functionalized silicon surfaces was foundto be at pH ~8.7 (WU et al., 2006).

2.5. SEM analysis of natural particles aswell as adsorbed materialA Cryo-FE-SEM, FEI Quanta 200 FEG, with


51

Edax EDX detector was used to investigateshape and chemical composition of naturalparticles. Slurry of suspended particles was puton a glass sample holder and dried.

2.6. Laser Particle Sizer (LPS)Natural material was sampled from encrusta-tions. The solid samples were elutriated.Suspensions were separated into grain sizeclasses < 2 µm and ≥ 2 µm using the Atterbergmethod. The particle size distribution (PSD) ofthe resuspended colloids was analyzed with alaser particle analyser from Beckman & Coulter(Laser Particle Sizer (LPS) LS 13320). Analyzedparticle size range was 0.04 µm to 2000 µm.The particle analyzer applies the technology ofelastic light scattering (ELS) for particle sizecharacterization. The size calculation is basedon either Fraunhofer diffraction or the Mietheory. To get information of particle size in thesub-micrometer range a so called polarizationintensity differential scattering (PIDS) technolo-gy is implemented in the particle analyzer.Several optical models can be applied to opti-mize the PSD results, depending on the natureof the solid, the suspending fluid and thewavelength of the incident laser beam. Verycritical values are both the refractive index andthe adsorption coefficient. The physical con-stants were taken from the literature (CARPER,1999). In the case of particle aggregates, par-ticularly for the poorly crystalline Fe oxideshydroxides, an effective refractive index wascalculated after the suggestions in literature(GREGORY, 2009) and applied to the collecteddata. This approach takes into account thetypical difference between the refractive indexof a single solid particle and a particle aggre-gate. The aggregate may show a clearly lowerdensity than the single grain and hence alower refractive index. According to these con-siderations, the effective refractive index canbe obtained from the ratio (f, volume fraction)of the total solid volume of the aggregate andthe total volume of the aggregate (enclosingvolume):

D = fractal dimension, k = number of particles.For the iron oxides a fractal dimension (D) of2.5 was chosen and the particle number wasset to 50’000 for 500 nm spherical aggregates(according to Gregory, 2008). The volume weigh -t ed relative refractive index, mF, is given by

with m = relative real refractive index (singlegrain/suspension medium). So we obtain aneffective refractive index, RIeff

For the PSD analysis of the iron oxides a RIeff of1.47 is chosen. An adsorption coefficient of0.1 is taken. This is an appropriate value foriron oxide as well as for silicateaggregates(CARPER, 1999).

2.7. XRD analysis of natural particlesMineral particles from precipitation experi-ments as well as from resuspended grain coa-ting material were analyzed using a PhilipsX'Pert MPD diffractometer that was equippedwith a PW 3373/00 Cu LFF x-ray tube and aPW 3050/10 goniometer. The chosen scanrange for poorly crystalline iron oxides hydroxi-des was from 10 – 70° 2Θ with a step size of0.04° 2Θ, time per step was 20 s. Better cry-stallized material from encrustations were ana-lyzed with a scan range of 2 - 75° 2Θ, step size0.02° 2Θ, time per step 2 s.

2.8. ICP-OESThe concentration of Fe, S, P, Si, Al, Ca, Na, K,Mg, Mn, Zn, Cu, Mo of mine water samples andof precipitation samples was measured using aICP-OES (PerkinElmer, Optima 4300 DV).Filtration of the mine water and from the preci-pitation experiments was performed using a fil-ter with pore size of 0.025 µm. The filter cakeswere digested with HClO4 before analysis.

2.9. Flow-through experimentsThe colloidal deposition experiments using sili-con wafer samples were conducted in a rect-angular parallel-plate channel fluid cell madeof Teflon. The inner dimensions of the cellwere 60 mm x 10 mm x 3.4 mm. The substra-


52

te was oriented parallel to the flow direction.After use, the particle suspension was discar-ded. For comparison and in order to quantifythe influence of half pores on the substratesurface, the colloid deposition experimentswere conducted for both plane and rough sub-strates. Exposition time was 40 min. The Pecletnumber and the Reynolds number for theparallel plate flow chamber were calculatedbased on the following equations (BAKKER etal., 2002):

(Qpp: flow rate; rc: particle radius; w: width ofthe flow chamber; b: half-depth of a parallelplate flow chamber, Doo: particle diffusion co -ef ficient; ρ: fluid density, and v: viscosity)The flow rate in cell was adjusted to obtain therequired Peclet number. For experiments withindividual colloid deposition experiments theimplemented Peclet number range was 10-4 to1. For the here applied flow rates, the Reynoldsnumbers is in the range of 0.004 to 26.Adsorption experiments with micrite surfacewere performed in a Teflon flow-through cell.Particle concentration was adjusted to 24 x 106

and 24 x 107 particles/ mL in 10-2 M NaCl solu-tion and pH = 9 was adjusted using 0.01MNaOH. Before the micrite surface was exposedto a colloidal suspension, it was polished andsubsequently etched in a flow-through cell for60 and 120 min in 0.0004M H2SO4 solution(pH = 4.7) to produce a variance in surfaceroughness of the micrite samples. A systematiccalcite dissolution study by Lea et al. (2001)showed that dissolution kinetics is significantlyreduced by an increased carbonate concentra-tion (>900 µM) in the solution. Therefore, inorder to minimize the dissolution of calciteduring the deposition experiments, and toavoid the instability of the colloidal system dueto the released calcium into the system, a con-centration of 700 µM of NaHCO3 was added

to the colloidal solution. To sustain electrostically unvaforable condi-tions close to the PZC of calcite, the experi-ments were conducted at the close point ofzero charge of calcite (pHpzc=8.8-8.9) (SCHMIDTet al., 2009; SOMASUNDARAN and AGAR,1967). Experiments were performed at roomtem perature (22 °C).

2.10. Vertical Scanning Interferometry (VSI) The characterization and quantification of sur-face topography and roughness variations wascarried out using a ZeMapper vertical scanninginterferometer, manufactured by ZemetricsInc., Tucson, AZ. VSI is an optical surface scan-ning method that provides a large field of viewof up to several mm2 and a vertical resolutionof < 1 nm and a lateral resolution of about 200nm (LUTTGE et al., 1999). For our measure-ments we applied white-light interferometrymode as well as blue-light monochromaticphase-shift mode. Several Mirau objectiveswere utilized. SPIP software (Image MetrologyA/S) was used for surface data analysis as wellas for particle characterization and quantifica-tion. Statistical topography parameters, so-cal-led roughness parameters (e.g., DONG et al.,1992; DONG et al., 1993; DONG et al., 1994a;DONG et al., 1994b), were applied to evaluatesurface topography data obtained by VSI mea-surements. We applied converged roughnessparameters (FISCHER and LÜTTGE, 2007) to getinformation about size, frequency, and distribu-tion of surface building blocks responsible forvariations in surface topography variations.

2.11. Quantification of colloidial deposi-tion at calcite surfaces The individual number of colloids adsorbedwas counted using SPIP software datasets. Thekinetics of the colloidal deposition was deter-mined by calculating the dimensionlessSherwood number (Sh) (KLINE et al., 2008),

where ap is the radius of the particle, C0 is thebulk colloidal concentration, D∞ is the bulk dif-


53

fusion coefficient and, J is the deposition fluxand was found by determining the initial slopeof the number of deposited colloids versustime and the value was normalized to the areaunder consideration.

2.12. 3D computed tomography ofnewly-formed minerals in sandstonesA Phoenix x-ray Nanotom (180 kV / 15W nano-focus computed tomography (nano CT) system)was used to get spatial information about thedistribution of iron oxide encrustation in sands-tones. Sandstone samples were cut to a size of3 mm x 3 mm x 3 mm. Samples were mountedto a glass stick. Resulting voxel size of the obtai-ned data set was 3 µm x 3 µm x 3 µm.

3. Results and DiscussionThis chapter provides information about threemain parts of our research performed duringthe second year of this project period. The firstsection shows results from our experimentalwork about particle adsorption at well-defi-ned, machined surface structures (chapter3.1). The second section (chapter 3.1) providesinformation about both sources and sinks ofcolloids at a location used as a »natural labo-ratory« to study the fate of particles in theenvironment. Chapters 3.1 - 3.4 present andcompare results obtained from experiments

using both natural rock as well as well-definedmachined sample surfaces as adsorbent.

3.1. Impact of surface functionalization,surface topography, and particle size onparticle retentionFavorable conditions (APTES coated surfaces)vs. unfavorable conditions (silicon oxide layer):Regardless of particle diameter, the colloiddeposition flux is always higher for APTES coa-ted surfaces compared to a silicon oxide surfa-ces (Fig. 1). The simple reason for this result isthe existence of attractive vs. repulsive electro-static forces (positively charged silica surfacewith exposed amino groups vs. negatively char-ges silicon oxide surfaces) (KLINE et al., 2008).More interesting is the size-depending beha-vior of J as well as the difference in J for roughvs. smooth surfaces. A minimum of depositionflux J was found for 1 µm colloids (Fig. 1). Theslight increase in J for colloids with d = 2 µmvs. d = 1 µm could be caused by the commoninterception (Fig. 2). In the absence of an ener-gy barrier, for non-Brownian particles the gra-vitational force is not an important factor (ELI-MELECH, 1994). The 0.3 µm colloids experien-ced the highest ratio in J for smooth vs. roughsurface deposition (pit diameter = 400 nm, pitspacing = 400 nm, pit depth = 100 nm). The Jratio decreases with an increase in particle size.Thus, it can be noticed that the effect of surfa-

Figure 1: Comparison ofcolloid de position flux ofcarboxylated latex colloidsof different sizes (constantfluid flow velocity = 1.3 x10-5 m/s) for smooth andrough surfaces. The influ -ence of surface roughnesson particle retention is re -duced for larger colloidsand under electrostaticallyfavorable conditions.


54

ce roughness is more pronounced for smallercolloids compared to larger colloids. Com -pared to the classical filtration theory (CFT)(Fig. 2), the observed colloid deposition beha-vior is in partial agreement with the predictionsof CFT. Filtration theory does not consider thesurface charge effects, physical forces acting atthe interface, and substrate surface geometryfor calculation. The theory assumes that collo-ids were irreversibly retained in the primaryminimum of the DLVO interaction energy dis -tribution (BRADFORD and TORKZABAN, 2008;ELIMELECH and OMELIA, 1990; JOHNSON etal., 2007; JOHNSON and TONG, 2006). Hen ce,the here observed differences in size-depen-dant colloid deposition rates between experi-mental results and calculations according tothe filtration theory can be explained by theinfluence of surface topography deviations. According to fundamental studies about physi-cal forces acting during particle-surface inter-action (BRADFORD and TORKZABAN, 2008),several mechanisms may play a role either forattachment or detachment, such as rolling,dragging, lifting (Fig. 3). As reported in the lite-rature, rolling plays an important role to remo-ve a particle from the surface under laminarflow conditions. Particle rolling can be initiatedregardless of its size even at the smallest flowrates (i.e., at lowest Reynolds numbers) (BUR-DICK et al., 2001). In order to remove a parti-

cle from the surface, the hydrodynamic torqueapplied to the particle must exceed that of theadhesion torque (TORKZABAN et al., 2007).Unlike a smooth surface, a rough surface canchange the center of rotation and therefore incase of a rough surface a larger applied torqueis required to remove a particle from the surfa-ce (BURDICK et al., 2005). Because of the herediscussed influence of hydrodynamic forces onthe efficiency of particle retention, a detailedexperimental approach was performed toquantify the impact of fluid-flow velocity vari-ation on the deposition of particles at roughsurfaces. The results are reported in chapter3.6 and compared to the particle depositionresults using micrite surfaces.The impact of polydispersity on the efficiencyof particle retention at rough surfaces wasinvestigated. A general result is that an increa-se in polydispersity leads to a decrease in thedeposition efficiency. Moreover, as the polydi-spersity increases (from 0.3-0.43 µm vs. 1-2µm) the difference in deposition efficiency bet-ween smooth vs. rough surfaces decreases.This can be explained by the hydrodynamicinteractions between particles (i.e., the so-cal-led hydrodynamic blocking effect (ADAMCZYKet al., 1995; DABROS, 1989)). When a particleapproaches to an already adsorbed particle, itwill be redirected from the initial (shortest)path due to EDL repulsions. For our investiga-

Figure 2: Collision frequen -cy efficiency for a singlecollector at fluid velocity1.3 x 10-5 m/s, viscosity =8.9 x 10-4 Pa s, colloiddensity = 1.05 g/cm3, fluid density = 1 g/cm3.


55

tions, the hydrodynamic blocking was found tobe significant for rough surfaces under flowconditions with Pe > 1. The resulting shadowarea is a narrow zone with an extension in thesize range of the attached particle (VAN LOEN-HOUT et al., 2009). Hence, the electrohydro-dynamic blocking increases with an increase inparticle size. Hematite colloids were used for comparisonexperiments. The experiments were carried outto quantify differences in the adsorption beha-vior of similar-sized and similar-shaped parti-cles of different mineralogy. We observed anincreased deposition of hematite colloids com-pared to polystyrene was observed. Again, thealready discussed influence of roughness wasreported for both favorable and unfavorableconditions. For a given fluid velocity (1.3 x 10-5

m/s), the deposition ratios (rough vs. smoothsurfaces) under electrostically unfavorable con-ditions were calculated to be 5.47 (hematite)vs. 2.41 (polystyrene). A potential reason ofthe remarkable difference in retention efficien-cy could be the difference of the particles sur-

face topography (DRELICH, 2006; HOEK andAGARWAL, 2006; MARTINES et al., 2008).

3.2. Natural colloids in an alum slate minesystem

Precipitation of particlesThe origin, transport, and retention of ironoxide/hydroxides was studied in an abandonedalum slate mine. During oxidative weathering,iron sulfides are removed from the black, orga-nic-rich alum slate (FISCHER et al., 2009).Source for iron oxide colloids in this system isthe dissolved matter in the acidic mine waterof the Morassina mine, Thüringisches Schie fer -gebirge (FISCHER et al., 2007). At sample site»Oberer See«, the acidic water (pH ~ 2.5) con-tains relatively high concentrations of iron andsulfur. Precipitation experiments in the labora-tory using sampled mine water showed theoccurrence of first precipitations after increaseof pH to 2.8. Subsequently, the pH was incre-ased to 4.5. The precipitations consisted of a2-line ferrihydrite (2LFh) (Fig. 4). In the mentio-

Figure 3: Schematic profilediagram to visualize thecommon forces at a fluid-solid body interface. As anexample, a particle with d= 0.3 µm is shown at thestructured surface appliedfor the study reportedhere.

Figure 4: Precipitation experimentsusing mine water (pH = 2.5) sho-wed the occurrence of 2LFerrihydrite (Fh) as first precipita-tions after increase of pH to 2.8.Subsequently, the pH was increasedto 4.5. In this pH-range ferrihydriteis not stable with respect toschwertmannite (Sh).


56

ned pH-range (2.5 – 4.5), ferrihydrite (Fh) isnot stable with respect to schwertmannite(Sh). The XRD- and ICP-OES-analysis showedthat there is a slow (in the range of months)transformation from 2-line ferrihydrite to Sh.This transformation came along with an incor-poration of sulfate and hence with a decreasein pH. The initial precipitation of 2-line ferrihy-drite instead of the thermodynamically favoredschwertmannite was caused by high rates ofhydrolysis in the laboratory experiments (135µmol OH / mmol Fe / min). Under natural con-ditions however the direct precipitation of fer-rihydrite has not been observed in sulfate-richenvironments. This is caused by common slowhydrolysis rates in natural systems. Such hydro-lysis conditions may also prevail for the mostsections of the Morassina mine. Ullrich et al.(2005) showed that the dominant secondaryminerals inside the mine consist of schwert-mannite (Fig. 5) or XRD-amorphous iron-phos-phate-sulfate phases.

Particle coatings (encrustations)Iron oxide encrustations were sampled in amine water streambed. Encrustations consistof pure 2-line ferrihydrite. SEM images showthe spherical shape of 2-line ferrihydrite parti-cles (Fig. 6). Particle diameter is around 200 -500 nm. The spherical particles may also formaggregates with diameters of up to severalmicrons. Aggregates may have an isometric aswell as a platy shape. The XRD results show thatthe flaky-shaped particles do not consist ofsheet silicates. Particle size distribution was ana-lyzed using laser particle sizer (LPS). The resultsshow for virtually all subsamples a bimodal sizedistribution. The first size mode confirms thesize range of about 200 – 500 nm, visible inSEM images. The second size mode showshowever a large portion of aggregated colloidalparticles. According to the XRD results, the app-lied optical model to evaluate the data collectedby LPS was calculated for Fh aggregates. For adetailed discussion of calculation of the opticalconstraints for porous aggregates analyzed byLPS see Gregory (2009).

Figure 5: Schwert -mannite was for-med during preci-pitation experi-ments in the labo-ratory using sam-pled mine water(left: SEM image,right: LPS particlesize analysis)

Figure 6:Particles sampledfrom encrusta-tions formed inthe streambed ofthe minewaterof the Morassinaoutflow (left:SEM image of 2-line ferrihydrite(2LFh) particlesand aggregates,right: LPS parti-cle size analysis.


57

Currently, in-situ adsorption experiments areperformed in the streambed to get informationabout the adsorption kinetics of particles in themine water. Substrate material of differenttypes of lithographic structures on silicon sam-ples was exposed to the mine water. The in-situ experiments mimic the topographic rangeknown from natural rock surfaces (FISCHER etal., 2008). Due to the well-defined differencesin pit size, shape, and spacing the expectedresults will provide important constraints forthe evaluation of the results obtained by labo-ratory experiments.

3.3. Particle adsorption experiments: The micrite surfaceAdsorption experiments were performed utili-zing a suspension of latex colloids (d = 1 µm)in a flow-through cell. Experiments were car-ried out close to the PZC of calcite (pH = 8.9,see Darbha, 2010) under electrostaticallyslightly unfavorable conditions (pH ~ 9). Thethree applied micrite limestone samples sho-wed surface roughness values of Rq = 70 nm,Rq = 85 nm, and Rq = 100 nm, respectively.After adsorption, the spatial distribution of theparticle density on micrite surfaces was analy-zed as a function of field-of-view. The resultshows a relatively broad range (Fig. 7). Par ti -cularly, surface sections < 20 µm show a signi-ficant increase of the density range. The obser-

ved range was found for all of the roughnesstypes (Fig. 7). According to previous resultsabout adsorption at rough surfaces, the meanvalues of particle density (for large field-of-view sections) show a strong and positive cor-relation to the root-mean-square roughnessRq. The increase of density dispersion wasfound for all roughness types as a function ofthe decrease of the field of view. In fact, thereis also an increase of the dispersion range withincreasing roughness. The process of disper-sion is however not completely governed byroughness variations. We therefore concludethat the dispersion reflects an intrinsic proper-ty of the rock that could be related to the den-sity of grain boundaries of this very dense,nonporous rock. We also conclude that thisresult could act as a general explanation forthe spatial inhomogeneity of particle density atrock surfaces after adsorption under electro-statically unfavorable conditions.

3.4. An analog experiment: The impact oflattice-»constant« variation of a machi-ned, well-defined silicon surface on parti-cle retentionThe conclusion about the impact of grainboundaries that modulates the roughness vari-ations of the etched rock surface was checkedvia an analog experiment. A silicon wafer was

Figure 7: Surface area-nor-malized particle density onrough micrite surfaces as afunction of field-of-view.Note the broad dispersionof particle density for sur-face sections < 20 µm.


58

machined with pits of constant depth (100nm) and a step-wise variation of the spacing ofits pit pattern (Fig. 8). Two types of particles (d= 60 nm, d = 100 nm) were utilized for adsorp-tion experiments in a flow-through cell. Again,the number and surface-normalized density ofadsorbed particles was analyzed. As a firstresult, the overall amount and distribution ofadsorbed particles shows a negative correla-tion to the pit size (Fig. 9). Particle density wasalso quantified as a function of pit density, i.e.,the density of lines or pit walls on a given sur-

face section (Fig. 10). The positive correlationbetween the normalized particle density andthe surface line density provides a robust con-firmation the conclusion made in the previouschapter about the impact of grain boundarieson particle retention (Fig. 11).

3.5. Kinetics of particle adsorption on mi -crite, impact of particle concentration, andfluid-flow velocity on the deposition of par - ticles on a rough micrite surface

Figure 8: A silicon wafer was machined with pits of con-stant depth (100 nm) and a step-wise variation of thespacing of its pit pattern. Particle adsorption experimentsusing latex colloids (d = 60 nm, d = 100 nm) were per-formed using this sample as an analog for the variationin grain boundary density of mineral aggregates.

Figure 9: Surface area-normalized particle density (60 nmand 100 nm particles in pits, 60 nm and 100 nm parti-cles outside of pits) as a function of pit size of the analy-zed sample region

Figure 10: Surface area-normalized particle density as afunction of pit wall density (line density, ld)

Figure 11: Visualization of adsorbed latex colloids (brightdots) on a micrite surface. The particles adsorbed at grainboundaries of the calcite grain aggregates. The grainboundaries form surface steps (h ~ X0 nm-X00 nm)generated during calcite dissolution.


59

Repetitive experiments were performed utili-zing micrite surfaces with Rq = 85 nm in aflow-through cell to study the impact of para-meter variation on the efficiency of particleretention at a rough rock surface. Mean values of particle density data measuredat large field-of-view yielded 0.014 µm ±0.002 µm2 after duration of 60’ (Fig. 12). After120’, the density increased to 0.033 ± 0.002µm2. No saturation was found during thisshort period of adsorption. The preferred sitesof adsorption, i.e., the surface steps at grainboundaries are still almost unoccupied after anexposition period of 120’. The dispersion ofparticle density was however similar for bothreaction periods. This underlines again theimportance of specific sites that are importantfor the adsorption of particles on rock surfacesunder unfavorable conditions (Fig. 11).The variation in particle concentration of theapplied suspension has no congruent impacton the mean density of adsorbed particles: Theincrease of concentration by an order of mag-nitude resulted in an increase of the mean par-ticle density of just about 5. Also the densityrange is relatively small for small sections (l =10 µm). We conclude that the simple increasein particle concentration is not able to providea congruent increase in the particles surfaceimpact probability (see also discussion aboutthe shadow effect, chapter 3.1).

The variation in fluid flow velocity has a con-gruent impact on the mean density of adsor-bed particles. This is, however, an inverse cor-relation: The 10-fold increase in fluid-flowvelocity results in a reduction of the mean den-sity of adsorbed particles. Because of thisremarkable result the impact of fluid flow velo-city variation on the deposition of particles wasobject of a detailed study (see next section).

3.6. Impact of fluid-flow velocity on theadsorption of particles at rough surfacesA detailed experimental approach showed theimpact of fluid-flow velocity variation on thedeposition of particles at rough surfaces.Experiments were carried out using a regularpit pattern on silicon wafers (pit diameter =400 nm, pit spacing = 400 nm, pit depth = 100nm). Particle diameters were 300 nm and 500nm, respectively. The number of particles wit-hin pits and outside of pits was analyzed as afunction of fluid-flow velocity (Fig. 13). The

Figure 12: Adsorption of 1µm latex colloids on roughmicrite surfaces. Surface area-normalized particle densityis shown as a function of field-of-view. Exposition timewas 60’ and 120’, respectively.

Figure 13: Surface-normalized particle density as a func-tion of fluid-flow velocity. Adsorption experiments werecarried out using regular pitted silicon wafer samples andlatex colloids (d = 300 nm, d = 500 nm)


60

impact of fluid-flow velocity on the efficiencyof particle retention is very different for parti-cles inside vs. outside of pits. Particles outsideof pits attached close to pit walls.For small particles (d = 300 nm) deposited inpits, a positive and linear correlation wasfound between particle density and fluid-flowvelocity (Fig. 13). Particles attached to the sur-face outside of pits show however a morecomplicated behavior as a function of fluidflow-velocity: Both particle sizes show an initi-al positive correlation. Subsequently, a satura-tion plateau was found. Here, a further increa-se of fluid-flow velocity does not result in afurther increase of particle retention. Above acritical velocity, the net particle retentiondecreases as a function of fluid-flow velocity.Further increase of fluid-flow velocity results inno net deposition of particles. The fluid-flowrange that results in the reported particle den-sity plateau is similar but not equal for bothparticle types. The results about particle adsorp-tion as a function of fluid-flow velocity discus-sed in the previous section represent two pointsof the complicated function presented here,the plateau situation (lower velocity) as well asthe negative correlation (higher velocity).

4. Outlook: The evolution of fluid-rockinterfaces and the formation of graincoatingsThe results presented here show a variety ofinput parameters that are able to govern theretention of colloids at rough surfaces underelectrostatically unfavorable conditions. Know -ledge and application of the parameter rangesinvestigated here enable quantitative predic-tions of particle retention in nature and allowfor a prediction of dominating processes dur -ing fluid-rock interaction. The example of sandstone diagenesis (Fig. 14)provides an example how the evolution of par-ticle coatings at fluid-rock interfaces governsthe further diagenetic evolution. For the exam-ple shown here, the formation of thick ironoxide coatings on sand grains hinder the preci-pitation of high-volume cement minerals, suchas calcite or quartz. During burial diagenesis,the iron oxide-rich sandstone experienced the-refore a higher reduction of the intergranularvolume compared to the sandstone withoutiron oxide coatings. The quantitive data about particle retention maytherefore also provide predictive conclusionsabout the evolution of fluid-rock interfaces.

Figure 14: Example of twotypes of grain coatings insandstone. Bright laminae(1) consist of grains withclean surface, high inter-granular volume (IGV),and high amounts ofcements. Hematite graincoatings (2) are responsi-ble for a different diage-netic evolution that leadsto remarkable lower IGV.The upper image is anoptical micrograph.Images (1) and (2) werecollected using an industri-al 3D computed tomogra-phy system (Phoenix x-rayNanotom)


61

5. AcknowledgementsThe authors thank André Filby and JohannesLützenkirchen (Institut f. Nukleare Entsorgung(INE), Forschungszentrum Karlsruhe, Ger ma -ny), Tim Salditt and Mike Kanbach (Institut fürRöntgenphysik, Univ. Göttingen), MichaelSeibt and Volker Radisch (Institut für Halblei -ter physik, Univ. Göttingen), Stephan Herming -haus and Matthias Schroeter (MPI für Dynamikund Selbstorganisation, Göttingen), Rolf S.Ar vidson and Everett Salas (Rice Univ.,Houston) for analytical help and fruitfuldiscussions. We thank for financial support(grant # 03G0719A) coming from GEOTECH-NOLOGIEN R&D program.

6. ReferencesAdamczyk, Z., Siwek, B., and Szyk, L., 1995.Flow- Induced Surface Blocking Effects inAdsorption of Colloid Particles. J Colloid InterfSci 174, 130-141.

Alonso, U., Missana, T., Patelli, A., Ceccato, D.,Albarran, N., Garcia-Gutierrez, M., Lopez-Tor -rubia, T., and Rigato, V., 2009. Quantificationof Au nanoparticles retention on a heteroge-neous rock surface. Colloid Surface A 347,230-238.

Alonso, U., Missana, T., Patelli, A., Rigato, V.,and Ravagnan, J., 2007. Colloid diffusion incrystalline rock: An experimental methodologyto measure diffusion coefficients and evaluatecolloid size dependence. Earth and PlanetaryScience Letters 259, 372-383.

Auset, M. and Keller, A. A., 2006. Pore-scalevisualization of colloid straining and filtrationin saturated porous media using micromodels.Water Resour Res 42, -.

Bakker, D. P., Busscher, H. J., and van der Mei,H. C., 2002. Bacterial deposition in a parallelplate and a stagnation point flow chamber:microbial adhesion mechanisms depend on themass transport conditions. Microbiol-Sgm 148,597-603.

Bales, R. C., Hinkle, S. R., Kroeger, T. W., Stock -ing, K., and Gerba, C. P., 1991. BacteriophageAdsorption during Transport through Porous-Media – Chemical Perturbations and Rever -sibility. Environ Sci Technol 25, 2088-2095.

Baumann, T., Fruhstorfer, P., Klein, T., andNiessner, R., 2006. Colloid and heavy metaltransport at landfill sites in direct contact withgroundwater. Water Research 40, 2776-2786.Bhattacharjee, S., Ko, C. H., and Elimelech, M.,1998. DLVO interaction between rough surfa-ces. Langmuir 14, 3365-3375.

Bowen, W. R. and Doneva, T. A., 2000. Atomicforce microscopy studies of membranes: Effectof surface roughness on double-layer interac-tions and particle adhesion. J Colloid Interf Sci229, 544-549.

Bradford, S. A., Simunek, J., Bettahar, M., VanGenuchten, M. T., and Yates, S. R., 2003.Modeling colloid attachment, straining, andexclusion in saturated porous media. EnvironSci Technol 37, 2242-2250.

Bradford, S. A. and Torkzaban, S., 2008. Col lo -id transport and retention in unsaturated po -rous media: A review of interface-, collector-,and pore-scale processes and models. VadoseZone Journal 7, 667-681.

Burdick, G. M., Berman, N. S., and Beaudoin,S. P., 2001. Describing hydrodynamic particleremoval from surfaces using the particle Rey -nolds number. Journal of Nanoparticle Re -search 3, 455-467.

Burdick, G. M., Berman, N. S., and Beaudoin,S. P., 2005. Hydrodynamic particle removalfrom surfaces. Thin Solid Films 488, 116-123.

Carper, J., 1999. The CRC Handbook of Che -mis try and Physics. Libr J 124, 192-+.

Chang, Y. I. and Chan, H. C., 2008. Corre la -tion equation for predicting filter coefficientunder unfavorable deposition conditions.Aiche J 54, 1235-1253.


62

Cooper, K., Gupta, A., and Beaudoin, S., 2000.Substrate morphology and particle adhesionin reacting systems. J Colloid Interf Sci 228,213-219.

Dabros, T., 1989. Interparticle HydrodynamicInteractions in Deposition Processes. ColloidSurface 39, 127-141.

Darbha, G. K., Schäfer, T., Heberling, F., Lüttge,A., and Fischer, C., 2010. Retention of LatexColloids on Calcite as a Function of SurfaceRoughness and Topography. Langmuir 26,4743-4752.

Das, S. K., Schechter, R. S., and Sharma, M.M., 1994. The Role of Surface-Roughness andContact Deformation on the HydrodynamicDetachment of Particles from Surfaces. JColloid Interf Sci 164, 63-77.

Delos, A., Walther, C., Schafer, T., and Buchner,S., 2008. Size dispersion and colloid mediatedradionuclide transport in a synthetic porousmedia. J Colloid Interf Sci 324, 212-215.

Dong, W. P., Sullivan, P. J., and Stout, K. J.,1992. Comprehensive Study of Parameters forCharacterizing 3-Dimensional Surface-Topo gra -phy.1. Some Inherent Properties of ParameterVa riation. Wear 159, 161-171.

Dong, W. P., Sullivan, P. J., and Stout, K. J.,1993. Comprehensive Study of Parameters forCha ra cterizing 3-Dimensional Surface-Topo -graphy. 2. Statistical Properties of Parameter Va -riation. Wear 167, 9-21.

Dong, W. P., Sullivan, P. J., and Stout, K. J.,1994a. Comprehensive Study of Parameters forCharacterizing 3-Dimensional Surface-To po -graphy. 3. Parameters for Characterizing Am - plitude and Some Functional-Properties. Wear178, 29-43.

Dong, W. P., Sullivan, P. J., and Stout, K. J.,1994b. Comprehensive Study of Parameters forCharacterizing 3-Dimensional Surface-Topo gra -phy. 4. Parameters for Characterizing Spatial

and Hybrid Properties. Wear 178, 45-60.

Drelich, J., 2006. Adhesion forces measuredbetween particles and substrates with nano-roughness. Miner Metall Proc 23, 226-232.

Elimelech, M., 1994. Effect of Particle-Size onthe Kinetics of Particle Deposition under At t -rac tive Double-Layer Interactions. J Colloid In -terf Sci 164, 190-199.

Elimelech, M. and Omelia, C. R., 1990. Kineticsof Deposition of Colloidal Particles in Porous-Media. Environ Sci Technol 24, 1528-1536.Filby, A., Plaschke, M., Geckeis, H., and Fang -hanel, T., 2008. Interaction of latex colloids withmineral surfaces and Grimsel granodiorite. Jour -nal of Contaminant Hydrology 102, 273-284.

Fischer, C., Karius, V., and Thiel, V., 2007. Or -ga nic matter in black slates shows oxidativedegradation within only a few decades. Jour -nal of Sedimentary Research 77, 355-365.

Fischer, C., Karius, V., Weidler, P. G., and Lütt -ge, A., 2008. Relationship between microme-ter to submicrometer surface roughness andtopography variations of natural iron oxidesand trace element concentrations. Langmuir24, 3250-3266.

Fischer, C. and Lüttge, A., 2007. Converged sur -face roughness parameters – A new tool toquantify rock surface morphology and reactivi-ty alteration. Am J Sci 307, 955-973.

Fischer, C., Schmidt, C., Bauer, A., Gaupp, R.,and Heide, K., 2009. Mineralogical and geo-chemical alteration of low-grade metamorphicblack slates due to oxidative weathering.Chemie der Erde - Geochemistry 69, 127-142.

Geckeis, H., Schafer, T., Hauser, W., Rabung,T., Missana, T., Degueldre, C., Mori, A., Eiken -berg, J., Fierz, T., and Alexander, W. R., 2004.Results of the colloid and radionuclide reten-tion experiment (CRR) at the Grimsel Test Site(GTS), Switzerland – impact of reaction kine-tics and speciation on radionuclide migration.


63

Radio chim Acta 92, 765-774.

Gregory, J., 2009. Optical monitoring of parti-cle aggregates. J Environ Sci-China 21, 2-7.

Heroux, J. B., Boughaba, S., Ressejac, I., and Meu -nier, M., 1996. CO2 laser-assisted removal of sub -micron particles from solid surfaces. J ApplPhys 79, 2857-2862.

Hoek, E. M. V. and Agarwal, G. K., 2006. Ex ten -ded DLVO interactions between spherical par-ticles and rough surfaces. J Colloid Interf Sci298, 50-58.

Hoek, E. M. V., Bhattacharjee, S., and Eli me -lech, M., 2003. Effect of membrane surfaceroughness on colloid-membrane DLVO interac-tions. Langmuir 19, 4836-4847.

Johnson, W. P., Li, X. Q., and Yal, G., 2007.Colloid retention in porous media: Mechanisticconfirmation of wedging and retention inzones of flow stagnation. Environ Sci Technol41, 1279-1287.

Johnson, W. P. and Tong, M. P., 2006.Observed and simulated fluid drag effects oncolloid deposition in the presence of an energybarrier in an impinging jet system. Environ SciTechnol 40, 5015-5021.

Kline, T. R., Chen, G. X., and Walker, S. L.,2008. Colloidal deposition on remotely con-trolled charged micropatterned surfaces in aparallel-plate flow chamber. Langmuir 24,9381-9385.

Kosmulski, M., 2002. The pH-dependent sur-face charging and the points of zero charge. JColloid Interf Sci 253, 77-87.

Kosmulski, M., 2004. pH-dependent surfacecharging and points of zero charge – II. Up -date. J Colloid Interf Sci 275, 214-224.

Kosmulski, M., 2006. pH-dependent surfacecharging and points of zero charge III. Update.J Colloid Interf Sci 298, 730-741.

Kosmulski, M., 2009. pH-dependent surfacecharging and points of zero charge. IV. Updateand new approach. J Colloid Interf Sci 337,439-448.

Lea, A. S., Amonette, J. E., Baer, D. R., Liang,Y., and Colton, N. G., 2001. Microscopiceffects of carbonate, manganese, and stron-tium ions on calcite dissolution. GeochimCosmochim Ac 65, 369-379.

Luttge, A., Bolton, E. W., and Lasaga, A. C.,1999. An interferometric study of the dissolu-tion kinetics of anorthite: The role of reactivesurface area. Am J Sci 299, 652-678.

Martines, E., Csaderova, L., Morgan, H., Curtis,A. S. G., and Riehle, M. O., 2008. DLVO interac -tion energy between a sphere and a nano-pat-terned plate. Colloid Surface A 318, 45-52.

Matijevic, E. and Scheiner, P., 1978. Ferric Hy -drous Oxide Sols .3. Preparation of UniformPar ticles by Hydrolysis of Fe(Iii)-Chloride, Fe(Iii)-Nitrate, and Fe(Iii)-Perchlorate Solutions. JColloid Interf Sci 63, 509-524.

Mitchell, D. F., Clark, K. B., Bardwell, J. A.,Lennard, W. N., Massoumi, G. R., and Mitchell,I. V., 1994. Film Thickness Measurements ofSio2 by Xps. Surf Interface Anal 21, 44-50.

Mori, A., Alexander, W. R., Geckeis, H., Hauser,W., Schafer, T., Eikenberg, J., Fierz, T., De guel -dre, C., and Missana, T., 2003. The colloid andradionuclide retardation experiment at theGrimsel Test Site: influence of bentonite collo-ids on radionuclide migration in a fracturedrock. Colloid Surface A 217, 33-47.

Prescott, S. W., Fellows, C. M., Considine, R. F.,Drummond, C. J., and Gilbert, R. G., 2002. Theinteractions of amphiphilic latexes with surfa-ces: the effect of surface modifications andionic strength. Polymer 43, 3191-3198.

Raming, T. P., Winnubst, A. J. A., van Kats, C.M., and Philipse, A. P., 2002. The synthesis andmagnetic properties of nanosized hematite


64

(alpha-Fe2O3) particles. J Colloid Interf Sci249, 346-350.

Schafer, T., Geckeis, H., Bouby, M., and Fang -hanel, T., 2004. U, Th, Eu and colloid mobilityin a granite fracture under near-natural flowconditions. Radiochim Acta 92, 731-737.

Schafer, T., Seher, H., Hauser, W., Walther, C.,Degueldre, C., Yamada, M., Suzuki, M., Mis sa -na, T., Alonso, U., Trick, T., and Blechschmidt,I., 2009. The Colloid Formation and Migration(CFM) project at the Grimsel Test Site (Swit -zerland): Results from the homologue tests.Geochim Cosmochim Ac 73, A1168-A1168.

Schaldach, C. M., Bourcier, W. L., Shaw, H. F.,Vi ani, B. E., and Wilson, W. D., 2006. Theinfluence of ionic strength on the interaction ofviruses with charged surfaces under environ-mental conditions. J Colloid Interf Sci 294, 1-10.

Schmidt, M., Heberling, F., Finck, N., Stumpf,T., Schlegel, M., Dardenne, K., and Bosbach,D., 2009. Institute for Nuclear Waste DisposalAnnual Report 2008. In: Geckeis, P. D. H. (Ed.).Forschungszentrum Karlsruhe GmbH, Karls -ruhe, Karlsruhe.

Somasundaran, P. and Agar, G. E., 1967. ZeroPoint of Charge of Calcite. J Colloid Interf Sci24, 433-+.

Tong, M. P. and Johnson, W. P., 2006. Excesscolloid retention in porous media as a functionof colloid size, fluid velocity, and grain angula-rity. Environ Sci Technol 40, 7725-7731.

Torkzaban, S., Bradford, S. A., and Walker, S. L.,2007. Resolving the coupled effects of hy -drodynamics and DLVO forces on colloid att -achment in porous media. Langmuir 23, 9652-9660.

Ullrich, B., Hantsch, K. U., Müller, K., Seidel,H., and Ullrich, B., 2005. Sekundär mineral bil -dung en des Alaunschieferbergwerkes »Moras -sina« bei Beiträge zur Geologie von ThüringenN.F. 12, 41-69.

van Loenhout, M. T. J., Kooij, E. S., Worm ees -ter, H., and Poelsema, B., 2009. Hydrodynamicflow induced anisotropy in colloid adsorption.Colloid Surface A 342, 46-52.

Wu, Z. J., Xiang, H., Kim, T., Chun, M. S., andLee, K., 2006. Surface properties of submicro-meter silica spheres modified with aminopro-pyltriethoxysilane and phenyltriethoxysilane. JColloid Interf Sci 304, 119-124.Zheng, H. P., Berg, M. C., Rubner, M. F., andHam mond, P. T., 2004. Controlling cell attach-ment selectively onto biological polymer-collo-id templates using polymer-on-polymer stam-ping. Langmuir 20, 7215-7222.

Zhuang, J., Qi, J., and Jin, Y., 2005. Retentionand transport of amphiphilic colloids underunsaturated flow conditions: Effect of particlesize and surface property. Environ Sci Technol39, 7853-7859.


65

Surface and Wetting Properties of DiageneticMinerals and Sedimentary Grains in ReservoirRocks (NanoPorO)

IntroductionCurrent secondary oil recovery measures allowfor the recovery of a maximum of about 33 %of the oil in a reservoir. The remaining almosttwo third of the energy carrier are lost due todecreasing pressure, pore blocking, water inva-sion or hydrocarbon fluid adhesion to the rock.Therefore, fluids weakening the adhesion ofhydrocarbons to pore walls and increasing per-meability of the rock by e.g. mineral cementdissolution, are injected to the deposits inorder to augment production. All such measu-res however, including enhanced oil recovery(EOR) methods such as the injection of super-critical CO2 may increase the recovery factor ofthe original oil in place by further 10 % only.Consequently, most of the hydrocarbon wealthin oil and gas reservoirs cannot be extractedand is lost to future generations. It is thus ofhigh importance to understand the fundamen-tal wetting processes in the pore space, to bet-ter develop the potential of oil and gas pro-duction from reservoir rocks and to secure thefossil energy supply.

In siliciclastic oil and gas reservoir rocks, thepore space typically faces mineralogically vary-ing sedimentary grains and various diageneticminerals. Most common are mineral surfaces

of quartz, feldspar, phyllosilicates, carbonatesand iron oxides and hydroxides. This mineralo-gy surrounding the pore space, and the surfa-ce chemistry, topography and roughness onthe micro and nano-scale rule the wettingbehavior and adhesion properties of hydrocar-bon fluids, water, or CO2 to the pore walls. Thedispersion, migration, adhesion and reactivity offluids in rocks depends also on pressure andtemperature conditions, nevertheless, particu-larly the morphology of pore walls and the poreand pore-throat shapes, have a significant im -pact on the behavior of the water-gas contactdepth (WGC) and on the potential recovery ofhydrocarbons from the given reservoir rock.

Each episode of fluid transport through therock leaves a significant and characteristictrace in cement mineralogy, pore morphology,permeability, but often also in sediment grainor bioclast alteration. Such processes can bedefined down to the nanoscale and play a cru-cial role in further mobility of hydrocarbons inrocks (Hassenkam et al., 2009). Sediment wet-ting varies with the alteration of the surfacesand depends on surface roughness, surfacecharges, and the chemical composition of theliquid phase (Al-Futaitsi et al., 2003; Al-Futaisiand Patzek, 2004). The interfacial tension of

Altermann W. (1)*, Drobek T. (1), Frei M. (1), Heckl W.M. (2), Kantioler M. (1), Phuong K.L. (1), Stark R.W. (3),

Strobel J. (4)

(1) LMU, Geology, Sedimentology, Crystallography, [email protected], [email protected]

muenchen.de, [email protected], [email protected], [email protected]

(2) TU Munich, TUM School of Education & Deutsches Museum München, [email protected]

(3) TU Darmstadt, Center of Smart Interfaces, [email protected]

(4) RWE Dea, Hamburg, [email protected]



66

the fluids strongly depends on the compositionof the coexisting phases (Sutjiadi-Sia et al.,2008). In addition, the presence of a supercri-tical (sc)CO2 phase can affect the wetting pro-perties of the other phases due to massexchange as a function of pressure (Foullac etal., 200; Sutjiadi-Sia et al., 2008; Jäger andPietsch, 2009). It is, however, not clear howthe specific conditions of the reservoir (p, T,surface chemistry and morphology) affect theinterfacial tension of the relevant fluids. Tounderstand the fundamental physico-chemi-stry helps to design new technologies for terti-ary exploitation measures and for CO2-storage(Carbon Capture and Storage: CCS). In thisproject we investigate the relationship betweenvarious minerals or grains facing the pore space,to fluids in reservoir rocks and to CO2 and scCO2

(Altermann et al., 2008). In the following wereport on the geological and physical character-ization of the reservoir rock under investigationand on the characterization of wetting of roughmodel surfaces in scCO2.

Characterisation of the rocks and thepore spaceThe RWE Dea AG, Hamburg, provided coresand plugs of sandstone reservoir rocks with

varying physical properties such as permeabili-ty, porosity, grain sizes and mineralogy fromnorth German oil and gas deposits. The sam-ples are well indurated, but still slightly friable,fine to medium grained, with sparry cementsand with abundant pale and dull grains (seee.g. Fig. 1). From each sample several petro-graphic sections were obtained. The porositywas stained blue but no other stains (e.g. forfeldspars or carbonates) were applied. In thinsection, the samples may show some indistinctalternating lamination of coarser and finergrains and partly with differing cement andopen pore space volume. Some samples howe-ver, exhibit very distinct lamination defined bygrain size, mineralogy and porosity changes.Most sandstone samples are structurallymedium mature but compositionally heteroge-neous and immature. The grains are very wellto well rounded, sometimes sub-roundedwhen finer, but appear also facetted, subroun-ded to poorly rounded because of dissolutionunder pressure, along grain to grain contacts(sutured contacts) and because of grain disso-lution along boundaries during diagenesis(comp. below). The sorting of grain sizes andelongation are moderate to poor. Most grainsare rather subspheroidal although distinctly

Figure 1: Grain under single (left side) anddouble (right side) polarised light. Scale in theupper right corners is 100µm. Chlorite is fil-ling the pore space overgrowing he matiteskins and the grains as pointed out by thearrows in picture on the left. The open porespace in the upper right corner of the upperpicture pair displays haematite followed bychlorite growing tangentially into the openpore (enlargement in the lower pictures). Thelarge elongated quartz grain in the upper partof the upper picture is syntaxially overgrownby quartz ce ment. The original grain boundaryis pointed out by the arrow. In the upper rightpicture, calcite cement (right arrow) is alsovisible. The left arrow points to an alteredfeldspar with sericite inclusions. Sericite ma trixcan be also found between some grain boun-daries. In the lower, enlarged view chloritegrowing tangentially (arrows) is overgrowinghaematite skins (arrows). The open pore spacedisplays small fans of chlorite facing the openpore. Fluid inclusion trails can be seen in thequartz grains (pointed out by arrows in thelower right picture. Sutured grain contact canbe recognised between the two grains on lefthand side of the lower pictures.


67

elongated grains occur as well. The elongationseems not to depend on the grain composi-tion. The sorting varies from laminae to laminaebut is always not very high and smaller grainsalways are compacted between larger grains,with rare coarse sand sizes and common fineand medium grain sizes present. Rounding ofgrains strongly depends on the size, fine sandbeing significantly less rounded and less sphero-idal than medium and coarse sand grains. Allthese properties are somewhat obliterated bywidespread syntaxial quartz and albite over-growth, which appears unambiguous onlywhen observed under crossed polarisator.

Grain composition is of about 60 % to 70 %quartz grains. Albite, plagioclase and K-feld-spar grains are also present and common.Quartzite, recrystallised chert, vein quartz andigneous rock fragments constitute less than 5 % of the grain content, the composition ofthe sandstones hints towards a cratonic sourcewith mixed sedimentary and igneous rocks.

This is supported by the presence of detritalphyllosilicates (muscovite and altered, greenishbiotite) compacted between the quartz, feld-spar and rock fragment grains. Few rounded,detrital, opaque ore grains (heavy minerals)were also found. The matrix varies between siltsize detrital grains, clay minerals (primary andas alteration products of some unknown, pro-bably volcanic and igneous rock fragment pre-cursor), and patches and interstices with hae-matitic filling (possible alteration).

All samples exhibit a complicated polyphasediagenetic history. Sutured contacts and graindissolution features evidence significant fluidmovement during diagenesis. The diageneticminerals were precipitated from migrating bri-ans in the pore space, as synthaxial and normalquartz overgrowth cement, but also very com-monly as albite cements, confirmed by EDAXanalyses (see Fig. 2 and 3). Very often, the firstgeneration of cement/matrix is that of chloriteand clay minerals and /or haematite skins sur-

Figure 2: The left upperSEM micrograph displaysan overview of the grainsurface with idiomorphicminerals growing on thegrain into the open porespace. The right handimage shows a close upview of this overgrowth,displaying idiomorphicmineral habitus of chlorite,quartz, albite (not welldeveloped) and halite skin(Salz) and the respectiveEDX measurements. In thehalite diagram (upper left)feldspar content is visiblebelow the halite encrusta-tion and thus Al and Kpeaks are visible as well.


68

rounding the grains. In a later diagenesis quartzand albite coatings were generated. Clay mine-rals, chlorite, haematite coats are present as thesecond and third cement generation as well.Partial decementation – dissolution of these neo -morphic minerals, is evident in laminae, wheresuch cements are much thinner and very irregu-lar, leaving out abundant open pore space andby partial filling of the interstices with calcite, asmost probably the latest stage of cementation.This is particularly clear where calcite is theonly remaining pore filling cement and indirect contact to the detrital grains, but pas-sing (overlapping) laterally onto quartz andalbite cements. In such cases, it becomes appa-rent that the process was not a gradual repla-cement of former cements by calcite, butinstead, the early cements were partly dissol-ved and transported aside in solution, and onlysubsequently, calcite was precipitated frommigrating fluids. Calcite together with othercements thus, often fills resorption embay-ments but also dissolution pits in the grains.

SEM investigations were performed on singlegrains broken off from the core samples, pik-ked with pincers, mounted on a conventionalSEM mount carrier disk with double-sided stik-king tape, and coated with sputtered carbon.Elemental mapping was conducted by theEDX. In most cases cements fully cover thegrains by multiple coats of various minerals.The outer most layers very often exhibit idio-morphic crystals facing the open pore space.The involved minerals are clay mineral plate-lets, albite twins, chlorite fans, idiomorphic,pseudohexagonal quartz needles and abun-dant salt (halite - NaCl) as hoppers and skins.The faces of these idiomorphic crystals areabsolutely flat and smooth on sub micrometerscale. Where the coatings build a tangentialskin without idiomorphic faces, the roughnessis just below 1µm scale and can be well obser-ved by the SEM. Typical crystalline surfaceswith pitted regular crystallographic pattern arerecognizable. As could be expected from theevaluation of the cements under optical micro-

Figure 3: The left picturedisplays an overview ofgrain and cements withrespective spots of EDXanalysis, as shown in dia-grams below. Chlorite,sericite, calcite, halite(Salz, NaCl) and quartzwere found. The righthand picture displays anEDX elemental map of theview in left picture. Thelarge area of calcitecement appears on theright side of the picture.Most of the other area isoccupied by quartzcement.


69

scope, only calcite does not produces idiomor-phic crystals. It has been observed as brokensurfaces only, because it is the latest stage ofcementation usually filling the pore space asone crystal, poikilotopic infill of several neigh-boring pores.

AFM investigations were performed on singlegrains or small, cemented and bond graingroups of 2-10 grains, mounted on a smallmetal plate sample carrier and fixed by a com-mercial nail varnish. The video camera mountedon the AFM table allows for a high magnifica-tion observation of the grains (500x). It was dif-ficult, but nevertheless possible, to relocate thesame spots under the AFM and SEM. The mea-sured surface roughness of the sample dependson the scale of observation. At all scales howe-ver, the roughness is significant and the topo-graphy as shown in the profiles in both direc-tions, vertically and horizontally, is always steepand distinct. This provides problems as to thechoice of the right AFM tip. The measurementprecision achieved in vertical direction was of up

to few tens of nm on ideal surfaces; surfaces ofsteep and high roughness however, were notpossible to measure exactly.

Further roughness and topography measure-ments were performed by a digital laser scan-ning microscope (laser profilometer, LSMKeyence). The LSM delivers digitally sandwi-ched surface images of the sample, with highdepth of focus and a quantitative topographydata set in µm resolution (Fig. 4). The advanta-ge of such measurements is the ability of theLSM to exactly resolve even very steep topo-graphy, although the resolution is by an orderof magnitude lower than that of an AFM or alight interferometer. Fig. 5 shows an examplewhere the grain surface is covered by cementand hematite. Additionally, needles (illite) growinto the pore space. Such needles can affectthe fluid transport in the porous network andretain hydrocarbons. This example also showsthat a microscopic inspection of the pore spaceis essential to identify relevant morphologicalfeatures and that the complicated surface

Figure 4: The LSM and optical photographs show cements of different shape within a pore of the investigated sandsto-ne. On the left hand side albite blades are visible. In the center of the pore needles of illite grow into the pore space onquartz overgrowth of a clastic grain. Some of the needles seem also to grow from the albite surface.


70

topography can not be simply described ormeasured by an AFM on a larger (>10 µm) scale.

As it is often difficult to identify very smallgrains or cements by optical microscopy, andalmost impossible to identify the mineralogyby the AFM or LSM, therefore the above inve-stigations were supported by micro-Ramanspectroscopy in order to aid the mineral identi-fication, performed mostly on thin sectionsand on single grain mounts. Our micro–Ramandevice is attached to a confocal microscope,which allows for three-dimensional imagingand for simultaneous chemical analysis. Theconfocal configuration leads to a small samplevolume, from which Raman scattering can berecorded. Lateral and vertical cross-sections ofthe sample can easily be generated with highspatial resolution, whereby lateral and vertical1-µm-thick optical slices are acquired from thesurface to the maximum depth of 25 µm. Thetechnical set up [confocal Raman microscope,alpha300 R; WITec GmbH, Ulm, Germany,frequency doubled Nd:YAG laser (532 nm, Pmax = 22.5 mW) and a 100× objective (Nikon,NA=0.90, 0.26 mm working distance)] is des-cribed in detail in Kremer et al. (in press).

Due to the above described properties, thepore space in the investigated samples is veryheterogeneous and has a wide range of mine-ralogical and roughness characteristics. The

observed variety of the open pore morphologydepends strictly on the shape of the detritalgrains, on the vesicular cement minerals oftangential and needle-like habitus and on thefringing dissolution and pressure dissolution ofthe grains, that are mostly densely packed andgrain supported. It is thus chiefly the diagene-tic history and not the sedimentary history inthese samples that rules the gross and nano-morphology of the pores. In order to simplifythe search for a suitable target for experi-ments, the pores were classified according tothe key mineralogy, into four main classes,each comprising several subclasses:

1) Not overgrown, rare grain surface/porespace contacts of varying mineralogy, butmainly quartz, feldspar and common volca-niclasts and heterogeneous rock fragmentsurfaces.

2) Quartz cement/pore space contact of botry-oidal or hypidiomorphic quartz.

3) Calcite cement/pore space contact of vary-ing shapes but mostly blocky and needleshaped or irregular, corroded calcite.

4) Heterogenous haematite, illite, chlorite albi-te cement/pore space contacts of differingmorphologies.

Small Angle Neutron ScatteringA combination of small angle neutron scatte-

Figure 5: (left) Boundary of grain and part of grain surface (G) The grain covered by hematite skin has been stressed atthe concave areas (dissolution depressions) and cement has overgrown the surface subsequently, while at the poreboundary illite needles (circle) are growing into the open pore space (Po). The sub-euhedral minerals with platy habitusare albite (dotted arrow in the right picture).


71

ring (SANS) and ultra-small angle neutron scat-tering (USANS) experiments on dry and partial-ly wetted reservoir rock samples has been car-ried out at the Institut Laue-Langevin (ILL), inGrenoble, France in September 2010. Themea surements will give insight into the inter-play between texture, pore surface rough-ness, size distribution of the cement over-growth and the distribution of connate waterin the pore network.

Molecular scale interactions between fluidsand mineral surfaces have a crucial impact onpore volume estimation, on the understandingof hydrocarbon migration, and on the possibledegree of exploitation. These interactions alsoinfluence the assessment of CO2 storage capa-city and retention mechanisms or of the poten-tial of geothermal energy generation in aqui-fers (Pruess and Azaroual, 2006; Pruess, 2008).The study aims to reveal nanoscale processes inthe porespace of sedimentary rocks in orderto explain the migration processes and the per-meability for fluids, gas and supercritical CO2

within sedimentary rocks.

For reservoir rocks it is important to study theintact porous network within a representative-ly large volume of a bulk sample. Sandstones,mainly composed of quartz grains, show diver-se heterogeneous textures, accessory minerals,and varying grain sizes of 5 ~ 100 �m. Notably,

capillary bound water in the pore system appe-ars to reduce the permeability. The penetrationpower of neutrons allows to investigate theporosity of such bulk samples with SANS. Thismethod is more powerful in combination withultra small angle neutron scattering (USANS)for the study of the microstructure of rocksover a large range of scales (Radlinski, 2006;Triolo et al., 2000; Sen et al., 2002). The frac-tal dimension derived from the SANS/USANSdata can be correlated to microscopy data(Wong and Howard, 1986; Wong and Bray,1988; Radlinski et al., 2004; Anovitz et al.,2009) and used as a parameter for the correla-tion of porosity and permeability of rocks inconjunction with their pore topology (Pape etal., 1999; Pape and Clauser, 2000). At the beamlines D11 and S18 at ILL SANS and USANS ex -periments were carried out on the same setof samples. Pore filling and capillary conden-sation of fluids in porous media was observedwith SANS and USANS by using contrast-mat-ching H2O/D2O mixtures (Broseta et al., 2001;Erko et al., 2010).

In Figure 6 (left) data measured on sandstonereservoir rock samples of different depth anddifferent porosity are shown. The scatteringintensities follow a power law behavior, whichis typical for porous rocks with an exponent ofabout -2.2, which corresponds to a mass frac-tal topography of the pore space with a

Figure 6: (left) Ultra small angle neutron scattering data of different plug samples from a quartz sandstone reservoirrock of a north German gas deposit. The intensities show a power law behaviour. With an exponent of about -2.2, indicating a mass fractal topography. (right) Filling the pores with D2O or a contrast-matching mixture of D2O and H2Oreduces the scattering intensity.


72

dimension of 2.2. Filling the pores with D2O ora mixture of D2O and H2O, contrast-matchingto quartz and feldspar, massively reduced thescattering intensity. The remaining scatteringcontrast is mainly caused by the surfaces ofcalcite cement, heavy minerals and otherminor constituents of the rock.

Wettability of rough surfaces at reservoirconditionsThe wettability of the reservoir rock has a largeinfluence on the productivity of a hydrocarbonreservoir. In oil fields, depending on the type ofwetting characteristics of the rocks, differenttechnologies for developing a field are neces-sary. One of the most prominent effects of themicro- and nano-morphology on wetting is thesuper-hydrophobicity, better known as the»lotus effect« on glass or ceramics (Barthlottand Neinhuis, 1997; Neinhuis and Barthlott,1997; Spori et al., 2008). The lotus effect iscaused by an only partial wetting of the roughsurface profile by the liquid phase in the so-cal-led Cassie-Baxter regime (Wenzel, 1936;Cassie and Baxter, 1944). So far, however, it isnot clear how the presence of a supercriticalphase and mass transport between the wet-ting and the supercritical phase affect super-hydrophobicity.

To investigate the fundamental processes ofwetting under reservoir conditions we have

applied a high pressure / high temperature con -tact angle goniometer (Fig. 7). The instrumenthas a cylindrical high pressure view cell (HPVC),which can be filled with different fluid or gase-ous media. It is possible to study wettability ofsurfaces and surface tension of different mediawith the sessile drop and pendant dropmethod. Experiments can be carried out at rea-listic reservoir conditions of up to 69 MPa and180 °C with different fluids and CO2.

To study the effect of temperature and pressu-re on the wettability of rough surfaces, amodel surface was created that resembles thesurface of a lotus leaf (Fig. 8). We used a cera-mic aluminum oxide replica of a two-step-photolithographic pillar structure. The samplewas coated with a chemically stable, hydropho-bic fluorosilane layer. Dynamic contact angle(CA) measurements of water drops in fluid aswell as supercritical CO2 were performed at dif-ferent temperature and pressure conditions. Thepillar surface shows a superhydrophobic beha-vior with advancing CA of 150-160°, with aslight increase of the CA with pressure (0.1MPa-1) and a hysteresis of about 15° (Fig. 9).The contact angles at 100 °C show larger slope(0.4 MPa-1) and increased hysteresis (25°) thanthe data measured at room temperature. Thetemperature and pressure dependence of thecontact angle can be influenced by the interfa-cial tension of the phases as well as buoyancyeffects due to variations of the density.

Figure 7: High Pressure Viewcell (Krüss GmbH, Ham burg,Eurotechnica, Bargteheide).The cylindrical cell has twooptical windows. A dropcan be placed on the sam-ple surface through a capil-lary. The cell can be filledwith two different gaseousor fluid media. Liquid CO2can be pressurized up to 69MPa and heated to 180 °C.The cell is equipped with acontact angle measurementsystem.


73

The experiment shows that wetting phenome-na caused by surface morphology can alsooccur under reservoir conditions. Main impacton the details of the wetting (contact angle,hysteresis) can be expected from variations inmass density or surface tension due to masstransfer with the supercritical phase.

ConclusionsThe pore space morphology of reservoir rocksdepends on the size and shape of the clasticdebris and can be drastically changed duringdiagenesis due to influence of factors such ascompaction, cementation, and pressure disso-lution. The above observations have led to the

conclusions that the pore surfaces at the firstphase are rather smooth on a µm scale, andthe pores are well-connected facilitatingmigration of fluids. The subsequent stages ofdiagenesis, including fluid migration itself,have changed and complicated however, thepore spaces morphology. Pores were filled byminerals and new micro-pores were opened bydissolution of diagenetic minerals. In addition,the grains and cements were partially dissol-ved, forming new pores within some sedimen-tary layers or laminae. The overgrowth on cla-stic grains by secondary minerals producedrough surfaces, leading to more rugged poresurfaces. The flow capacity of hydrocarbon flu-ids has been decreased due to authigenic

Figure 8: SEM image of a cera-mic replica of a pillar structure.The large pillars have a diameterof 100 µm, the small ones 10µm. The sub-micron scale topo-graphy of the sample is domina-ted by the grain size of the alu-minium oxide particles of theslurry used for casting. Due tothe undercut shape of the mas -ter, part of the small pillarsbroke off at separation of thereplica from the negative (D.Spori, ETH Zurich).

Figure 9: (left) Microphotograph of a water drop on the superhydrophobic pillar structure in liquid CO2. The surfacewas moved with respect to the capillary, to measure the advancing and receding contact angle. (right) Contact anglesof water in CO2 at different temperatures and pressures. The data at 100 °C (supercritical CO2) show a higher slope(0.4°/MPa) and a stronger hysteresis (25°) than the data at room temperature (0.1°/MPa and 15°).


74

minerals forming barriers and blocking the oiland gas flow. Micro-pore structures and over-grown grain surfaces have formed larger sur-faces increasing the adhesion capacity of flu-ids. Small angle neutron scattering data indica-te a mass fractal geometry of the pore spacewhich spans over several orders of magnitude.In the next step the correlation between scat-tering data (bulk) and the microscopically mea-sured roughness (surface) will be investigated.This correlation provides input for the assess-ment of the inner roughness of the rock.

The wetting experiments show that the samewetting phenomena (super-hydrophobicity) asknown from atmospheric conditions also occurunder reservoir conditions. Static and dynamicwetting properties, however, depend on themass transfer on the phases and thus on thetemperature and pressure. Further wettingexperiments on model surfaces and on typicalminerals like quartz, micas, calcite or feldsparsshall provide further insight into the dynamicwetting behavior.

Our research and experiments aim to answerquestions imposed by problems occurring inpractical oil and gas exploitation activities. It isdirected by theoretical petro-physical conside-rations and practical cases, as discussed withthe RWE-Dea, Hamburg, the main oil and gasproducer in Germany.

AcknowledgementsWe thank the beamline scientists H. Lemmel(TU Wien, Atominstitut) and P. Lindner (ILLGrenoble) for their support in carrying out theneutron scattering experiments. We thank D.Spori and C. Cremmel (ETH Zürich, Laboratoryfor Surface Science and Technology) for theceramic replicas.

ReferencesAl-Futaisi, A., Qaboos U, S., Patzek, T.W. andU.C. Berkeley, 2003. Three-phase hydrauliccon ductances in angular capillaries. SPE J. 8(3), 252-261.

Al-Futaisi, A., and T.W. Patzek, 2004. Se con -dary imbibition in NAPL-invaded mixed-wetsediments. J. Contam. Hydrol. 74: 61-81.

Altermann, W., Heckl, W.M., Stark, R.W.,Strobel, J. and Ch. Wolkersdorfer, 2008. Nano-structure and wetting properties of sedimenta-ry grains and pore-space surface (NanoPorO).Geotechnologien Science Report No. 12. Ko or -di nierungsbüro Geotechnologien, Pots dam,ISSN 1619-7399, 58-69.

Anovitz, L.M., Lynn, G.W., Cole, D.R., Rother,G., Allard, L..F., Hamilton, W.A., Porcar L. andM.H. Kim, 2009. A new approach to quantifi-cation of metamorphism using ultra small andsmall angle neutron scattering. GeochimicaCosmochimica Acta, 73 (24), 7303-7324.

Barthlott, W. and C. Neinhuis, 1997. Purity ofthe sacred lotus, or escape from contaminationin biological surfaces. Planta. 202, 1-8.

Broseta, D., Barre, L., Vizika, O., Shahidzadeh,N., Guilbaud, J. P. and S. Lyonnard, 2001. Ca -pil lary condensation in a fractal porous me -dium. Phys. Rev. Lett., 86 (23), 5313-5316.

Cassie, A.B.D. and S. Baxter, 1944. Wettabilityof porous surfaces. Transactions of the FaradaySociety. 40: 0546-0550.

Erko, M., Wallacher, D., Brandt, A., and O. Paris,2010. In-situ small-angle neutron scattering studyof pore filling and pore emptying in orderedmesoporous silica. J. Appl. Crystal lo gr., 43, 1-7.

Fouillac, C., Sanjuan B., Gentier, S., and I.Czernichowski-Lauriol, 2004. Could sequestra-tion of CO2 be combined with the develop-ment of enhanced geothermal systems? In:Third Annual Conference on Carbon Captureand Sequestration, Alexandria.

Hassenkam, T., Skovbjerg, L.L., and S.L.S.Stipp, 2009. Probing the intrinsically oil-wetsurfaces of pores in north sea chalk at subporeresolution. Proceedings of the NationalAcademy of Sciences (PNAS). 106: 6071-6076.


75

Jäger, P. and A. Pietsch. 2009. Characterizationof reservoir systems at elevated pressure. J.Petrol. Eng. 64, 20-24.

Kremer, B., Bauer, M., Stark, R.W., Gast, N.,Altermann, W., Gursky, H-J., Heckl, W.M. andJ. Kazmierczak (in press). Raman and atomicforce microscopy test for chlorococcalean affi-nities of problematic Silurian microfossils(»acritarchs«). Journal of Spectroscopy.

Neinhuis, C. and W. Barthlott, 1997. Cha rac -terization and distribution of water-repellent,self-cleaning plant surfaces. Annals of Botany.79: 667-677.

Pape, H., Clauser, C. and J. Iffland, 1999. Per -me ability prediction based on fractal pore-spacegeometry. Geophysics, 64 (5), 1447-1460.

Pape, H. and C. Clauser, 2000. Variation of per -meability with porosity in sandstone diagenesisinterpreted with a fractal pore space model.Pure Appl. Geophys., 157 (4), 603-619.

Pruess, K., 2008. On production behavior of en -hanced geothermal systems with CO2 as wor-king fluid. Energy Conversion and Manage -ment. 49, 1446-1454.

Pruess, K. and M. Azaroual, 2006. On the fea-sibility of using supercritical CO2 as heat trans-mission fluid in an engineered hot dry rockgeothermal system. In: 31st Workshop onGeothermal Reservoir Engineering, Stanford,CA. SGP-TR-179.

Radlinski, A.P., Ioannidis, M.A., Hinde, A.L.Hainbuchner, M. Baron, M., Rauch, H. and S.R.Kline, 2004. Angstrom-to-millimeter character-ization of sedimentary rock microstructure. J.Colloid Interface Sci., 274 (2), 607-612.

Radlinski, A.P., 2006. Small-angle neutron scat-tering and the microstructure of rocks. Reviewsin Mineralogy and Geochemistry, 63, 363-397.

Sen, D. Mazumder, S. and S. Tarafdar, 2002.Pore morphology and pore surface roughning

in rocks: A small-angle neutron scattering inve-stigation. J. Mater. Sci., 37 (5), 941-947.

Spori, D.M., T. Drobek, S. Zurcher, M. Ochsner,C. Sprecher, A. Muehlebach, and N.D. Spencer,2008. Beyond the lotus effect: Roughness,influences on wetting over a wide surface-energy range. Langmuir. 24, 5411-5417.

Sutjiadi-Sia, Y., P. Jaeger and R. Eggers. 2008.Interfacial phenomena of aqueous systems indense carbon dioxide. J. Supercrit. Fluid., 46,272-279.

Triolo, F. Triolo, A. Agamalian, M.M. Lin, J.-S.Heenan, R.K. Lucido, G. and R. Triolo, 2000.Fractal approach in petrology: Combining ultrasmall angle, small angle and intermediateangle neutron scattering. J. Appl. Crystallogr.33 (1), 863-866.

Wenzel, R.N., 1936. Resistance of solid surfa-ces to wetting by water. Ind. Eng. Chem. 28,988-994.

Wong, P.Z. and A.J. Bray, 1988, Porod scatte-ring from fractal surfaces. Phys. Rev. Lett., 60(13), 1344-1344.

Wong, P.Z. and J. Howard, 1986. Surface roug-hening and the fractal nature of rocks. Phys.Rev. Lett., 57 (5), 637-640.


76

Reactivity of Calcite/Water-Interfaces (RECAWA):Molecular level process understanding for tech-nical applications

AbstractRECAWA intends to develop a fundamentalunderstanding of the reactivity and dynamicsof calcite surfaces during crystal growth inaquatic systems. Therefore, the calcite-watersurface has been characterised at differenthydrochemical conditions and surface diffrac-tion measurements have been applied toprobe if changes in zeta potential are accom-panied by changes in the calcite-water interfa-ce structure. The interaction of calcite withSelenium is investigated in a mixed flow reac-tor (MFR) study which is accompanied by GI-EXAFS analyses to characterize the Se adsorp-tion species.The interaction of calcite surfaces with phos-phate and phosphonates is the focus of twosub-projects. The sorption/precipitation me cha - nisms of PO4

3- and HPO32+, respectively, on dif-

ferent calcite powders (precipitated calciumcarbonates and limestone powders) are investi-

gated to determine sorption isotherms andphase transformations on the applied surfa-ces which is of great interest with regard tophosphate recycling and/or water treatment.Ana lytical work includes SEM, XRD and XAFSspectroscopy for a detailed characterisationof the calcite powders in different sections ofthe isotherms.Experimental investigations on the interactionof limestone powder and two superplasticizershave been conducted with regard to concreting.These experiments were done at alkaline condi-tions and lead to massive reactions and changesin the zeta potential of the solutions and limes-tone powder when superplasticizers added.Additional results of quantum-mechanical cal-culations and force-field modelling demonstra-ted that the Double Defect Method is able toquantitatively predict mixing properties ofvarious iso-structural binary and ternary carbo-nate solid solutions. In particular, low equili-

Stelling J. (1), Neumann T.* (1), Kramar U. (1), Schäfer T. (2), Heberling F. (2), Winkler B. (3), Vinograd V. (3),

Arbeck D. (3), Müller H.S. (4), Haist M. (4), Glowacky J. (4), Vucak M. (5), Fischer U. (6), Pust C. (6), Huber J. (7),

Bosbach D. (8)

(1) Karlsruhe Institute of Technology (KIT), Institute of Mineralogy and Geochemistry (IMG), e-mail: [email protected],

[email protected], [email protected]

(2) Karlsruhe Institute of Technology (KIT), Institute for Nuclear Waste Disposal (INE), e-mail:


(3) Goethe Universität Frankfurt (UF), Institut für Geowissenschaften (IfG), e-mail: [email protected],


(4) Karlsruhe Institute of Technology (KIT), Institute of Reinforced Concrete Structures and Building Materials (IfMB), e-

mail: [email protected], [email protected], [email protected]

(5) Schaefer Kalk GmbH & Co. KG, e-mail: [email protected]

(6) Rheinkalk Akdolit GmbH & Co. KG, e-mail: [email protected], [email protected]

(7) Lafarge Zement Wössingen GmbH, e-mail: [email protected]

(8) Forschungszentrum Jülich GmbH, Institute of Energy Research (IEF), e-mail: [email protected]



77

brium retention levels of SO42- and SeO4

2- incarbonates imply that the reasonably largeconcentrations SO4

2- and SeO42- in carbonates

reported in previous studies should be attribu-ted to non-equilibrium entrapment effects.Data obtained from atomistic modelling com-plement our data determined using labmethods. Further experimental and analytical,work including in-situ AFM studies usingIceland spar single crystals, is in progress andwill help to interconnect the present results ofthe sub-projects.

1. IntroductionCarbonates are of high economic interest andrepresent an important resource for the chemi-cal and pharmaceutical industry, glass andpaper industry, construction material industry,for the production of fertilizers and for the qua-lity of drinking water. Due to the abundance ofcarbonate rocks in the earth crust, carbonatesare considered as a mineral mass product.As the thermodynamically most stable andmost abundant carbonate phase, calcite playsa prominent role for numerous natural proces-ses in our environment. Calcite affects in va ri -ous ways the global circulation of matter, regu-lates the pH and controls the chemical compo-sition of natural aquatic systems, and has theability to fix or structurally incorporate biologi-cally active and, depending on the concentra-tion, hazardous elements such as phosphorus,calcium, magnesium, iron, arsenic, seleniumand other trace elements. Adsorption reactionsare usually rather fast compared to co-precipi-tation reactions. In particular with respect tokinetic aspects of trace element sorption, themolecular level reactions at the mineral/waterinterface need to be understood in variousfields of application to improve the effective-ness of industrial calcite mass products such asPCCs or filter material for water treatment.Atomistic models allow studying sorption ofcomplexes and molecules at the mineral-waterinterface and the formation of solid solutions.These aspects of the interaction of calcite sur-faces with dissolved reactants require differentsimulation approaches. Sorption processes are

modelled using Molecular Dynamics, whilesolid solution formation is treated with statisti-cal thermodynamic tools which combine quan-tum mechanical calculations and Monte Carlosimulations.To achieve the goals of the joint research pro-ject, experimental and analytical work is divi-ded in sub-projects which concentrate on dif-ferent aspects of calcite surfaces.

2. The calcite-water surface and its inter-action with Selenium

2.1. Structure and speciation of the calci-te-water interfaceTo characterize the calcite surface potential asa function of the composition of the aqueouscontact solution we measure calcite zeta po -ten tials using two methods. (i) Electrokineticmea surements are performed in a BrookhavenInstruments zeta potential analyzer usingphase analyses light scattering (PALS). For PALSmeasurements we use equilibrium solutions atthree different CO2 partial pressures, at atmos-pheric pressure. Suspensions are pre-equilibra-ted until the theoretically expected pH is rea-ched. In equilibrium with pure CO2 pH range is5.8 to 6.8, in equilibrium with air (360 ppmCO2) pH range is 7.5 to 9.6, and in equilibriumwith N2 (6 ppm CO2) pH range is 8.4 to 10.3.We observe an increase in the iso-electric point(IEP) with decreasing CO2 partial pressure. (ii)Streaming potential measurements are perfor-med on an Anton Paar SurPASS electrokineticanalyzer. Streaming potential is measured as afunction of pH performing pH titrations in a pHrange from 5.5 to 11. As at these conditionssolutions are undersaturated with respect tocalcite, the calcite must be dissolving duringthe measurements. Solution analysis proofs,however, that this does not have a significantinfluence on the solution composition duringthe experiments. Hardly any pH dependence isobserved in streaming potential measure-ments. Zeta potential increases with increasingCa2+ concentration and decreases with increa-sing CO3

2- concentration.


78

Surface diffraction measurements are appliedto probe if changes in zeta potential are acco -mpanied by changes in the calcite-water inter-face structure. We perform in-situ surface dif-fraction measurements at the GSECARS undu-lator beamline at sector 13 of the AdvancedPhoton Source. Five crystal truncation roddatasets are recorded at the solution condi-tions listed in Table 2.1.

Main results of the interface structure analy-ses are:– The part of the calcite(104)–water interface

structure, to which surface diffraction mea-surements are mainly sensitive, does notchange significantly, even for the investiga-ted extreme differences in the compositionof the contact solutions.

– No indication for calcium or carbonateinner–sphere complexes on the flat calci-te(104)–face is observed.

– Two well ordered layers of water are identi-fied at 2.35±0.10 Å and 3.24±0.12 Å abovethe calcite surface. Water molecules of thefirst layer are located above the surface cal-cium ions and those of the second layer arelocated above the surface carbonate ions.

– In contact with solution mainly surface car-bonate ions relax slightly from their bulkposition and tilt towards the surface byabout 4°.

– A gradual increase in Debye–Waller factorsfrom the bulk crystal to the solution isobserved.

– Elevated Debye–Waller factors of watermolecules in the second water layer in data-sets 3 and 4 (high Ca2+ concentration in

solution), as well as the drop in diffuse elec-tron density at about 4.2 Å above the surfa-ce in dataset 5 (no Ca2+ in solution), mightbe indicative for calcium outer–sphere com-plexes.

A Basic Stern Surface complexation model isdeveloped that, in accordance with the inter-face structure analyses, considers only outer-sphere adsorption of ions other than protonsand hydroxide. The observed zeta potentialsare well reproduced by the model. A majoradvantage compared to previous models (vanCappellen et al., 1993; Pokrovsky et al., 2000;Wolthers et al., 2008) is that a physically rea-sonable value for the Helmholtz-Capacitance(0.45 F/m2) is obtained. Together with theStern layer thickness, estimated from the sur-face diffraction results (~3.5 Å), this value canbe used to estimate the relative permittivity ofthe interface water, to be about 18.

2.2. Selenite (Se(IV)O32-) and Selenate

(Se(VI)O42-) interaction with calcite

The MFR study on selenite / selenate coprecipi-tation is ongoing. MFR experiments at pH 7.4show that the pH has an influence on theincorporation. As expected anion adsorptionand subsequent incorporation is increased atlower pH. Partition coefficients at pH 7.4 are~10-3 for selenate and ~10-2 for selenite, whichstill shows that selenium incorporation intocalcite is unfavorable, but the values are relati-vely large compared to the extremely low va -lues measured at pH 10.3, which were belowthe detection limit of the method. Comparison of the macroscopic calcite growth

Table 2.1: Composition and chemical conditions in the contact solutions used for the surface diffraction study.


79

rate obtained in MFR experiments in presenceand absence of selenite / selenate (Fig. 2.1)shows the retarding effect, these anions haveon calcite growth.

Selenium K-edge EXAFS is used to characterizethe adsorption species of selenate at the calci-te water interface. Adsorption samples are pre-pared in batch type adsorption experiments inpre-equilibrated calcite suspensions at pH 8.3and 7.5 using calcite powder (Merck calciumcarbonate suprapur) or freshly cleaved Icelandspar single crystal platelets as substrate. InitialSeO4

2- concentration is 2 × 10-3 mol/L. EXAFSspectra measured on powder adsorption sam-ples (black circles in Fig. 2.2) do, despite thelower signal to noise ratio, not deviate signifi-cantly from a reference EXAFS spectrum mea-sured on a 0.1 mol/L NaSeO4 solution (thickblack line in Fig. 2.2). Therefore these spectraare considered inappropriate to characterizethe adsorption species. Only spectra measured on single crystal sam-ples using grazing incidence to achieve surfacesensitivity, show an additional contribution tothe EXAFS most visible around k = 10.7 Å-1

(dashed line in Fig. 2.2) that results in oscilla-tions in the Fourier transform spectra between

R = 2.7 Å and R = 3.7 Å. These are likely con-tributions to the adsorption species spectrumoriginating from the calcite surface. DetailedEXAFS data analysis is still in progress.

3. Fixation and phase transformation ofphosphate at calcite surfacesIt is well known that phosphate is crucial forincreased primary production, but its availabili-ty as a non-renewable resource is decreasingdramatically (Gilbert, 2009; Driver et al., 1999).On the other hand, severe environmental pro-blems result through the discharge of nutrientrich waters from urban, agricultural and indu-strial sources. This leads to the eutrophicationof water bodies and the enrichment of phos-phate in sewage and wastewaters (Smith,2003). Therefore, the immobilization of phos-phate dissolved in eutrophic water bodies andwastewaters is a promising approach to solvethis problem and to recycle phosphate.In the past, calcite was used for the remedia-tion of eutrophied water to immobilise dissol-ved phosphate (Babin et al., 1994; Prepas et al,1997; Hart et al., 2003). Laboratory tests demon-strate the success of calcite application andshow that a calcite barrier can reduce the

Figure 2.1: Macroscopic calcite growth rates obtained in MFR experiments on pure calcite (Cc, circles),in presence of selenite (Se(IV), squares), and selenate (Se(VI), diamonds).


80

phosphate flux from nutrient-rich sediments tothe water. The mechanisms of phosphate fixa-tion at the calcite surface are decisive for theefficiency and sustainability (Babin et al., 1994;Berg et al., 2004), but poorly understood atthe atomic scale to date. Calcite has its pointof zero charge between pH 8.0 and 9.5 (Soma -sundaran & Agar, 1967), while many otherminerals have negative surfaces already at pH< 7. Hence, calcite exhibits a positive surfacecharge and can be used for anion adsorption.A detailed explanation of the processes ofphosphate fixation and phase transformationon calcite surfaces under varying hydrochemi-cal conditions is investigated. The scientific andtechnical objectives cover:i. the clarification of the mechanisms of phos-

phate fixation and phase transformationson calcite surfaces

ii. the characterisation of the hydrochemicalconditions and crystallographic parametersfor optimised and effective fixation of dis-solved phosphate on calcite surfaces

iii. the identification and development of calci-te products for the technical application

3.1. Experimental and analytical methodsThe uptake of phosphate (PO4

3-) by calcitefrom saturated calcite solutions has been exa-mined using three types of calcite powders, (i)a powdered powder Devonian limestone(Sancy, FR) with a specific surface area of 1.76m2/g (SSA, estimated by N2-BET, sample A), (ii)precipitated CaCO3 (Merck p.a., 102066, 0.2

m2/g; sample M), and (iii) a precipitated CaCO3

delivered by Rheinkalk Akdolit (PCC, sample B,37.45 m2/g). Different amounts of calcite pow-der (0-50 g/L) were added to saturated calcitesolutions (SICaCO3 = 0.6 - 1.2, Ca2+ = 1.8 - 2.8mmol/L, TA(HCO3

-) = 4.0 - 8.0 mmol/L), produc -ed from Merck p.a. chemicals (CaCl2, NaHCO3)and double-distilled H2O (18.2 MΩ·cm) as des-cribed by Lin & Singer (2005). Phosphate wasadded in concentrations of 3 to 1000 µmol/Las H3PO4 or Na3PO4×6H2O. pH of the solutionswas adjusted to 8.0 using HCl or NaOH. PO4

3-

concentrations of the calcite solutions are com -parable to conditions of pore water and hypo-limnion of eutrophic water bodies (Belzile et al.,1996). The batch solutions and the powderswere shaken for 24 h (Fig. 3.1).Afterwards the solutions and powders wereseparated by vacuum filtration. Residual con-centrations of Ca2+ and PO4

3- were analysed byAAS, ICP-OES and UV-VIS spectrophotometry,respectively. Total alkalinity (HCO3

-) was chek-ked using an Aquamerck quick test (Merck11109) and pH values were determined usinga WTW SenTix 81 electrode attached to aWTW pH330 pH-meter. Changes in the mor-phology of the calcite sample powders werechecked with REM. P concentrations and spe-ciation of the powders were analysed byEDXRF and XANES spectroscopy, respectively.

3.2. ResultspH values remained stable at 8.10±0.15 duringthe experimental run. (Over)saturated calcite

Figure 2.2: k3 weighted EXAFS (left) and Fourier Transform (right) spectra measured on: a liquid SeO42- reference

(thick black line), powder adsorption sample at pH 7.5 (circles), and single crystal adsorption sample measured at gra-zing incidence at pH 7.5 (dashed line).


81

solutions show decreasing trends for Ca2+ andHCO3

- and P concentrations in dependence ofthe SSA, respectively. Overall removal of phos-phate from the batch solution ranges from 2 % to 100 %, dependent on the applied cal-cite powders. In contrast to experimental seriesof sample A and M, phosphate was eliminatedfrom the batch solution of sample B seriescompletely at all starting concentrations of 3to 1000 µmol/L PO4

3-. The addition of smallamounts of calcite B (~10 g/L) results in aphosphate elimination of min. ~70 %, where-as calcite A and M result in a decrease of max.~70 %. Adding high amounts of calcite pow-der to the batch solution (≥25 g/L) results in acomplete phosphate elimination in the sampleB experimental series, whereas the PO4

3- con-centration decreases by max. ~86 % for sam-ple A and M. In the latter case, the residualPO4

3- load of the solution is increasing withincreasing starting concentration (see. Fig.3.2 for details).

In general, phosphate concentration decreaseswith increasing amount of calcite powderadded to the solution. Depending on the star-ting phosphate concentration of the batchsolution, total phosphate uptake by calciteincreases more rapidly in sample B series incomparison to sample A and M (Fig. 3.2 & 3.3).These results show that the ability of phos phateuptake can be considered as a function of theSSA which is crucial in terms of adsorption andprecipitation mechanisms and phase transfor-mations, respectively.To evaluate these mechanisms, sorption iso-therms have been determined. These iso-therms are in good agreement with those ofEiche et al. (2008). Van Cappellen (1991) divi-ded sorption isotherms for phosphate on calci-te into four sub-sections (Fig. 3.3 a) Section (A)marks a range of concentration of calcite (Cc)and PO4

3- in which only adsorption on the cal-cite surface occurs (high Cc/PO4

3- values). (B)shows the precipitation of meta-stable Ca-Pcompounds, (C) the transformation into stable

Figure 3.1: Scheme of a batch sorption experiments.

Figure 3.2: Normalized post-experimental PO43- concentrations of experimental series starting with a) 30 µmol/L and

1000 µmol/L.


82

compounds and (D) the adsorption of phos-phorus on the newly formed Ca-P surface (lowCc/PO4

3-). These sections are different for theapplied calcite powders of the experimentalseries. Figure 3.3 b) shows an example forexperimentally determined sorption isotherms(30 µmol PO4

3-/L) which can be divided in dif-ferent sections. On the basis of these observa-tions further analytical investigations will becarried out.

Processed calcite powders have been analysedusing SEM imaging and, as a novel approach,

XAFS spectroscopy (in particular XANES) is usedto identify the speciation of phosphorus andphase transformations on the calcite surface.By applying these methods, a main goal is tounderstand the different phases of fixation andsorption and the transition zone (changeoverfrom A to C in Fig. 3.3 a), respectively, on thecalcite surface at the molecular scale.First results show that XANES spectroscopy atthe phosphorus K-edge is a suitable tool toidentify the speciation of phosphorus on calci-te (Fig. 3.4). Two characteristic features areobserved at the P K-edge, the phosphate whi-

Figure 3.3: a) Schematic sorption isotherm for phosphate on calcite surfaces according to van Cappellen (1991), b) experimentally determined sorption isotherm for starting concentration of 30 µmol PO4

3-/L.

Figure 3.4: XANES spectra recordedat the P K-edge of phosphate-dopedPCC powders and crystalline referen-ces materials, dashed lines markenergies at 2152.0 eV (phosphate»whiteline«) and 2154.6 eV.Analyses were conducted at theSUL-X and INE beamline at ANKA(KIT), respectively.


83

teline at 2152.0 eV and a shoulder at 2154.6eV. The latter is characteristic for Ca-Phos pha -tes and decreases with phosphorus concentra-tion on the calcite powder, which shows achange in the coordination of phosphate. Theanalysed calcite samples are representative forparts (A) to (C) of the schematic sorption iso-therm (see Fig. 3.3 a). SEM imaging has beenperfomed for processed sample powders Aand B. Only the limestone powder shows theoccurrence of new phases in and therefore dif-ferences to the starting material. These phaseshave already been identified as Octa-Calcium-Phosphate (Freeman and Rowell, 1981).

3.3. SummaryThree calcite powders have been processed inbatch sorption experiments. Based on thepost-experimental solution chemistry, powdershave been chosen for further analytical work.Merck CaCO3, the limestone powder and thePCC powder show different shapes of thesorption isotherms. Phase transformationshave been determined by SEM imaging for thelimestone powder surfaces, but more detailedwork including using SEM and EDX on all pow-ders is necessary and in progress.First XANES results at the P K-edge show twodifferent spectral details and a change of thefeature at 2154.6 eV with decreasing P con-centration. Further XAFS spectroscopy (inclu-ding XANES/EXAFS) at both the Ca K-edge(4038 eV) and the P K-edge (2146 eV) is plan-ned. XPS and XRD analyses of the powders

and AFM studies using Iceland spar single cry-stals, respectively, are in progress and will helpto complete the characterisation of the atomicand molecular environment of phosphate atcalcite surfaces.

4. Interactions of etidronic acid and PCCpowders during the calcite crystal growthPrecipitated calcium carbonates (PCC), a puri-fied, refined or synthetic calcium carbonate,are an industrial mass product commonly usedduring the production of pharmaceutical pro-ducts, varnish and colours, paper, cosmeticsand in the food processing industry (e.g. Soutoet al., 2008). Moreover, organic additives suchas etidronic acid (HEDP, HPO3

2+), citric acid ortartaric acid are used in a wide field of indu-strial application of calcite chelating, dispersingor complexing agents in the production ofpaper, cosmetics and pharmaceutical products(Dunn et al., 1994). Our first experiments willfocus on the interaction of HEDP during thecalcite crystal growth.First batch sorption experiments investigatingthe interaction of organic additives on the cal-cite crystal growth, especially of PCCs, wereconducted in saturated and oversaturated cal-cite solutions and solution free of calcite.These solutions were produced as described insection 3.1. HEDP (etidronic acid, C2H8O7P2)was added in concentrations of 0 to 50 µmol/L(as organic agent) to the solution. For experi-ments 100 mg of calcite (Cc) powder (PCCs,

Figure 3.5: SEM images of processed limestone powders; a) section A, ~ 550 ppm P ≙ 6 at/nm2; b) section B, ~ 3600ppm P ≙ 40 at/nm2.


84

Merck p.a. CaCO3, Iceland spar) varying in spe-cific surface area (0.3 to 23.8 m2/g) was addedto 20 mL solution. Additionally, PCCs vary incrystal morphologies (scalenohedral, prismatic,rhombohedral). Afterwards, sample vials wereclosed and shaken for 24 h (Fig. 3.1). After theexperiments, all solution were filtered andchecked for alkalinity and pH. Ca2+ and P con-centrations were determined AAS and/or ICP-OES. Phosphorus concentrations of the proces-sed calcite powders were determined byEDXRF analyses.pH values remained at 8.15 to 8.35 during theexperimental run. (Over)saturated calcite solu-tions show decreasing trends for Ca2+ andHCO3

- concentrations in dependence of theSSA. Differences can be observed in solutionswhere Cc powder with a max. SSA of ~2 m2/gare used as crystal seeds. Here, the concentra-tion of Ca2+ drops more efficiently when HEDPis available. Overall decrease of Ca2+ and HCO3

-

leads to a noticeable decrease of the calcitesaturation, which was calculated using thePHREEQC software (Parkhurst & Appelo, 1999).Three experimental series with HEDP were con-ducted. Starting concentrations of 5 and 50µmol/L were adjusted for (over)saturated calci-te solutions and solutions without Ca2+ and

HCO3- (Fig. 4.1). HEDP is eliminated from a

pure etidronic acid solution by 80 to 90 %.Depending on the SSA of the applied Cc pow-der, up to 65 % of the starting HEDP is remo-ved from the calcite solution. So far, no syste-matic interconnection of HEDP sorption andcrystal morphology has been observed for SSA~ 7 m2/g. EDXRF analyses of Cc powders showP concentrations of 1000 to 1500 ppm.

First results show that the Ca2+ concentrationof a calcite solution decreases stronger whenHEDP is present. Differences in the decrease ofHCO3

- compared to solutions without HEDPhave not been observed. From our experi-ments, this is valid for calcite powders withSSA ~ 2 m2/g. This observation is supported byresults reported by Lin & Singer (2005) orSawada et al. (2003) and shows possible inhi-bition of calcite growth at these conditionsdue to precipitation Ca-P compounds. ForHEDP a linear decrease dependent on the SSAof the Cc powder has been observed.Changes in the calcite morphology and the Pspeciation and bonding at the calcite powdersurfaces will be tracked using additional analy-tical methods (SEM, XRD, XPS, XAFS) and, fur-thermore, within in-situ AFM studies. Promi -

Figure 4.1: Remaining HEDPconcentration in solutionafter experiment.


85

sing XANES measurements have been conduc-ted at the P K-edge (Fig. 3.4), further analysesare planned.

5. Reactivity of calcite surfaces duringconcreting Natural limestone plays a key role in the manu-facturing of Portland cement and thus for theproduction of a broad bandwidth of buildingproducts, such as concrete. In order to reducethe ecological impact of modern cements,limestone is added to the cement as a fillermaterial, thus reducing the cement clinkercontent. With increasing replacement rate,however, pronounced changes in the fresh andhardened concrete behaviour can be observed.At the fresh state, limestone strongly deterio-rates the working mechanisms of modern con-crete admixtures, such as superplasticizers, lea-ding to pronounced problems in the placingand handling of the fresh concrete. This beha-viour is normally attributed to the TOC contentof the natural limestone, as this might interactwith the organic matter of the superplastici-zers. Systematic investigations regarding thisimportant mechanism are not yet available.Further, limestone powder could be identifiedto trigger early hydration by acting as nucleusfor the formation of calcium silicate hydrates,which are responsible for the strength of a har-dened concrete (Matschei et al., 2007).One of the primary goals of the research sub-project described below, was to identify theinteraction mechanisms of modern concretesuperplasticizers with limestone powders (LSP)of different provenance and composition. Onthis basis, eluation tests investigating the dis-solution behaviour of limestone in both anaqueous environment (carrier liquid water) andin an alkaline environment (pH 13, 1.0 mol/dm³ NaOH/KOH solution) were carried out.The volume fraction of solids in the solutionwas adjusted to be 5 vol.%, allowing for anundisturbed dissolution process far away fromthe saturation limit. Two types of superplastici-zers, a polynaphtalene-sulphonate (PNS) whichacts electrostatically and a polycarboxylate (PC)

which acts sterically, were investigated. Eachadmixture was added to the carrier liquid at adosage of 0.5 mass.% of the limestone. Allinvestigations were carried out at 20 °C. Thesuspensions were continuously stirred duringthe investigation ensuring a homogeneousdistribution of the particles in the carrier liquid.At certain time intervals, the pH value, con-ductivity and temperature of the solution weredetermined. Further, samples of 25 cm³ of thesuspension were taken, filtrated and the filtra-te was analyzed for its Ca, Mg and K ion con-tent using the AAS method. Ca and Mg ionshave been identified in earlier studies to have asignificant influence on the rheological proper-ties of the cement suspension (Haist, 2009).As shown in Figure 5.1 (left), the Ca content inthe carrier liquid strongly increases (back -ground Ca content NaOH/KOH solution 0.14g/dm³; NaOH/KOH + PC 0.41 g/dm³; NaOH/KOH+PNS 5.4 g/dm³) in the first minutes afteraddition of the solvent, indicating that pro-nounced quantities of limestone dissolve in thecarrier liquid. This behaviour is independent ofthe presence of superplasticizer and the typeof superplasticizer. After the first dissolutionprocesses, the Ca content decreases again, in -dicating precipitation processes occurring inthe solution. This process is identical for solu-tions with and without polynaphthalene-sulfo-nates. For suspensions containing the admixtu-re on polycarboxylate basis, dissolution andprecipitation processes seem to repeat severaltimes, resulting in an oscillation of the totalcalcium content in the filtrate. The results furt-her indicate, that the organic admixtures beco-me chemically unstable over time and decom-pose after approx. 1400 min. (~24 h), resultingin a further decrease of the calcium content.Beyond this time, the precipitation process ofcalcium seems to continue, possibly accompa-nied by an oriented re-crystallization of calciumon the particle surface. However, from the loga-rithmic scaling of the time axis, no end-membercould be identified from the measured data.

Plotting the calcium content in the filtrate elu-ated by the limestone normalized on calcium


86

content in the solution (with or without admix-ture), it becomes obvious, that the dissolutionprocess of limestone and the release of calciumis strongly inhibited by the addition of bothadmixture types. The dissolution process itselfis ascribed to the chemical instability of thelimestone in the surrounding alkaline environ-ment. Calcium dissolves from the surface oflimestone powder and then precipitates againon the surface, possibly in a different crystalstructure. At the same time there is an interac-tion of the superplasticizer with the calcite sur-face. This inhibits both the dissolution and thecrystal growth, which could explain the timedependence of the calcium concentration inthe normalized diagram (see Fig. 5.1, right). Theeluation tests prove that limestone powder the-refore interacts with both investigated super-plasticizers, however in a different manor. Inorder to verify and characterise these processesin detail, surface sensitive analyses of the reac-tions between calcite and admixture will becarried out using AFM and ESEM methods.Additional eluation experiments are being car-ried out at the moment.Based on the results of the eluation tests, com-bined rheological and electro-acoustical inve-stigations on pure limestone powder suspen-sions and cement suspension with limestonepowder fractions were carried out. The rheolo-gical behaviour of the suspension was deter-

mined using both oscillatory and rotationalrheometer tests. First results of the rheologicaltests have been reported in (Müller et al.,2009). At the same time, the zeta potential ofthe particles and the medium agglomerate sizein the suspension are determined in situ usingelectro-acoustical measurement techniques.Therefore, a defined AC voltage is applied tothe sample, resulting in an electrical field in thesuspension. In case the suspended particles areelectrically charged with a charge of ζ (zetapoten tial in mV), the particles start to oscillatein the electrical field of amplitude E resulting inan oscillating pressure wave in the incompres-sible carrier liquid. This pressure signal is recor-ded using an ultrasound transducer. From theamplitude of the pressure signal p, the zetapotential ζ can be calculated using Eq. 5.1:

In Eq. 5.1 ρp and ρs denote the density of theparticle and the carrier liquid, respectively. Theparameter c denotes the speed of sound in thecarrier liquid. The phase content f in the testswas set to 0.44. hs, e0 and er describe thedynamic viscosity, the specific dielectrical con-stant and the permittivity of the carrier liquid.G(a) and f(l,w’) are correction terms. For furt-her details regarding this measurement techni-que see (Haist et al., 2009).

Figure 5.1: Calcium eluation of technical LSP in NaOH/KOH brine (pH 13, 1.0 mol/dm³) with and without superplastici-zer based on polynaphthalin-sulfonate, PNS, or polycarboxylate, PC: absolute Ca content in filtrate (left) and contentnormalized by background (right).


87

Figure 5.2 shows the influence of differenttypes of superplasticizers on the zeta potentialof limestone powder particles suspended inwater. Both the addition of polycarboxylate(PC) as well as the addition of polynaphthalin-sulfonate (PNS) leads to a strong reduction ofthe zeta potential of the particles. For theadmixture on PNS basis, this reduction is morepronounced than for the one on PC basis.However, also the suspension without super-plasticizer addition shows significant changesin the zeta potential in a time frame between17 and 25 min. after addition of water to thedry powder. The reasons for this reduction areassumed to be linked to the calcium eluationand re-precipitation observed in the eluationtests. However, the mechanisms for this pro-cess still have to be clarified by additional tests.The same is true for the pronounced peak in

the zeta potential time curve at the age ofapprox. 23 to 25 min. This peak is observedindependent of the presence and the type ofsuperplasticizer and also has to be clarified.The long-term influence of the cement repla-cement by limestone powder was studiedusing tests on hardened concrete. Therefore,prismatic specimens measuring 40 x 40 x 160mm³ were casted from the different cementpastes, demolded after one day and stored inwater until testing. On the specimens, thecompressive and the bending tensile strengthwere determined according to DIN EN 196-1.

As shown in Figure 5.3, the replacement ofcement by limestone powder leads to a syste-matic decrease in the compressive strength. Thereduction of cement, however, is not identical tothe reduction in strength. Despite the large scat-

Figure 5.2: Temporal developmentof the zeta potential of limestonepowder (LSP) in pure water, waterand polycarboxylate (PC) and wa terand polynaphthalin-sulfonate (PNS).

Figure 5.3: Influence of the ce mentreplacement by limestone powder(LSP) on the compressive strengthdetermined on prismatic specimens(40x40x160 mm³) normalized bythe compressive strength of speci-mens with pure cement.


88

ter, for cement replacements by LSP of up to 20mass.%, a cement reduction by 1 % leads to areduction of compressive strength of approxi-mately 0.7 % for the investigated limestonepowder. Further investigations are in progress.

6. Thermodynamics of mixing in carbona-te solid solutions from atomistic simula-tionsIn the past project period we have successfullymodelled the thermodynamics of cationic,(Ca/Mg)2+ = Cd2+and anionic,(CO3)2- = (SO4/SeO4)2-

substitutions in carbonates. The solid solutionshave been modelled with two different atomi-stic simulation approaches, namely, the DoubleDefect Method (DDM) (Vinograd et al., 2009)and the Single Defect Method (SDM), whichwe have developed for iso-structural and non-iso-structural substitutions, respectively. Thesubstitution (Ca/Mg)2+=Cd2+ has been studiedwithin two binary sections of the ternarysystem CaCO3-MgCO3-CdCO3. The descrip-tion of the binary section (Ca0.5Mg0.5)CO3-(Cd0.5Mg0.5)CO3, dolomite-Cd-dolomite, re qui -r ed the development of a ternary version ofthe DDM (Vinograd et al., 2010), which wasaccomplished within the project. The model-ling of the non-iso-structural solid solutionsCaCO3-CaSO4 and CaCO3-CaSeO4 with calciteand aragonite structure types was performedusing the SDM. The SDM provides an algo-rithm for the calculation of the thermodynamicproperties of virtual end-members with thecompositions of CaSO4 and CaSeO4, whichinvolves simulation of supercell structures of thehost phases (CaCO3 calcite/aragonite) with oneCO3

2- unit replaced with a SO42- or SeO4

2- unit.

6.1. The system CaCO3-MgCO3-CdCO3

The main aim of this study was to derive themixing properties along the CaCO3-CdCO3

(calcite) and (Ca0.5Mg0.5)CO3-(Cd0.5Mg0.5)CO3

(dolomite) binaries and thus to be able tomodel the co-precipitation of Cd2+ with calciteand dolomite. The secondary aim was to provi-de an additional test for a new set of empiricalpair potentials for Ca,Mg-carbonates develo-ped by Raiteri et al. (2010). The new force-field

model has shown excellent performance in thedescription of relative stability of all knownpolymorphs of CaCO3 and of hydration pro-perties of calcite. The task was to test its accu-racy in predicting the effects of mixing in thesolid solutions. These aims essentially requiredus to model the mixing properties within theternary system. The binary systems were model-led with the standard DDM (Vinograd et al.,2009). The simulation results are in the excellentagreement with the available experimentaldata. The mixing in the calcite-otavite binary isessentially ideal. This is in an agreement withthe recent experimental data of Katsikopouloset al. (2008). The calcite-dolomite and otavite-dolomite binaries show the tendency to the for-mation of the ordered intermediate compoundswith the dolomite structure. The predicted sub-solidus phase relations are in good agreementwith the experimental data of Goldsmith andHeard (1961) and Goldsmith (1972). The tem-perature-composition diagrams are shown inFigure 6.1 a) & b). Here we wish to emphasizethat this agreement did not required anyadjustment of the force-field parameters ofRaiteri et al. (2010). The Cd-O potential wasobtained by fitting to the formation energiesof several ordered compounds in the otavite-magnesite system calculated quantum-mecha-nically by Burton and van de Walle (2003). Theobtained agreement with the experimentaldata within the binaries suggests that themodel is sufficiently accurate for the composi-tions within the ternary system. Figure 6.2shows the free energy of mixing in the ternarysystem at 873 K calculated with the ternaryDDM. The miscibility gaps are shown in theprojection of the free energy surface onto theground plane. In Figure 6.3 the activities of thesolid solutions with the calcite and dolomitestructures are compared. Our results suggestthat the mixing of Cd and Ca within the orde-red dolomite structure is less ideal than withinthe calcite-otavite binary. This is the effect ofthe third component, MgCO3. The increase in»non-ideality« is due the contraction of thedolomite lattice relative to the structure of calci-te, which is caused by the presence of compactMg-layers. The next step of our research will be


89

to use the predicted activity-composition rela-tions for the calculation of Lippmann diagramsand for the assessment of relative retentioncapacity of calcite and dolomite for Cd2+.

6.2. The solid solutions of CaCO3-CaSO4

and CaCO3-CaSeO4 in calcite and aragoniteThe thermodynamic description of these solidsolutions is problematic. Indeed, all availablesolid solution theories including the DDM con-sider iso-structural solid solutions. The theoryrequires that the both end-members shouldbelong to the same space group. This condi-tion cannot be fulfilled, as it is not possible tobuild two iso-structural end-members due tothe different geometry of CO3

2- and SO42-/

SeO42- groups. The fortunate circumstance is

that due to the structural difference, thesesolid solutions are very dilute. Diluted solidsolutions are typically very disordered and thus,the regular model description is adequate.Figure 6.4 shows the enthalpies of mixing oftwo hypothetical regular solid solutions withcalcite and anhydrite structure types, for whichcomplete sets of the end-members are unavai-lable. We are concerned with the descriptionof mixing in the vicinity of pure calcite only.The enthalpy of mixing of a regular solid solu-tion in the diluted range can be well modelledwith a linear equation, which forms a tangentto the enthalpy of mixing curve. The white cir-cle denotes the excess enthalpy of a supercellstructure of calcite with a single defect of SO4

2-

substituting for CO32-. The enthalpy of this

Figure 6.1: The subsolidusphase relations in calcite-magnesite (a) and otavite-magnesite (b) systems si -mulated with the MonteCarlo method using thepairwise interactions com-puted with the DDM. Thesymbols are the experi-mental data from Gold -smith and Heard (1961)(a) and Goldsmith (1972)(b), respectively.

Figure 6.2: The Gibbs free energy of mixing inthe calcite-magnesite-otavite, CaCO3-MgCO3-CdCO3, system at 873 K simulated with theMonte Carlo method using the pairwise inter-actions computed with the ternary DDM. Thesolid curves outline the boundaries of themiscibility gaps.


90

structure necessarily belongs to the enthalpy ofmixing function of the disordered solid solu-tion, because a structure with a single defectcannot have any ordering. The enthalpy of thisstructure also falls on the tangent, because theratio of solute to host concentration is small.This implies that if it was possible to define avirtual end-member, which enthalpy falls onthe tangent line, then such a solid solutionwould be strictly ideal. The single defectmethod is based on the notion that the ent-halpy of the virtual end-member can be calcu-lated knowing the absolute enthalpy of thesingle defect structure, the enthalpy of thehost phase and the number of the exchangea-ble units in the supercell. Figure 6.5 shows thesupercell structures of calcite and aragonitewith a single defect of SO4

2- used in the study.Table 6.1 shows the results of the calculationof the enthalpies of virtual compounds withCaSO4 and CaSeO4 composition, which formideal solid solutions with calcite and aragonite.

The standard enthalpies of the virtual com-pounds can be calculated knowing the stan-dard enthalpy of a stable compound, whichhas the same composition as the virtual com-pound. Clearly, the enthalpy of this compoundshould be calculated with the same atomisticsimulation approach, which was used to calcu-late the properties of the virtual end-member.In the case of the CaCO3-CaSO4 solid solutionCaSO4 anhydrite is a natural choice for thereference compound. In the case of theCaCO3-CaSeO4 system, a stable compoundwith CaSeO4 composition does not exist.However, there is a possibility to estimate thestandard properties of a compound iso-struc-tural to anhydrite from the reaction (Eq.6.1.)

where CaSeO4*2H2O is a well characterizedcompound iso-structural to gypsum. Ourquantum-mechanical calculations suggest thatthe enthalpy change due to the reaction in Eq.

Figure 6.3: The activity-composition relations incalcite-otavite (a) anddolomite-Cd-dolomite (b)solid solutions computedwith the Monte Carlomethod.

Table 6.1: The total energies of the supercell structures and the total energies of the virtual end-members in eV calculated with SDM. The last column shows the energy of reference compounds,whose thermodynamic properties are assumed to be known.


91

6.1 is -10.85 kJ/mole. Neglecting the entropychange in this reaction and using thermodyna-mic data for CaSeO4*2H2O, CaSO4(anhydrite)and CaSO4*2H2O (gypsum) from the Nagra-PSI data base (Hummel et al., 2002), we obtai-ned -996.87 kJ/mol for the standard Gibbs freeenergy of CaSeO4 (anhydrite). Our next con-cern was to estimate the standard entropies ofthe virtual end-members CaSO4 (in calcite) andCaSO4 (in aragonite) by calculating the pho-non density of states of the relevant supercellstructures. The lattice dynamics calculationswere performed with the program GULP (Galeand Rohl, 2003). The force-field model wascombined from the model of Allan et al.(1993) for sulphates and a new model ofRaiteri et al. (2010) for carbonates. These cal-culations have shown that the standard entro-pies of the virtual compounds CaSO4 (in calci-te) and CaSO4 (in aragonite) are 121.7 and142.8 J/mol/K, res pectively. Knowing the stan-dard entropy of CaSO4 (anhydrite) of 105.32J/mol/K, and u sing the enthalpy values fromTable 6.1, we obtained the values of -1226.41and -1145.87 for the standard Gibbs free ener-gies of CaSO4 (in calcite) and CaSO4 (in arago-nite), respectively. Assuming that the entropychanges of the reactions

(Eq. 6.2)

and

(Eq. 6.3)

are the same as in the analogous reactions forsulphates, we obtained the values of -877.11and -781.81 kJ/mol for the standard Gibbs freeenergies of CaSeO4 (in calcite) and CaSeO4 (inaragonite), respectively.The obtained values permit straightforwardcalculations of the retention levels of SO4

2- andSeO4

2- in carbonates equilibrated with aque oussolutions. Assuming that the maximum possi-ble ion activity product [Ca2+][SO4

2-] in aqueoussolutions is limited by the equilibrium solubilityproduct of CaSO4*2H2O (gypsum), we can cal-culate the maximum retention levels of SO4

2- incalcite and aragonite from the reactions:

(Eq. 6.4)

(Eq. 6.5)

Figure 6.4: The virtual end-memberconstruction for CaCO3-CaSO4 sys -tem. The solid circles are the stablehost phases with calcite and anhydri-te structures. The white circle is theexcess energy of a supercell of calci-te with one CO3

2- group replacedwith a SO4

2- group. The grey circle isthe enthalpy of the virtual CaSO4end-member, which corresponds toideal calcite-type solid solution.


92

The maximum retention levels of SeO42- can be

obtained from the analogous reactions:

(Eq. 6.6)

(Eq. 6.7)

The equilibrium concentration of SO42+ in cal-

cite according to Eq. 6.4 is defined by theequation

(Eq. 6.8)

Therefore, the maximum equilibrium retentionlevel of SO4

2- in calcite at 298 K is 1.3*10-17.The analogous calculations for aragonite givean even smaller value of 8.3*10-32. The reten-tion levels of SeO4

2- in calcite and aragonitecalculated using the reactions in Eq. 6.6 andEq. 6.7 are 7.4*10-24 and 1.5*10-40, respecti-vely. These small values are essentially determi-ned by the predicted large enthalpies of thevirtual compounds relative to the enthalpies ofthe anhydrite-type phases.

6.3. SummaryOur results show that quantum-mechanicalcalculations and force-field modelling provideuseful tools for predicting equilibrium reten-tion levels of hazardous components in carbo-nate minerals. Particularly, we have demon-strated that the DDM is able to quantitativelypredict mixing properties of various iso-struc-tural binary and ternary carbonate solid solu-tions. We have also demonstrated the possibi-lity of a quantitative modelling of non-iso-struc -tural solid solutions with the SDM. The particularly interesting result is that theequilibrium retention levels of SO4

2- and SeO42-

in carbonates are very low. This implies thatthe reasonably large concentrations SO4

2- andSeO4

2- in carbonates reported in literature (Ree der et al., 1994; Pingitore et al., 1995)should be attributed to non-equilibrium ent-rapment effects.

7. Conclusion and OutlookIn the past project period, promising and use-ful results with regard to industrial applicationhave been achieved. The calcite(104)-waterinterface has been characterised by in-situ sur-face diffraction under varying hydrochemicalconditions. Moreover, experimental resultsinvestigating the interaction of the calcite sur-

Figure 6.5: The relaxed 2x2x2 and2x2x1 supercells of calcite (a) andaragonite (b) with one CO3

2- unitreplaced with SeO4

2- unit. The geo-metry minimization calculations wereperformed with CASTEP in GGA-PBEapproximation.


93

face and Se show that calcite is an efficientmedium to immobilise oxianions and EXAFSstudies characterized the adsorption species ofSe, respectively.Elimination of phosphate from calcite solutionsby calcite powders shows that P fixation is pre-dominantly a function of the specific surfacearea. Occurrence of sorption mechanisms, newCa-P phases and phase transitions, respective-ly, is different for each applied calcite powder.Systematic SEM and EDX studies accompaniedby further XRD and XAFS analyses will indicatepoints of changeover from sorption to precipi-tation. Additional AFM studies will help tounderstand inhibition and/or growth mecha-nisms on the calcite powder and single crystalsurfaces. The results achieved to date providea basis for forthcoming experimental andanalytical work and are promising for advan-ces in the application of calcite materials inwater treatment and phosphate recycling,respectively.Experiments investigating the interactions oflimestone powder and superplasticizers, withregard to concreting, show uncommon disso-lution and re-crystallisation pattern of calciteat alkaline conditions. Further analytical workis necessary to describe the observed beha-viour of the admixtures during the experi-mental run-time.Results of quantum-mechanical calculations andforce-field modelling are useful to predict equi-librium retention levels calcite minerals. In thefollowing project period, these methods will tryto solve or support specific problems investiga-ted with experimental methods in the lab.

AcknowledgementsWe wish to thank Dr. Dmitrii Kulik, Dr. EnzoCurti (Paul Sherrer Institute, Switzerland), Prof.Julian D. Gale (Curtin University, Perth,Australia) for the help in this project. The ato-mistic calculations were performed at theCenter for Scientific Computing at theUniversity of Frankfurt.

The ANKA (KIT) beamline staff at SUL-X (JörgGöttlicher, Ralph Steininger) and INE (Jörg

Rothe, Kathy Dardenne) is greatly acknowled-ged for support during test analyses andbeamtime measurements. Hartmut Gliemannis acknowledged for support during AFM ana-lyses at the Institute of Functional Interfaces(KIT-IFG).

ReferencesAusten, K., Wright, K., Slater, B. & Gale J.D.(2005) The interaction of dolomite surfaces withmetal impurities: a computer simulation study.Phys. Chem. Chem. Phys. 7, 4150–4156.

Babin, J., Prepas, E.E., Murphy, T.P., Serediak,M., Curtis, P.J., Zhang, Y. & Chambers, P.A.(1994) Impact of lime on sediment phosphorusrelease in hardwater lakes: The case of hyper-eutrophic Halfmoon Lake, Alberta. Lake andReservoir Management 8, 131-142.

Belzile, N., Pizarro, J., Filella, M. & Buffle, J.(1996) Sediment diffusive fluxes of Fe, Mn,and P in a eutrophic lake: Contribution fromlateral vs. bottom sediments. Aquatic Sciences- Research Across Boundaries 58, 327-354.

Berg, U., Neumann, T., Donnert, D., Nüesch, R.& Stüben, D. (2004) Sediment capping in eu -tro phic lakes - efficiency of undisturbed calcitebarriers to immobilize phosphorus. Appl. Geo -chem. 19, 1759-1771.

van Cappellen, P. (1991) The formation ofmarine apatit - A kinetic study. PhD thesis, YaleUniversity, New Haven.

van Cappellen, P., Charlet, L., Stumm, W. &Wersin, P. (1993) A surface complexationmodel of the carbonate mineral-aqueous solu-tion interface. Geochim. Cosmochim. Acta 57,3505-3518.

Casey, W.H., Rock, P.A., Chung, J-B., Walling,E.M. & McBeath, M.K. (1996) Gibbs energiesof formation of metal-carbonate solid solu-tions 2: the CaxSr1-xCO3 system at 298 K and1 bar. Am. J. Sci. 296, 1–22.


94

Clark, S. J., Segall, M. D., Pickard, C. J., Hasnip,P. J., Probert, M. J., Refson, K. & Payne, M. C.(2005) First principles methods using CASTEP.Zeitschrift fuer Kristallographie 220, 567-570.Driver, J., Lijmbach, D. & Steen, I. (1999) WhyRecover Phosphorus for Recycling, and How?.Environ. Technol. 20, 651-662.

Dunn, C.J., Fitton, A. & Sorkin, E.M. (1994)Etidronic Acid - A Review of Its Pharma co lo -gical Properties and Therapeutic Efficacy InResorptive Bone-disease. Drugs & Aging 5,446-474.

Eiche, E.; Berg, U.; Song, Y. & Neumann, T.(2008) Fixation and Phase Transformation ofPhosphate at Calcite Surfaces – Implicationsfor Eutrophic Lake Restoration AustralasianInstitute of Mining and Metallurgy Bulletin,292-302.

Freeman, J. S. & Rowell, D. L. (1981) Theadsorption and precipitation of phosphateonto calcite. Europ. J. Soil Sci. 32, 75-84.

Gale, J.D. & Rohl, A.L. (2003) The GeneralUtility Lattice Program (GULP). MolecularSimulations 29, 291-341.

Gilbert, N. (2009) Environment: The disappea-ring nutrient. Nature 461, 716-718.

Goldsmith, J.R. & Heard, H.C. (1961) Sub so li -dus phase relations in the system CaCO3-MgCO3. J. Geology 69, 45-74.

Goldsmith, J.R. (1972) Cadmium dolomite andthe system CdCO3-MgCO3. J. Geology 80,617-626.

Hummel, W., Berner, U.R., Curti, E., Pearson,F.J. & Thoenen, T. (2002) Nagra/PSI chemicalthermodynamic data base 01/01. Radiochim.Acta 90, 805-813.

Haist, M. (2009) Zur Rheologie und den physi-kalischen Wechselwirkungen bei Zementsus -pen sionen. Dissertation Universität Karlsruhe(TH), Germany.

Hart, B., Roberts, S., James, R., Taylor, J.,Donnert, D. & Furrer, R. (2003) Use of activebarriers to reduce eutrophication problems inurban lakes. Water Sci. Technol. 47, 157-163.

Katsikopoulos, D., Fernandez-Gonzalez, Á. &Prieto, M. (2008) Crystallization of the (Cd,Ca)CO3 solid solution in double diffusion systems:the partitioning behaviour of Cd2+ in calcite atdifferent supersaturation rates. Min. Mag. 72,433-436.

Kulik, D.A., Berner, U. & Curti, E. (2004) Mo -delling chemical equilibrium partitioning withthe gems-psi code. In B. Smith and B.Gschwend, editors, PSI Scientific Report 2003.Nuclear Energy and Safety IV, Villigen. PaulScherrer Institut.

Kulik, D.A., Vinograd, V.L., Paulsen, N. &Winkler, B. (2010) (Ca,Sr)CO3 aqueous-solidsolution systems: From atomistic simulations tothermodynamic modelling. Phys. Chem. Earth35, 217-232.

Lin, Y.-P. & Singer, P. C. (2005) Inhibition of cal-cite crystal growth by polyphosphates. WaterRes. 39, 4835-4843.

Matschei, Th., Lothenbach, B., Glasser, F. P.(2007) The role of calcium carbonate incement hydration. In: Cement and ConcreteResearch 37, 551-558.

Müller, H. S., Haist, M., Glowacky, J. (2009)Reactivity of calcite surfaces during concreting.In: GEOTECHNOLOGIEN Statusbericht 2009,Universität Bayreuth, November 2009.

Parkhurst, D.L. & Appelo, C.A.J. (1999) User’sGuide To PhreeqC (version 2) - A ComputerProgram For Speciation, Batch-reaction, One-dimensional Transport, And Inverse Geo che -mical Calculations. U.S. Geological Survey.

Perdew, J.P., Burke, K. & Ernzerhof, M. (1996)Generalized gradient approximation madesimple. Phys. Rev. Lett. 77, 3865-3868.


95

Pingitore, N.E., Meitzner, G. & Love, K.M.(1995) Identification of sulfate in natural car-bonates by X-ray absorption spectroscopy.Geochim. Cosmochim. Acta 59, 2477-2483.

Pokrovsky, O. S., Mielczarski, J. A., Barres, O. &Schott, J. (2000) Surface speciation models ofcalcite and dolomite/aqueous solution interfa-ces and their spectroscopic evaluation. Lang -muir 16, 2677-2688.

Prepas, E.E., Murphy, T.P., Dinsmore, W.P.,Burke, J.M., Chambers, P.A. & Reedyk, S.(1997) Lake Management Based on LimeApplication and Hypolimnetic Oxygenation:the Experience in Eutrophic Hardwater Lakes inAlberta. Water Qual. Res. J. Can. 32, 273-293.

Raiteri, P., Gale, J.D., Vinograd, V.L. & Winkler,B. (2010) A computational investigation of thebulk and aqueous surface properties of calcite,magnesite and dolomite (in preparation)

Reeder, R.J., Lamble, G.M., Lee, J-F. & Staudt,W.F. (1994) Mechanism of SeO42-substitutionin calcite: An XAFS study. Geochim. Cos mo -chim. Acta 58, 5639-5646.

Sawada, K., Abdel-Aal, N., Sekino, H. & Satoh,K. (2003) Adsorption of inorganic phosphatesand organic polyphosphonate on calcite.Journal of the Chemical Society. DaltonTransactions, 342-347.

Smith, V.H. (2003) Eutrophication of freshwa-ter and coastal marine ecosystems - A globalproblem. Environ. Sci. Poll. Res. 10, 126-139.Somasundaran, P. & Agar, G.E. (1967) ZeroPoint of Charge of Calcite. J. Colloid InterfaceSci. 24, 433-440.

Souto, E.C.S., Damasceno, J.J.R. & Hori, C.E.(2008) Study of Operational Conditions for thePrecipitated Calcium Carbonate Production.Advanced Powder Technology VI 591-593,526-530.

Vinograd, V.L., Sluiter, M.H.F. & Winkler, B.(2009) Subsolidus phase relations in the

CaCO3-MgCO3 system predicted from theexcess enthalpies of supercell structures withsingle and double defects. Phys. Rev. B79,10420-104209.

Vinograd, V.L., Paulsen, N., Winkler, B. & vande Walle, A. (2010) Thermodynamics of mixingin the ternary rhombohedral carbonate solidsolution (CaxMgyMn1−x−y)CO3 from atomi-stic simulations. CALPHAD 34,113-119.

Wolthers, M., Charlet, L. & van Cappellen, P.(2008) The Surface Chemistry of Divalent MetalCarbonate Minerals; a critical As sess ment ofSurface Charge and Potential Data using theCharge Distribution Multi-Site Ion Complex a -tion Model. Am. J. Sci. 308, 905-941.


96

Identification and modification of the surfaceproperties of calcite fillers as a basis for new,highly filled adhesives

AbstractPowdered calcium carbonates from naturalsources are being widely used as fillers in adhe-sive systems to improve processing and in ser-vice performance characteristics. The conver-ting of limestone and marble to adhesive fillermaterials typically includes grinding and insome cases precipitation and coating to adjustparticle size, processability and chemical reac-tion with the adhesive. It has been frequentlyobserved that calcium carbonate powder bat-ches with apparently similar particle-specificcharacteristics (e. g. density, chemical composi-tion and particle size distribution) may exhibitsignificantly varying processing properties interms of their effect on rheology, curing andadhesive performance of the adhesives formu-lation. This indicates that different calcium car-bonates as raw materials for fillers obviouslyfeature intrinsic characteristics which have notyet been identified and examined sufficientlyand whose effect on the processing and pro-duct characteristics of highly filled reactiveadhesive systems is essential to achieve a suffi-cient level of process stability, batch-to-batchreproducibility and last not least uniform pro-duct quality. The aim of this research projecttherefore is to reveal the structure-property-relationship between the manifold calcium car-bonate particle characteristics on one hand

and the physicochemical and technical proper-ties of the resulting adhesives formulation onthe other hand.

IntroductionUsing leading edge analytical methods, the keyparticle characteristics and their influence onadhesive properties shall be identified to esta-blish objectively measurable tolerance criteria.The resulting process window is intended toprovide an optimum balance of technologicaltarget values related to the storage and appli-cation properties of the adhesive as well as itsmaterial properties after curing within applica-tion specific limits.

The variable parameters of the test matrix con-sider the following factors:– geological source of raw materials– grinding process and grinding aids– grinding degree (entire spectrum of particle

size distribution)– modification state (uncoated, coated with

stearates)

The corresponding resulting analytical effort incharacterizing powdered calcium carbonateparticle properties is the following:– porosity (Hg pressure porosimetry)

Diedel R. (1), Dörr H. (1), Geiß P. L.* (2), Presser M. (2), Roth E. (3), Wittwer W. (3)

(1) Forschungsinstitut für Anorganische Werkstoffe – Glas/Keramik – GmbH (FGK), Heinrich-Meister-Str. 2,

56203 Höhr-Grenzhausen

(2) University of Kaiserslautern, Department Mechanical and Process Engineering, Workgroup Materials and Surface

Technologies (AWOK), Erwin-Schrödinger-Str., Building No. 58, 67663 Kaiserslautern

(3) Kömmerling Chemische Fabrik GmbH, Zweibrückerstr. 200, 66954 Pirmasens



97

– mineral surface (BET nitrogen adsorption)– chemical composition of the particle surface

(Tof-SIMS, XPS)– mineral composition and crystallinity (X-ray

diffractometry)– zeta potential in aqueous suspension (PCD

measuring cell)

After dissolving the different calcium carbona-te powders into representative adhesive for-mulations at a defined filler to prepolymer ra -tio the following adhesive properties are ofmajor importance:– curing kinetics (DSC measurements)– viscosity depending on shear rate (plate-

plate viscosimetry)– sedimentation and phase segregation

during storage over a period of 6 months– change in viscosity during a storage period

of 6 months– thixotropy and sagging characteristics

Chemical and physical analysis of the cal-cite fillersIn the first year of the project the physical andchemical properties of 21 calcite samples have

been analysed extensively. In general, theground samples (GC and GCC) showed a hig-her amount of contamination by other mineralphases (mainly quartz), larger primary grainsand a reduced agglomeration compared toprecipitated products (PCC samples). The uncoated filler products (GC) differ largelyfrom the synthetic precipitated ones, the sam-ples coated with stearic acid (GCC) havingintermediate properties. Depending on thequality regarding crystalline quality and chemi-cal purity these samples can be allocated toone group or the other.Based upon revision and validation of thegenerated data, specific analytical methodswhere employed to investigate the surfaceproperties of coated calcite. In particular thenature of the coating agent (mostly technicalstearic acid) has been in the focus of attention.

Simultaneous Thermal Analysis STA (ther-mogravimetry and differential scanningcalorimetry)The decomposition of the coating agent hasoccured at temperatures far below the thermaldecomposition temperature of carbonate (star-ting at 825 °C). Other components detected

Figure 1: STA of PCC 6 with derivative of the TG curve


98

were physical and chemically bounded CO2,(crystal-) water, early decomposing carbonatesor hydroxides, and miscellaneous organic con-taminants.

Fig. 1 shows the decomposition progress of acoated sample (PCC 6). The mass loss duringthe mean interval (here: 210 to 400 °C) is close– but not exactly – to the amount of the coa-ting agent (compared to the weight loss of theuncoated samples, as specified in table 1). Inaddition to CO2 and small amounts of water,due to the partial combustion of the organiccompound, a component was detected, whichcould be assigned to a technical stearic acidcomposition (a mixture of stearic acid and pal-mitic acid) by infrared spectroscopy respective-ly its decarboxylated residue. Not all sampleshowever exhibit a continuous gravimetric lossin weight. For some samples two or three dif-ferent temperature ranges of decompositioncould be observed.

For each sample the decomposition tempera-ture range was determined individually on thebasis of its gravimetric curve to minimizeweight loss not attributed to the decomposi-tion of the coating.Table 1 shows the gravimetric change of thecoated and uncoated samples in the decom-position range. The three groups of samples

can be clearly distinguished according to theircharacteristic loss in weight. The significantlyhigher values of the coated precipitated calci-tes in comparison to the coated ground fillersindicate a higher amount of coating agent dueto their smaller primary grain size. The weight loss of the samples PCC 2, 3 and 6is remarkably high and relates to the volumefracture of particles smaller than 2 µm, whilethe BET-values of these samples on a samelevel. Although GCC 7 and GCC 9 sampleshave similar specific surfaces compared to thePCCs, the mass loss rate is equal to those ofother GCCs. The data suggest that only smallamounts of the coating agent are chemicallyadhered to the particles surface area.

Zeta-PotentialThe zeta potential generally depends oneffects related to the particle surface. Hencethis parameter should be able to give evidenceof the surface occupancy. The potential wasdetermined using the DT 1200 equipmentfrom Quantachrome. The different calcite fil-lers have initially been measured in an aqueoussuspension (8 g Powder to 72 ml water) in ac -cordance with the usual procedure and titratedwith 0.1 M sodium hydroxide to show thedependence of the potential with increasingpH. Fig. 2 shows the correlation for the GCsamples. For samples GC1, GC2, GC3 and

Table 1: weight loss by decomposition of the coating agent


99

GC4 the potential decreases with increasingpH values.

While the zeta-potentials of these samples arelocated close to the isoelectric point sampleGC 5, which has a much higher specific surfa-ce (BET) and a finer grain size distribution thanthe other fillers, shows a much lower potenti-al. The observed strong variation of zeta-potential for different types of calcites has pre-viously been reported in literature (P. Ney; Zeta-Potentiale und Flotierbarkeit von Mineralen;Springer-Verlag KG; 1973; 96-101.)

Coated Calcites have a tendency to bloat inwater. Using pure ethanol in preparing thesuspension however carries the risk of distortingthe measurement by small amounts of waterabsorbed on different fillers due to their hygro-scopic properties. To eliminate the influ ence fil-ler humidity on the zeta-potential, GC, GCC andPCC samples were measured in a de fined etha-nol-water mixture containing10 w% of water.Table 2 gives an overview of the mean zeta-potential data obtained in this way and the cor-responding standard deviation (SD). The valuesin column »Data 2« indicate results of validationexperiments carried out to investigate the repro-ducibility of the described methodology.Since a well-defined correlation of these results

with other calcite properties could not yet beestablished, the data sets will be further exten-ded to provide a basis for statistical correlationanalysis, which is intended to reveal complexproperty-effect relations. For small values ahigher degree of deviation in the results needsto be taken into account. The observed varia-tions in the pH values of the samples (6.2 to 8)have been established after several hours ofhomogenisation. It should be noted that theph-values were obtained using a standard pHsensor not being approved for use in alcoholicsolution. Therefore the collected pH-values canbe used for a qualitative comparison of thebehaviour of different fillers and no statementcan be made about the absolute accuracy ofthe pH-values. It could however bee observedthat a change in pH has a significant effect onthe zeta-potential.Therefore in a next set of experiments the pH-value for all suspensions of fillers in a ethanol-water mixture was uniformly adjusted to a pH-value of 9.5 using an aqueous buffer solution(NH3-solution 25 % (8 w%) / ammonia hydro-gencarbonate (8 w%)). To avoid precipitation ofthe buffer the ethanol to water ratio had to bechanged to 50:50. The results of the zeta-potential measurements at pH 9.5 according tothis methodology are summarized in table 3.In contrast to the measurements in table 2

Figure 2: Zeta-potential as function of pH of uncoated calcites in an aqueous suspension


100

zeta-potentials at a uniform alkalinity of pH9.5 using an ethanol-water-ratio of 50:50 areall located on a negative scale. The differentresults indicate that the chemical environmenthas a strong influence on the zeta-potentialanalysis and that measurements using differentethanol /water ratios should not be directlycompared to each other.

Repeating seven measurements of the samplesGC 2, GCC 9 and PCC 4 exhibited standarddeviations of 23 %, 17 % and 11 %.In practice the homogenisation of the coatedsamples in the solution containing the chemi-cal buffer appeared to be difficult due to foa-ming and increasing viscosity of the suspen-sion. These effects increased with an increa-

Table 2: Zeta-potential in an ethanol-water-mixture (10 w%)

Table 3: Zeta-potential in an ethanol/ aqueous based buffer (50 w%) system with a constant pH value (9.5 ±0.1)


101

sing amount of coating agent but accurateresults still could be obtained.

Determination of porosityThe porosity was determined by Hg-porosimetry(with AutoPore IV 9520 from Micromeritics)using calcite powders as well as pellets, whichwere isostatically cold-pressed at 2000 bar.Both types of experiments have been used todetermine the open pore volume and the cavity.Compared to the results obtained using powdersamples the porosity of the pellets generally hasbeen shifted to lower values. This serves as anindication, that the porosity of the powderbefore compression can be characterized as acavity type of porosity. Additional SEM imaginggave no indication of open porosity up to adetection limit of approximately 0.6 µm.

DensityFor the full range of calcite fillers the densitywas measured using a helium pyknometer (tab -le 4). An increased amount of coating agentcauses a decrease of density. The difference bet-ween the measured density and the to the the-

oretical value for pure calcite (2.715 g/cm3) canhowever not exactly be related to the amountof coating agent, because technical stearicacid used in the coating process typically con-tains undefined fractions of palmitic acid andinorganic contaminants like SiO2 and MgCO3 .

Analysis of the surface interaction of coa-ting agent and calciteThe chemical and physical analysis of the diffe-rent coated and uncoated powdered calcitesamples reveals how the variable parametersof the test matrix relate to the characteristicparticle properties of the 21 types of fillers. Toanalyse the specific chemical and physical sur-face interaction between the coating agent andthe calcite surface, calcite crystals were split andthe reaction of the newly generated calcite cle-avage planes with stearic acid investigated usingFTIR-spectroscopy and ToF-SIMS.In a first set of experiments reference FTIR-spectra of the virgin calcite surfaces were col-lected at different angles of rotation of the cal-cite crystal in relation to the diamond tip of the

Figure 3: Pore size distribution of GCC 9 powder sample


102

Figure 4: Pore size distri-bution of GCC 9 pelletsample

Table 4: Density (measured with He-pyknometer)


103

FTIR-spectrometer. Due to the optical anisotro-py of the calcite crystal a constant angle ofrotation needs to be maintained in comparati-ve measurements. The virgin calcite surfacewas then coated with stearic acid at 70 °Cwhich is to a large degree similar to the indu-strial coating process of the powdered calcitefillers. The presence of stearic acid on the cal-cite surface could afterwards be easily detec-ted by FTIR-spectroscopy.To find out whether the nature of the surfaceadhesion between stearic acid and calcite isdominated by physical or chemical interaction,it was attempted desorption of excessive stea-ric acid was carried out in isopropanol atambient temperature. Surprisingly 30 secondsof immersion in isopropanol at ambient tempe-rature were sufficient to completely remove thestearic acid from the calcite surface, at leastbeyond the threshold of sensitivity of the FTIR-analysis. Compared to the spectra of the virgincalcite surface immediately after splitting, theIR-spectra of the desorbed calcite surfaces showcharacteristic peaks indicating the presence ofCO3

2- which could not be observed before thecoating and desorbing procedure.To take advantage of a higher resolution andsensitivity, additional analyses were conductedusing ToF-SIMS at the Institut für Oberflächenund Schichtanalytik (ifos) in Kaiserslautern. TheToF-SIMS measurements confirmed the resultsof the FTIR-experiments in so far, as most of the

stearic acid seemed to have been removed afterdesorption in isopropanol. However at the edgeof fracture planes a higher amount of residualstearic acid had collected which obviously resi-sted desorption due to a strong surface interac-tion. The total amount of this stearic acid residuefound by ToF-SIMS analysis however is below thedetection threshold of the FTIR-spectrometer.

The results above suggest, that a strong surfa-ce interaction between calcite and stearic acidoccurs predominantly at dislocations like edgesand fracture planes. Transferred to the case ofcoated powdered calcite fillers this observationcould lead to an explanation of the differencein filler properties due to parameters related tothe geological origin and parameters of thegrinding and coating process. Further experi-ments and analyses including defined coatingand desorption of stearic acid in combinationwith calcite crystals and calcite powders areplanned to gain further insight and understan-ding of the calcite-coating and calcite-polymerinteraction in highly filled adhesives.

Technical adhesive product validationVarious application-specific test methods havebeen established to characterize the uncuredadhesive material In terms of e.g. storage sta-bility, extrudability, sagging behaviour, sprea-ding and wetting. The difference in curing cha-racteristics, which have been previously investi-gated by Dynamic Scanning Calorimetry (DSC),in practice determine the pot life and the incre-ase of shore hardness as a measure for thecuring adhesives cohesive strength.In this part of the project the correlation bet-ween the chemical-physical properties of thedifferent calcite fillers (published in the firstscience report of this project) and the technicalproduct properties of a polyurethane adhesiveformulation including these fillers, have beeninvestigated.

Sedimentation behaviour after storageSedimentation during storage is an undesiredphaenomenon because segregation of theadhesive formulation leads to non-uniform

Figure 5: ToF-SIMS mapping of the distribution of stearicacid on the surface of calcite crystals after desorption inisopropanol (negative polarity)


104

product properties (Fig. 6). If segregationoccurs during storage, the liquid and solid pha-ses have to be remixed and homogenisedbefore application which causes increased eff-ort and expenses.After preparing adhesive formulations usingthe different types of uncoated and coated fil-lers the amount of supernatant was measuredafter a storage period of six months at roomtemperature and at 50 °C (Fig. 7).

Especially after storage at 50 °C a correlationbetween the used calcite filler and the sedi-mentation in the polyurethane adhesives Acomponent becomes obvious. The GC typeswith low BET values generally tend to segrega-te more than the GCC calcites with larger BET

values. GCC 8 represents in so far an excep-tion, as its level of sedimentation characteristicis more similar to the GC types of fillers.Because of their small particle size causinghigh adhesive viscosities, all PCC fillers exhibitno sedimentation at all.

Extrusion behaviourThe extrusion behaviour is an important featu-re directly related to the dispensing of an adhe-sive. In hand operated dispensing tools theextrusion behaviour relates to the effort neces-sary to squeeze a certain amount of adhesivesthrough the dispensing tip. The extrusion ratewas determined using a pneumatic cartridgedispenser at an operating pressure of 5000 hPaand measuring the amount of adhesive dispen-

Figure 6: Sedimentation afterdifferent storage times at 50 °C(1, 4, 8, 12 weeks)

Figure 7: Sedimentation behaviour of polyurethane adhesive with different calcites


105

sed in a defined period of time (Figure 8).Results in Figure 9 are expressed in g/s units.

Similar to the results shown in the chapterbefore the extrusion rate depends on the phy-sical characteristic of the calcite filler. The GCformulations with a low viscosity, based on thelow BET of the filler, have a higher extrusionrate in opposition to the GCC and PCC types.The same exceptions as in Figure 7 are visible,GCC 7 is similar to the PCC group and GCC 8looks like a GC calcite.

SaggingSagging relates to flow of an adhesive ropeunder the effect of gravity (Figure 10). To de -ter minate sagging according to EN ISO 14678,an adhesive rope with a certain diameter is app-lied to a vertical metal plate and the distance offlow after one minute is measured (Figure 11).

Figure 11 illustrates, that sagging predominant -ly occurs when uncoated GC fillers are beingused. GCC8 and GCC1 however also exhibit asignificant amount of sagging. The saggingcharacteristic of the GC types of fillers relates

to their BET value and particle size.Sagging does generally not occur when PCCtypes of fillers are being used in the adhesivesformulation.

Curing characteristicsCuring of the reactive adhesive starts aftermixing the A and B component. The term »potlife« relates to the period of time during whichthe mixed adhesive can be handled and app-lied without the risk of loss of adhesive perfor-mance due to an advanced level of curing lea-ding to a higher viscosity of the adhesive andpoor wetting properties. Once the transitionfrom the viscous to the solid state of the adhe-sive is accomplished, the shore hardness givesan indication of the attained level of cohesivestrength until the curing reaction is completed.

Pot lifeThe pot life was determined by measuring theincrease in viscosity after mixing both reactivecomponents. The pot life is considered to haveexpired, once the viscosity has doubled its initi-al value (Fig. 12).

It should be assumed, that the chemical reac-tion between the coating agent (stearic acid)and the adhesives formulation causes a reduc-tion in pot life. The exception represented byGCC2, PCC4 and PCC7 however suggest, thatinverse effects related to other particle proper-ties can cause prolonged pot life even in thepresence of coated fillers.

Shore A hardness The progression of the shore hardness in timerelates to the speed of cure of a certain adhe-sive formulation. This parameter is especiallyimportant to determine the time, when anadhesively bonded structure has developed asufficient initial load capacity and can be hand-led for further manufacturing steps. To accessthis characteristic a specimen with definedthickness (5 mm) is cured under normal clima-te at 23 °C and 50 % r. h. and the shore A hard-ness is measured (Fig. 13).As in the case of the pot life, the results of theshore A hardness after 6 hours of curing show

Figure 8: Measuring the cartridge extrudability of anadhesive-filler-formulation


106

Figure 9: Extrusion rate at 5000 hPa and 23°C with a polyurethane A component including different fillers

Figure 11: Flow behaviour of polyurethane A-component with different calcite fillers

Figure 10: Sagging of an adhesive rope on a vertical adherent surface


107

a non-uniform characteristic. To identify multi-dimensional property-effect-relations sophisti-cated mathematical methods of regressionanalysis have been applied in the course of theproject taking advantage of the generateddatabase now comprising sets of analyticalparticle characteristics, chemical and physicaladhesives properties as well as technical pro-duct features.

Analysis of property-effect-relations ofcalcitic fillers in adhesivesThe data based mathematical analysis is focu-sed on gaining insight into the influence ofintrinsic properties of calcitic fillers on the che-mical/physical as well as on the technical pro-perties of the corresponding formulated adhe-sives. For this purpose data from validatedmeasurements were provided for 21 differentcalcitic fillers, each of them belonging to oneof the three groups GC, GCC or PCC.

Figure 12: Pot life of 2 component polyurethane mixture with different fillers at 23 °C

Figure 13: Shore A hardness of a polyurethane mixture with different fillers after different curing time


108

For the basic characterization of the differentfillers various parameters concerning for exam-ple the chemical composition, the zeta poten-tial, the specific surface and the distribution ofparticle sizes were chosen. In order to keep thenumber of potential influence parameters assmall as possible, the shapes of the particledistributions were analyzed in such a way thatthey can well be approximated via 5 parame-tric bimodal densities. A further analysis of theso-defined input parameter space was carriedout to detect and eliminate multicollinearities.As a first result it could be confirmed that 6 ofthe input variables can be explained on a 95%-level via linear combinations of the remainingindependent ones.

In the next step of this subproject mappingsbetween the set of remaining influence para-meters and the set of the corresponding adhe-sive properties, the so-called output parame-ters, will be identified via the estimation ofmathematical regression models. In order toaccount for the unbalance between the num-bers of model parameters to be estimated andthe available measurements, a further reduc-tion of the input space via principal compo-nent analysis, respectively factor analysis willbe performed previously. As models for regres-sion purely linear models as well as modelsthat are non-linear in their parameters will beconsidered. In parallel a partial least squaresregression (PLS) approach, that simultaneouslyconsiders the structure of the output space,will be investigated. Finally the best model foreach of the adhesive properties will be identi-fied via cross validation techniques. A study ofthe corresponding relevance diagrams that arecomputed from the identified models then isexpected to reveal the impact of the individu-al calcitic filler properties on the chemical/phy-sical as well as on the technical properties ofthe corresponding formulated adhesives.

OAP process optimizationBased on classical quality management toolsaccording to Ishikawa, FMEA and R&R, theFGK has developed a method of process capa-bility analysis (OAP method: Output AcceptedProcess), with which natural mineral raw mate-rials (clays, kaolins, feldspars, carbonates) aretested for their influence and their stability inmanufacturing processes. To identify the domi-nant process parameters and their range, tocontrol the coating process in an industrialenvironment, a first inventory of the proces-sing steps, the critical parameters and theactual process control was performed accor-ding to the OAP® procedure in cooperationwith the industrial partner. In this inventory theactual process and product flow as well as thestatus of the process control (including thehandling of non-conformities) »as is« was ta -ken into account, indicating a high degree ofprocess capability and closed loop controls. Onthe basis of the chemical and physical analysisof the calcite fillers additional parameters havebeen identified, which may be used e. g. in thefiller coating process as a basis for raw materi-als and process control. Due to the lack of suf-ficient detection and measurement possibilitiesthese parameters can however only be asses-sed as sum-parameters, which impedes theirspecific use and integration in the actual pro-cess control.Therefore during the upcoming phase of theproject additional effort will be taken to qualifymethods like e.g. the measurement of the zeta-potential of a filler suspension and to imple-ment them into an enhanced process control,as described in the phase model presentationof the adhesive production (Fig. 14).


109

SummaryIn the current stage of the project the identifi-cation of the intrinsic, chemical-physical mate-rial characteristics of the filler particles to alarge degree has been accomplished. Theinfluence of various types of fillers on the flowbehaviour and curing reaction of adhesiveswas previously investigated by means of rheo-logy and analysis. Technical test methods,which are specific to the use of adhesives, havenow been applied adjacent to the analytical

techniques of the objective chemical and phy-sical property characterization. The technicalproduct validation comprises aspects related tostorage stability and processing properties(before curing), and »material properties« (af -ter the curing process). The acquired experimental data are being cor-related and evaluated by means of advancedregression analysis. As a first result it could beconfirmed that six of the filler property inputvariables can be explained on a 95 %-level via

Figure 14: Simplified phase model presentation of the adhesive production process using calcite fillers, including theadditional control factors specified for the coating process


110

linear combinations of the remaining indepen-dent ones. In the next step mappings betweenthe set of remaining influence parameters andthe set of the corresponding adhesive proper-ties, the so-called output parameters, are iden-tified via the estimation of mathematical re -gres sion models. In order to account for theun balance between the numbers of model pa -rameters to be estimated and the availablemeasurements, a further reduction of the in -put space via principal component analysis,respectively factor analysis will be performed.To analyse the specific chemical and physicalsurface interaction between the coating agentand the calcite surface, calcite crystals weresplit and the reaction of the newly generatedcalcite cleavage planes with stearic acid inve-stigated using FTIR-spectroscopy and ToF-SIMS. The ToF-SIMS measurements and theresults of the FTIR-experiments reveal that alarge amount of coated stearic acid can easilybe removed from the calcite surface by des-orption in isopropanol. However at the edge offracture planes a higher amount of residualstaric acid had collected which obviously resi-sted desorption due to a strong surface inter-action. Projected to the case of coated powde-red calcite fillers this observation could lead toan explanation of the difference in filler pro-perties due to parameters related to the geolo-gical origin and parameters of the grindingand coating process. Further experiments andanalyses including defined coating and desorp-tion of stearic acid in combination with calcitecrystals and calcite powders are planned togain further insight and understanding of thecalcite-coating and calcite-polymer interaction.To identify the dominant process parametersand their range, to control the coating processin an industrial environment, a first inventoryof the processing steps, the critical parametersand the actual process control was performedaccording to the OAP® procedure in coopera-tion with the industrial partner. On the basis ofthe chemical and physical analysis of the calci-te fillers additional parameters have been iden-tified, which may be used e.g. in the filler coa-ting process as a basis for raw materials andprocess control once the corresponding analy-

tical methods have been sufficiently qualified fortheir application in an industrial environment. In the future the knowledge of the technicallyrelevant characteristics of the calcite filledadhesives provides the basis for avoiding therisk of deviating batches due to fluctuations inthe supply of raw materials and the potentialto reduce the effort in selecting new cost-opti-mised sources of raw materials. As a result thedevelopment cycles of new adhesive productswith improved application characteristics areshortened and new adhesive products withimproved characteristics will be available toe.g. the transportation or the building andconstruction industry.


111

InProTunnel: Interfacial Processes betweenMineral and Tool Surfaces - Causes, Problemsand Solutions in Mechanical Tunnel Driving

AbstractDuring mechanical tunnel driving in fine grai-ned soil or rock the excavated material oftensticks to the cutting tools or conveying system,which may cause great difficulties in its exca-vation and transport. In the InProTunnel projectthis problem is faced on different scales parti-cularly for the method of Earth Pressure Ba -lanced (EPB) shield tunnelling. Major influencesfrom tunnel boring machine (TBM) operationare identified by project data analyses, clog-ging propensity is evaluated by a new labora-tory test, alternative manipulation techniquesare developed based on modifications of thephysico-chemical and electroosmotic beha-viour of clay minerals and will be tested in anespecially designed clogging test system.

1. IntroductionIn the course of the joint research project »Inter -facial Processes between Mineral and ToolSurfaces - Causes, Problems and Solutions inMe chanical Tunnel Driving« (InProTunnel) theproblem of clogging during tunnel driving infine grained soil or rock is investigated particu-larly for the method of EPB shield tunnelling.

Three university institutes and four industrialpartners are working on new techniques toidentify the relevant effects and interactions ondifferent scales. Furthermore, new manipula-tion methods for the avoidance or reduction ofadhesion-caused impacts are investigated. Theresulting proposals for solution will have apositive effect on the efficiency, sustainabilityand profitability of tunnelling projects.

In a first step, data deriving from TBM opera-tion, geological and environmental parametersare analysed (cf. ch. 2). Detected correlations areused to identify positive or negative effects withrespect to the sticking of excavated material tothe cutting tools or transportation equipment.

Secondly, new laboratory tests to investigateand classify different soils and rocks are deve-loped as until now no suitable test procedureor classification scheme for the clogging po -tential is available (cf. ch. 3). The adhesion ofclays or clayey soft rocks in mechanical tunneldriving has already been investigated in severalresearch projects (Jancsecz 1991, Wilms 1995,Thewes 1999, Burbaum 2009); nevertheless,no generally accepted (standardized) test cur-

Feinendegen M. (1), Spagnoli G. (2), Stanjek H. (3), Neher H.P. (4), Ernst R. (5), Weh M. (6),

Fernández-Steeger T. M. (2)*, Ziegler M. (1), Azzam R. (2)

(1) Chair of Geotechnical Engineering and Institute of Foundation Engineering, Soil Mechanics, Rock Mechanics and

Waterways Construction, RWTH Aachen University

(2) Chair of Engineering Geology and Hydrogeology, RWTH Aachen University

(3) Clay and Interface Mineralogy, RWTH Aachen University

(4) Ed. Züblin AG, Stuttgart

(5) Herrenknecht AG, Schwanau

(6) Marti Tunnelbau AG, Bern, CH



112

rently exists to determine the clogging beha-viour from a practical (tunnel) constructionpoint of view.

Clay minerals show a mechanical behaviourwhich cannot simply be extrapolated from clas-sical soil mechanics. The mechanical propertiesof fine-grained soils like clays are intimately rela-ted to the chemical properties of the respectiveminerals and pore fluids. Ad di tionally, it is wellknown, that electric fields strongly affect theme chanical behaviour of soft soils. Since theseprocesses are of great importance for the occur-rence of adhesion and/or clogging as well as forthe development of methods for their reductionthe acting mechanisms are investigated on themicro scale in a third part of the project (cf. ch. 4).

Finally, new manipulation techniques based onchemical modifications as well as electroosmo-tic methods will be tested in a realistic testsetup. These tests are presently in the designstate (cf. ch. 5).

2. Data Analysis from TBM tunnellingprojects

2.2. Project DatabaseThe data generated by tunnel driving withTBM (penetration, thrust forces, pressures ofannular grouting etc.) are crucial informationfor the interpretation and prediction of advan-ce rates and potential troubles. The same applies

for survey data and geological or geotechnicalinformation.

Based on a powerful SQL database an applica-tion to collect, store, process and analyse thedata of tunnel projects has been developed.Project partners have direct access to the data-base by means of a web-based graphical userinterface (GUI). The application is designed asa socalled »thin-client« based on a central ser-ver database. Thus the application is indepen-dent from operating system and can be acces-sed from anywhere via internet. Only an inter-net browser is needed to run the application.

The basic function of the database is the sto-rage of TBM data. Moreover, survey and geo-logical/geotechnical data or any other type ofprocess data referenced to chainage or timecan be incorporated. All kinds of data can beprocessed and analyzed in graphical or tabularformat regarding chainage or time. Stan dar diz -ed charts and tables as well as individual onescan be generated. Even a multiaxes evaluationof the data is feasible. Thus, the system iscapable to interconnect all data of a project tobecome aware of existing correlations. De tec -ted correlations can be used to identify positi-ve or negative effects from different parame-ters with respect to e.g. the degree of stickingof excavated material.

In the framework of the InProTunnel project da -ta from one tunnel project in southern Germany,

Figure 1: Graphical multi-axes evaluation of TBM and geological/geotechnical data


113

where clogging problems were encountered, areavailable in the database. In figure 1 values forthe local water inflow at the face, the relativeamount of sticking of excavated material tothe cutting wheel as well as the contact forceof the cutter head are given for an approxima-tely 3 km long tunnel part. More over, therespective geology is plotted in the background.The chart shows, that the clogging of the cutterhead is strongly related to the water inflow atthe face which is directly linked to the respec-tive geology. Another fact is, that the stickingof excavated material to the cutting tools leadsto an increasing contact force of the cutterhead, when regarding the respective geology.The needed energy and the correspondingconstruction costs increase.

2.2. Influences on clogging from geologyand TBM operationWhen regarding the processes that lead toclogging problems in a TBM drive one has todistinguish between primary and secondarycauses. Primary causes are mainly the follo-wing geotechnical conditions:– The composition of the subsoil, especially

the type and amount of clay minerals,– The slaking durability,– The water content of the soil prone to clog-ging and/or the availability of free water in theadjacent soil.In this connection not only one layer or type ofsoil has to be regarded, but all materials thatwill be excavated at the same time have to betaken into consideration.

The secondary causes of clogging result fromthe interaction between cutting technology andsubsoil. During the excavation and transport ofthe material the mechanical wear causes a lossof strength which may even lead to a completedisintegration of the composite structure. Thevarying consistency of the fine grain fractionand its percentage with respect to the total soilmass are major influences on the clogging pro-pensity. In several TBM drives it could be obser-ved, that a higher sand fraction in the excavatedmaterial led to considerably lower adherences.Adherences may occur at all surfaces that are

in contact with the excavated material duringexcavation, transport or disposal. The materialfirst sticks to the tool surfaces due to adhesionforces. Clogging problems will follow whenthe amount of adhering soil leads to a narro-wing in the transportation ways: – The material is compacted and pressed to

the tool surfaces or already adhering soil. Bythis, also material that normally shows noclogging propensity often tends to stick.

– Because of the possibly high pressure theadhesive forces may be considerably higheras well.

– If the inner strength of the soil (cohesion) ishigher than the forces acting during excava-tion and (unpressurised) transport soilagglomerations can persist. When pressedthrough narrowings these aggregates maybe squeezed or sheared. This leads to anexposure of their inner parts, where thewater content and the clogging tendency isoften much higher than in the rest of theexcavated material.

If the adhering soil has no lateral support, highadhesive forces and/or cohesion is needed toprevent it from breaking of. Usually clogging isobserved much less on moving parts, where notonly gravitational forces like on stationary parts,but also rotational or shear forces are acting.However, the mechanical wear of the materialin connection with the machinery working onthe TBM leads to a massive rise of the tempera-ture, which again results in very tough and per-manent adherences. This can be very unfavou-rable, when these hard adherences act as akind of support for softer material. For exam-ple, during a TBM drive the roller bit boxes areoften filled with fairly wet soil. On the frontsideand the backside this material is sliding (rota-ting) over the material in the excavation cham-ber or at the tunnel face, which frequently leadsto the described effects. Such kind of adheren-ces can be very long-lasting and can be carriedfar beyond the areas of subsoil that is originallyprone to clogging.

As already seen before (cf. Fig. 1) the occur-rence of clogging is very much dependent on


114

the availability of water and the resulting con-sistency of the excavated material. When theoriginal subsoil is relatively dry and only thetunnel driving operations cause a contact withwater, then sometimes layered adherences areformed. In figure 2 heavily protracted adheren-ces with such a layered structure are shown.The original colours are grey in the inner part(1), brown (2), grey (3) and light grey on thesurface. When the picture was taken, theouter cutting wheel openings were almost freefrom clogging. Only the roller bit boxes sho-wed the protracted adherences from earlieroccurrences of water inflow. In the cross sec-tion A-B it can be seen how the adhering soilhas been pulverized in the (rotating) contactsurface to the material in the excavation cham-ber, has lost water due to the high temperatu-re and changed its colour from a dark to a ligh-ter grey. This very hard layer acts as a kind ofprotection for the softer soil within the rollerbit box, which had to be removed by hand inthe end.

3. Laboratory testing of the cloggingbehaviour

3.1. Cone pull-out testFor a better identification and quantification ofthe mechanisms affecting the clogging beha-viour on a laboratory scale, the so called »conepull-out test« has been developed (Feinen de -gen et al. 2010). The equipment and the testpro cedure are shown in figure 3.

The sample material is compacted in a stan-dard proctor device, a steel cone is insertedinto a pre-drilled coneshaped cavity and loa-ded for 10 minutes with the magnitude of theapplied load between 3.8kN/m² and 189kN/m²depending on the consistency. The load is thentaken off and the specimen is placed in a teststand where the cone is pulled out with a velo-city of 5 mm/min. The tensile forces and thedisplacements are recorded.

Different clay types with varying mineralogyand plasticity have been tested up to now in anumber of test series with different cones andsoil consistencies. Table 1 shows the relevantproperties of some selected clays.

3.2. Preliminary testsIn the following, some exemplary results ofpreliminary tests on the so called »clay 3«, ame dium plastic clay from a quarry in the Wes -terwald (Germany) are illustrated. One impor-tant difference -amongst others- was, that thetest cones in the later main experiments (cf. ch.3.3) consisted of stainless steel, which has dif-ferent surface characteristics than the con-structional (black) steel that was used in thepreliminary tests. The results thus cannot becompared directly.

It should be mentioned, that all curves shownin the following normally represent the meanvalues of four tests. Only when the deviation istoo large, the respective data are neglected. Figure 4 shows the progress of the vertical ten-

Figure 2: Adherences in the roller bit box of an EPB-TBM


115

sile stresses for clay 3 tested with differentcone inclinations at a consistency of IC = 0.70.It can be seen, that with the »nearly flat« cone0 (10°) tensile forces can only be measured fordisplacements less than 3mm, while with the»steep« cone 4 (72.6°) they are acting over aquite large range up to 11mm. After severalcomparative tests, only cone 3 (58°) was usedfurthermore, since it provided the most cha-racteristic results for all analysed soils.In figure 5 the respective results for differentconsistencies tested with cone 3 are shown.

Here the stiff material (IC = 0.85) shows quitehigh tensile stresses at very short displacementways whereas for the softer material the maxi-mum decreases with tensile forces still actingover large ways. An integration of these tensi-le stress-displacement curves delivers the »PullEnergy« which represents the energy that isneeded to pull the cone out of the soil.

Figure 3: Cone pull-out test

Table1: Tested clays


116

In addition to the tensile stress measurements,after each test the mass of adhering soil (cf. fig.6) was determined by weighing. This parameteris defined as »adherence« in the following.

When plotting the pull energy, which was deri-ved from the tensile (= bond) stresses, over theconsistency and comparing it to the measuredadherences, a good correlation could be obser-ved from the first tests series (fig. 7). Both theclogging potential as well as the adherenceshow relatively high values in a soft to stiffconsistency and a decrease towards the »wet«and the »dry« side. This corresponds quite wellwith the experiences from practice (Weh et al.2009a, b). Clogging problems mainly occur,when the material in the excavation chamberis in or gets into a plastic state. They are usual-ly much smaller when the soil is very dry ornear liquid.

3.3. Main experimentsIn the main experiments four different clayswere examined: clay 13, like clay 3 from theWesterwald; clay 14, the extremly plastic socalled »Ypresian Clay« (= »London Clay«); clay15, the again well-known »Boomse Klei« andclay 16, which represents a pure smectite. In figure 8 the curves of the pull energy overthe consistency are shown for these four clays,whereas in figure 9 the corresponding ad her -ences are plotted. It can easily be seen, that the(quantitative) correlation of these curves is notalways as good as in the preliminary tests, butthe (qualitative) shape is still somehow similar.Nevertheless, from figure 9 a direct compari-son of the four soils can be drawn. Clay 14,the »Ypresian Clay«, shows a characteristicsteep developing with a very high maximum ofadherence (1150 g/m²) at a consistency of IC =0.54. The curves for clay 15, the »Boomse

Figure 4: Tensile stress - displacement curves for differentcone inclinations

Figure 5: Tensile stress - displacement curves for differentconsistencies

Figure 6: Adhering soil for IC = 0.20 and IC = 0.4 (0° = viewing direction)

Figure 7: Pull energy and adherence over consistency


117

Klei« and the pure smectite (clay 16) are moreeven with considerably lower adherences of275 g/m² and 350 g/m² respectively. Finally, forClay 13 from the Westerwald the maximumvalue of 700 g/m² lies in between, but the cur -ve is a bit irregular.

3.4. Evaluation of the clogging potentialBased on the results obtained so far, a draft ofa classification scheme to quantify the clog-ging potential of different fine-grained soilshas been developed (fig. 10). Classes of high,medium and low clogging potential are assig-ned to the diagram with the adherences deri-ved from the cone pull-out tests plotted overthe consistency.

One can directly identify the very high risk ofclogging for clay 14. In particular, the sharpincrease of the curve up to the maximum of1150 g/m² is a strong indication for problemsto be expected. Also Clay 13 with a maximumadherence of 700 g/m² will lead to extensive

clogging. But also for the two other clays note-worthy adherences have to be reckoned with.

In this connection it should be mentioned, thatall clays presented here were deliberately cho-sen as »sticky« clays. Other samples were exa-mined that showed adherences of less than100 g/m² and can thus be defined as »notprone to clogging«. Anyhow, the ranges ofhigh, medium and low clogging potential areup to now defined arbitrarily. There is still astrong need for a verification by the EPB tun-nelling praxis!

4. New manipulation techniques

4.1. Theoretical background of the chemo-mechanical coupling of claysClay minerals are characterized by strong elec-trical attractive and repulsive forces that varysignificantly in magnitude depending on mine-ralogical composition and the charge on theirsurface as well as on their edges. Thus the claymineralogy becomes important in geotechnicalengineering with fine-grained soils. The regionin the particles size vs. stress field wherechemo-mechanical coupling may take placereflects the relevance of double layer pheno-mena as well as the relative balance betweenlocal contract-level electrical forces and boun-dary-skeletal forces resulting from Terzaghi’seffective stress (Santamarina et al. 2002). The chemo-mechanical coupling plays animportant role for the engineering response ofthe material. Numerous authors have already

Figure 8: Pull energy over consistency

Figure 10: Draft of a classification scheme

Figure 9: Adherence over consistency


118

investigated the problem (Mesri and Olson1970; Anandarajah and Zhao 2000; Spagnoliet al. 2010a). Dielectric constant and electroly-te concentration of the pore fluids are impor-tant controlling parameters, particularly in thecase of 2:1 expanding clays such as Smectite.Kaolinite (1:1 non-expanding clay) and Illite(2:1 non-expanding clay) do not respond in thesame way. In fact, they have a very thin diffu-se double layer (DDL) expansion (Gajo andMaines 2007) due to the negligible isomorphicsubstitution. With increasing salt concentra-tion the DDL thickness decreases leading tosmaller specimen volume. For the followingdiscussion, it will be assumed that the DDL the-ory is applicable.

Clay particle systems are frequently describedas a series of parallel clay particles. The Pois -son-Boltzmann equation (Mitchell and Soga2005) for a single particle can be used toobtain the midplane electrolyte concentrationand potential between two particles. Anapproximate indication of the influences ofparticle spacing and pore fluids chemistry canbe seen in terms of thickness of the DDL.However, the variation of liquid limit forKaolinite with fluids of different dielectric con-stant or electrolyte concentration is almostnegligible compared to the variation whichoccurs for Smectite (Fig. 11). The difference inchemo-mechanical behaviour of Kaolinite andSmectite is probably due to the existence oftwo different mechanisms governing the liquidlimit. One mechanism might be controlled by

the thickness of the DDL which governs theliquid limit. Sridharan and Venkatappa Rao(1975) stated that the liquid limit of soils ismainly controlled by the amount of DDL heldwater. The most important conclusions, con-cerning the structure of the double layer asfunction of the electrolyte concentration (and /or dielectric constant) of the fluid is, that theextension of the double layer in solution decre-ases with increasing electrolyte concentration(or decreasing dielectric constant).

This can be seen from the fact that the liquidlimit for Smectite is 455 % with water and89% with a non polar fluid. Since liquid limit isthe amount of fluid which must be added to asoil to allow the layers most distant from thesoil particle to acquire the properties of freewater (Sridharan et al. 1988), it can be statedthat the fluid content at the liquid limit is con-tributed also by the fluid content due to thediffuse double layer. Therefore, the liquid limitis a measure for water (or fluid) held as doublelayer water (held with rigidity) plus water (orfluid) in the liquid state. For this reason, thedecrease in liquid limit for Na-Smectite withnon-polar fluids can be explained by a sup-pressed or thin double layer which is in accor-dance to equation (1).

(1)

where 1/k is the DDL thickness (in nm), T is theabsolute temperature given in K, n0 the ionicconcentration in the bulk solution in mol/l, e0

Figure 11: Variation of the liquid limit of Smectite and Kaolinite for different dielectric constant (left) or electrolyte con-centration (right) of pore fluids (after Spagnoli et al. 2010b)


119

and e are the electric permittivity of vacuum andthe relative dielectric constant of the pore fluid,respectively, whereas u² is the valence of theprevailing cation. The constants are the elemen-tary charge (e = 1.602·10-19 C) and the gas con-stant (R = 8.3145 J/mol·K) (Israelachvili 1991). For Kaolinite a change in the fluids dielectricconstant (or electrolyte concentration) doesnot lead to any appreciable change in doublelayer thickness. In addition, it could be statedthat the DDL for non-swelling clays is verysmall or not existing. Hence, the diffuse doublelayer approach does not apply for such clays.

4.2. ζ-potential measurementsThe electrokinetic behaviour of clay suspen-sions was investigated using an electroacoustictechnique (DT1200 ζ-potential analyzer, Dis -persion Technology, Inc., USA), which is basedon the determination of the dynamic mobilityobtained from the electrokinetic sonic amplitu-de signal (ESA) of the charged particles in non-dilute suspensions (Galassi et al. 2001). Simul ta -neously the electrophoretic mobility µ is calcula-ted using the following equation (Duk hin andGoetz 2002):

(2)

where e is the dielectric constant of pore fluids,e0 the dielectric constant of the vacuum, ρp thedensity of the particles, ρs the density of thedispersion, �ρm the density of the medium, G isa function, Du the Dukhin number, w the fre-quency and j the volume fraction (Dukhin andGoetz 2002). The advantage of this techniqueis the ability of propagating ultrasoundthrough samples that are not transparent for

light and therefore this technique offers a uni-que opportunity to characterize concentrateddispersions (Dohnalová et al. 2008).

In figure 12 the variation of the ζ-potential fordispersions of Kaolinite and Na-Smectite withdifferent ionic strength and pH is shown. It canbe observed that for Na-Smectite, the ζ-poten-tial is negative for the whole pH interval. It isessentially independent of pH. In contrast, theζ-potential of Kaolinite becomes more negati-ve with increasing pH. Increasing the ionicstrength the values of ζ-potential become lessnegative. With a 1M electrolyte concentrationfor both materials, the ζ-potential is slightlypositive (between 2 and 3 mV). Experimentalresults show that the ζ-potential is in generalmore negative at high pH, low ionic concen-tration and low ion valence (Santamarina et al.2002). The increasing negative ζ-potential ofKaolinite in case of pH rise can be attributed tothe deprotonation of the aluminol group onthe edges of the Kaolinite, forming a comple-xed anion which contributes to the rise of thenegative surface charge and subsequently toan increase of the repulsive forces among par-ticles. The different electrokinetic behaviour ofthe tested clays results from the fact thatKaolinite has OH- termination sites on thegibbsite face as well as on the edges, whileSmectite has OH- sites only on the edgeswhich play a negligible role in determining theelectrokinetic parameters (dominated by theconstant negative charge of faces) of Smec tites(Benna et al. 1999). Others have found that theζ-potential of Smectite, though relatively con-stant at - 30 mV between 10-5 and 10-2 M, didfluctuate up to 10 mV with changes in the

Figure 12: ζ - potential measurements for Kaolinite (left) and Na-Smectite (right)


120

concentration of indifferent electrolyte (Hunter1981). As only 5% of the negative charge ofthe 2:1 clay minerals is controlled by the pHvalue, whereas 50% or more of the surfacecharge of 1:1 minerals are controlled by the pHvalue (Das 2008), it can be stated, that the ζ-potential of swellable clays is less sensitive topH variations than that of 1:1 minerals. Re gar -ding the positive values of the ζ- potential for1M NaCl, we are not sure whether this is dueto the rapid sedimentation in the measuringchamber or due to the higher ionic strength.

4.3. Variation of the clogging of clays Several modified direct shear tests were per-formed for Kaolinite and Na-Smectite mixedwith water and 1 M NaCl to reach a consisten-cy of 0.5. The sample was gently places at ametal surface, loaded with very low pressure(15-75 kPa) for 10 minutes and then shearedwith a shear rate of 0.5 mm/min. With this testthe shear resistance of the interface betweenclay and metal may be described by theCoulomb criterion. It has two components: the

adhesion (resistance at zero normal stress anda frictional component (adhesive friction; des-cribed by the angle d). In this case, the adhesi-ve shear strength of clay on metal is less thanthe applied shear stress and also less than theinternal shear strength of the clay (Kooistra etal. 1998). Spagnoli et al. (2010c) described thestrong variation of adhesive shear strength forNa-Smectite for different pore fluids while forKaolinite no variation occurred. It is interestingthat for Na-Smectite with water as pore fluid,the shear resistance is independent of the nor-mal stress. The results (fig. 13) show for Na-Smectite a slight decrease of adhesion strengthbut a general increase in internal shear strength.From a pure mechanical point of view, clog-ging occurs during modified shear tests if:

(3)

It is theoretically possible to predict if stickingwill occur when the adhesion (a), the cohe-sion (c), the adhesive friction angle (d), theinternal friction angle (j) and the normal

Figure 13: Modified shear test on Ka -o linite (left) and Na-Smectite (right)for water and 1 M NaCl (after Spag -noli et al. 2010c)

Figure 14: Variation of the clogging for Na-Smectite mixed with water or 1 mol/l NaCl (left) and with additional application of an electrical field (right)


121

stress (sn) are known.Starting from the assumption stated by theDDL theory the clogging of Na-Smectite mixedwith water and 1 mol/l NaCl, adjusted to diffe-rent consistencies (0.55, 0.65, 0.7, and 0.85)was investigated using the cone-pull out test.Figure 14 (left) shows that the clay mixed withthe 1 mol/l NaCl fluid does not stick any moreto the cone. In fact, the increased internalshear strength of the clay due to the mixingwith a fluid with higher electrolyte concentra-tion, leads to a general drop of the stickinessof the material.

Geotechnical engineers have investigated andapplied electroosmosis to consolidate andstrengthen clayey soils and mine tailings sincethe 1940s (Casagrande 1947). Other success-ful field applications of electroosmotic consoli-dation included Bjerrum et al. (1967) and Lo etal. (1991). Davis and Poulos (1980) studied theuse of electroosmosis to aid pile driving in clay-ey soil. However, electroosmosis could also beused to reduce the clogging of the material.Therefore, experiments with the cone pull-outtest were performed. The procedure is thesame as explained before. In this case the conewas used as cathode while the proctor pot wasthe anode. A potential of 2.5 V was used in alltests by applying DC for 10 minutes. After thistime the cone was pulled out. Fig. 14 (right)shows the results.

The drop in clogging for both clays is clearlyvisible. By applying an electric charge to the

steel parts, water can be transported throughthe clay by electroosmosis to the interface bet-ween the clay and the steel. This creates a filmof water at the clay-steel interface and there-fore reduces the adherence. However, the useof DC could be a disadvantage. Extended elec-trical laboratory investigation was performedby using AC. Not only current intensity but alsowater flow depend on the electric impedanceof the electrode-soil system (ERS impedance), i.e., the current intensity depends on the soilresistivity and on the electrode-soil contact(ERC) impedance. It has been shown that forDC measurements the contact impedance isoften crucial for the system impedance where-as for high frequencies (e. g. 10 kHz) the in flu -ence of the contact impedance is of minor im -portance. Electrical impedance spectroscopy(IS) measurements were carried out (two elec-trodes) to study the properties of the electro-de-soil system (ERS) and especially the influen-ce of the contact impedance on the ERS impe-dance. IS measurements on clay samples wereperformed in the frequency range from 1 Hz to1 MHz using the IAI impedance analyzerPSM1735. Fig. 15 shows that AC signals aremore effective in reducing adherence than DCsignals due to the lower resistivity. Therefore,with AC generally lower field strengths areneeded to reduce the adherence time.

However, application of electroosmosis duringoperational tunnel heading could lead to pro-blems in the tunneling practice. By applying anegative charge to the steel surface of the

Figure 15: Resistivity of theelectrode-soil system versusfrequency obtained by ISmeasurements for Smectite


122

TBM, this steel surface is protected from corro-sion. However the anode will corrode fastercompared to an uncharged surface. The anodewill not only wear out faster; a rusted anodewill also have a lower conductivity and as aresult the amount of water transported byelectroosmosis will be lower. Furthermore, thevoltages used to generate an electroosmoticflow will electrolyze the water. Therefore hydro-gen and oxygen gas will develop due to the fol-lowing half reactions for water electrolysis:at the anode: 2H2O − 4e- ➝ O2 (g) + 4H+

at the cathode: 2H2O + 2e- ➝� H2 (g) + 2OH-

The oxygen developed at the anode will pro-bably not cause problems. However the hydro-gen gas is flammable, and if mixed with oxy-gen at the right concentration it can explode.When an electric current passes through a clayspecimen, the clay is heated and the tempera-ture will rise. As the specific resistance of clayis many times larger than that of steel, the heatdevelopment in the clay is many times largerthan in the steel electrodes. Also the influenceof the electrical current and magnetic fields oncomputers and electrical machines inside theTBM is difficult to predict.

5. Clogging test systemTo transfer and prove the results of the newstandardized adhesion laboratory test and toverify new manipulation techniques, a Clog gingTest System (CloggTS) will be developed. It isdesigned to support the geotechnical cohesicesoft ground evaluation and conditioning relatedto clogging affinity in early project phases. Thetest also aims at coming closer to some geolo-gical-geotechnical aspects and to the reality ofTBM layout with respect to material excavationand conveying processes in EPB-TBM. Somemajor demands on the test procedure are: – To construct a robust, effective and econo-

mic clogging test system adapted to multi-ple use.

– To log those operational TBM data relevantfor detecting clogging. Among those arevisual notes, advance rate, face contactforce of the cutting tool, torque of cutting

and conveying tools, cutting tool revolution(rpm), temperature, and if so, the amountof additives given into the system. It is plan-ned to vary operational data during testing.

– To avoid or control unwanted processes andother effects on operational data. Amongthese are face support managing (includingmaterial conditioning, excavation modevariation), not well documented geologicalchanges and TBM technical adaptations.Empirically, these effects may complicatethe assessment of the clogging affinity ofthe subsoil in TBM drives.

– To enable a systematic and complex investi-gation of clogging and soft ground materialconditioning. This includes the detection ofindex properties and other characteristics ofthe ground and additives. Besides classicalparameters also parameters relevant to clog-ging like ground water saturation, slakedurability, adhesion stress, ζ-potential, claymineralogy, void ratio, swelling potential orcation exchange capacity (CEC) are involved.

– To test geomaterial samples taken undistur-bed if possible. High quality samples (borecores) are preferred. By this, further geologi-cal/geotechnical ground characteristics thatmay be of importance for the adhesion/clog-ging propensity can be involved. Amongthose are low-scale geological interbeddingand fissuring (e. g. fissured clays), groundwater saturation, anisotropic ground proper-ties and mixed tunnel face conditions.

– A complex approach with well definedtesting conditions may contribute to a bet-ter understanding of the sometimes puzz-ling situation with respect to stickiness ofsoft ground materials.

– To assess the transferability of medium-scaletest results to TBM projects.

– To evaluate the efficiency and applicabilityof chemical or physical conditioning (claysurface manipulation) of soft ground mate-rials regarding technical, operational andenvironmental aspects as well as the reuseof the excavated material.

The clogging test ground samples in bore cores(diameter approx. 100mm, length approx. 0.5


123

- 1.0 m) are cut using open steel cutting tools.The opening ratio and the design of thesetools will be varied. Driving and rotation of thecutting tool assembled on a drill pipe is reali-zed by a linear rotary drive on a guidance. Thebore cores to be cut are fixed to prevent themfrom sliding and extruding. The material trans-port is realized with a screw conveying systemwhich is affixed to a drill pipe. The test systemis inclined as for the screw conveyor position inEPB drives. A simplified plan of the cloggingtesting system is shown in figure 16.

The process measuring and control units areused to set and vary testing parameters.Advance rate, face contact force and possiblyrevolution of the cutting wheel are plannedto be varied during tests. These parametersshall simulate a range of operational datatypical for TBM drives in clays, clay soils andclayey soft rocks.

For a conditioning the of material an adapteddosage unit is used. Water, electrolyte solu-tions, slurries or foam can be added to thesystem in front or behind the cutting tool usinga channel inside the drill pipe. For possibleelectrokinetic tests the cutting tool could beused as cathode. For foam conditioning anadapted foam generator can be used.

6. SummaryIn the joint research project InProTunnel theproblem of clogging during EPB tunnel drivingin fine grained soil or rock is investigated ondifferent scales.

From the analyses of project data, amongstothers based on a SQL database, some factorsof prime importance for the occurrence ofclogging problems were determined. In addi-tion to the evident parameters (clay) mineralo-gy and particle size of the excavated material,in particular the significant influence of thewater content and/or water inflow and theresulting (change of) consistency of the mate-rial could be identified.

With the cone pull-out test a practical labora-tory test to investigate different soils withrespect to their clogging behaviour could bedesigned and put into practice. Based on thenewly defined parameter »adherence« a draftfor a classification scheme for the cloggingpotential has been developed. Anyhow, thenew tests still needs to be verified by the EPBtunnelling praxis.

In the context of a comprehensive examinationof the chemomechanical coupling of clays theinfluence of different pore fluids and theirimpact on the mechanical behaviour has beeninvestigated. Shear tests and cone pull-out

Figure 16: Clogging test system (prototype)


124

tests show the effectiveness of modified elec-trolyte concentrations as well as the applica-tion of an electric charge to the steel parts withrespect to a reduction of the adhesion.

Finally, the different factors affecting the clog-ging propensity, that have been identified sofar, will be studied in a TBM-like medium-scaletest. The respective test setup is currently in theconcept state.

ReferencesAnandarajah, A. and Zhao D. (2000). Triaxialbehavior of kaolinite in different pore fluids,Jour nal of the Geotechnical Engineering Divi -sion, ASCE, 126, 2, 148-155.

Benna, M., Kbir-Ariguib, N., Magnin A. and Ber -gaya, F. (1999). Effect of pH on rheologicalpro perties of purified sodium bentonite sus -pen sions. Journal of Colloid and Interface Sci -ence, 218, 442-455

Bjerrum, L., Moum, J. and Eide, O. (1967). Ap p -lication of electroosmosis to a foundation pro-blem in a Norwegian quick clay. Geotechnique,17, 3, 214-235.

Burbaum, U. (2009). Adhäsion bindiger Bödenan Werkstoffoberflächen von Tunnelvor triebs -ma schinen. Institut für Angewandte Geowis -sen schaften. Technische Universität Darmstadt.Casagrande, L. (1949), Electroosmosis in soil.Geotechnique, 1, 3, 159-177.

Das, M.B. (2008). Advanced Soil Mechanics,Taylor and Francis

Davis, E.H. and Poulos, H.G. (1980). The reliefof negative skin friction on piles by electroos-mosis. Research report R-367, School of CivilEngineering, University of Sidney, Australia.

Dohnalová, Ž, Svoboda, L. and Šulcová, P.(2008). Characterization of kaolin dispersionusing acoustic and electroacoustic spectroscopy.Journal of Mining and Metallurgy, 44, 63-72

Dukhin, A. S. and Goetz, P. J. (2002). Ultra -sound for characterizing colloids. Particle size,zeta potential, rheology, Elsevier

Feinendegen, M., Ziegler, M., Spagnoli, G.,Fernández-Steeger, T., Stanjek, H. (2010). Anew laboratory test to evaluate the problem ofclogging in mechanical tunnel driving withEPB-shields. In: Proc. EUROCK 2010, Lausanne,Switzerland, Taylor & Francis Group, London.Gajo, A. and Maines, M. (2007). Mechanicaleffects of aqueous solutions in organic acidsand bases on a natural active clay, Géotech -nique, 57, 8, 687-699

Galassi, C., Costa, A.L. and Pozzi, P. (2001).Influence of ionic environment and pH on theelectrokinetic properties of ball clays, Clays andClay Minerals, 49, 263-269

Hunter, R. J. (1981). Zeta Potential in ColloidScience. Academic Press, London

Israelachvili, J. (1991). Intermolecular and sur-face forces, Academic Press

Jancsecz, S. (1991). Definition geotechnischerParameter für den Einsatz von Schildvortriebs -maschinen mit suspensionsgestützter Orts brust.In: STUVA-Tagung 1991, Düsseldorf. Alba-Fachverlag.

Kooistra, A., Verhoef, P.N.W., Broere, W.,Ngan-Tillard, J.M.. and van Tol, A.F. (1998).Appraisal of stickiness of natural clays fromlaboratory tests. In R. Rijkers and M. vanStaveren, (eds.), Engineering Geology andInfrastructure, 101-113

Lo, K. Y., Ho, K. S. and Inculet, I. I. (1991). Fieldtest of electroosmotic strengthening of softsensitive clay. Canadian Geotechnical Journal,28, 1, 74-83.

Mesri, G. and Olson, R.E. (1970). Shearstrength of montmorillonite. Géotechnique20(3), 261-270.


125

Mitchell, J.K. and Soga, K. (2005). Funda -mentals of Soil Behavior. John Wiley and SonsSantamarina, J.C., Klein, K.A., Palomino, A.and Guimaraes, M.S. (2002). Microscale as -pects of chemical-mechanical coupling: in -terparticle forces and fabric. chemomecha-nical coupling in clays; from nanoscale toengineering applications, Di Maio, Hueckeland Loret editors, 47-63

Spagnoli, G., Fernández-Steeger, T., Feinen de -gen, M., Stanjek, H. and Azzam R. (2010a).The influence of the dielectric constant andelectrolyte concentration of the pore fluids onthe undrained shear strength of smectite. Soilsand Foundations, 50, 5 (accepted)

Spagnoli, G., Feinendegen, M., Neher H., Fer -nán dez-Steeger, T. and Azzam R. (2010b). Ver -kle bungsproblematik zwischen Tonen undTBM. Veröffentlichungen des Grundbau insti tu -tes der Technischen Universität Berlin, 7thOctober 2010, 50 (accepted)

Spagnoli, G., Fernández-Steeger, T., Azzam, R.,Feinendegen, M. and Stanjek, H. (2010c). Theinfluence of different pore fluids compositionson the shearing behaviour of clays. 11th Con -gress of the International Association forEngineering Geology and the Environment,IAEG, 5-10 September 2010, Auckland, NewZealand (ac-cepted)

Sridharan, A. and Venkatappa, Rao G. (1975).Mechanism controlling the liquid limit of clays,Proc. Istanbul Con-ference on Soil Mechanicsand Foundation Engineering, 1, 75-87.

Sridharan, A., Rao, S. M. and Murthy, N. S.(1988). Liquid limit of kaolinitic soils, Geo -technique, 38, 2 191-198

Thewes, M. (1999). Adhäsion von Tonbödenbeim Tunnelvortrieb mit Flüssigkeitsschilden.Institut für Bodenmechanik und Grundbau, Nr.21. Gesamthochschule Wuppertal.

Weh, M., Ziegler, M., Zwick, O. (2009a). Ver -klebungen bei EPB-Vortrieben in wechselndemBaugrund: Eintritts-bedingungen und Gegen -maßnahmen. In: STUVA-Tagung 2009, Ham -burg. Gütersloh: Bauverlag.

Weh, M., Zwick, O., Ziegler, M. (2009b). Ma -schi nenvortrieb in verklebungsanfälligem Bau -grund, Teil 1: Tunnel 28(1), 25-36 und Teil 2:Tunnel 28(2), 18-28.

Wilms, J. (1995). Zum Einfluß der Eigen schaf -ten des Stützmediums auf das Verschleiß ver -halten eines Erddruckschildes. Fachgebiet Bau -betrieb und Bauwirtschaft, Nr. 12. Universität-Gesamthochschule Essen.

Zimnik, A.R., van Baalen, L.R., Verhoef, P.N.W.,Ngan-Tillgard, D.J.M. (2000). The adherence ofclay to steel surfaces. In: GeoEng 2000: AnInternational Conference on Geotechnical andGeological Engineering. Melbourne, Aus tra lia.,Lancaster, Technomic Publ. Basel.


126

SIMSAN - Simulation-supported developmentof process-stable raw material componentsand suspensions for the production of ceramicsanitary ware on the basis of modified mineralsurfaces

AbstractThe article focuses on the status of the rese-arch after the second project year. In additionto the basic characterisation of the chosenreference material mixture for sanitary slipcasting, as presented in previous publicationsfor the Geotechnologien Science Reports, it fo -cuses especially on the characterisation resultsof the basic raw material components, whichin the second year of the project were investi-gated. These are the basis for the funtionalcharacterisation of novel body concepts andreference systems for the experimental valida-tion of the simulation model. The developmentof first body concepts, derived from the mine-ralogical characterisation of the raw materials,is supported using a principal FMEA-based as -sesment of influences on the functional beha-viour of the body as casting slip. Parameteranalysis and statistical design of experimenttechniques are being implemented to experi-

mentally evaluate basic mineralogical influen-ces within the context of a slip compositionincluding differerent additive contributions.Significant interactions have to be specified andand to parameters effects ranked for furtherdevelopment. These insights will be the basisfor laboratory pilot scale tests, which will in thenext project phase focus on identifying andvalidating processing parameter influences.

The characterisation results have furthermorebeen used to develop simplified reference sys -tems for the experimental validation of thesimulation. After a short standstill in the mo -dell development, caused by the relocation ofthe simulation expertise provided by CME fromRWTH Aachen to the Materials and ProcessSimulation Group in Bayreuth, the results arebeing transferred into a binary, up to ternarymodel approach, based upon a simplified com-position and particle system. The simplifica-

Engels M. (1)*, Agné T. (2), Bayard E. (2), Diedel R. (1), Emmerich H. (6), Emmerich K. (3), Haas S. (5),

Latief O. (4), Peuker M. (1), Steudel A. (3), Vuin A. (5), Yang H. (6)

(1) Forschungsinstitut für Anorganische Werkstoffe – Glas/Keramik – GmbH (FGK), Höhr-Grenzhausen

(2) Villeroy & Boch AG, Mettlach/Saar

(3) Karlsruhe Institute of Technology (KIT), Centre of Competence for Material Moisture (CMM), Eggenstein-

Leopoldshafen

(4) Stephan Schmidt KG (SSKG), Dornburg

(5) Zschimmer & Schwarz GmbH & Co. KG (Z&S), Lahnstein

(6) Materials and Process Simulation (MPS), Fakultät für Angewandte Naturwissenschaften, Universität Bayreuth**,

Bayreuth

* Coordinator of the Project

** Former allocated at the CME Center for Computational Engineering Science of RWTH Aachen


127

tion, performed in varying degrees, is based onthe translation of the identified mineralogy ofthe components into particle systems with par-ticle size and ratio aspects, charge specifica-tions of identified surfaces and surface poten-tial effects in different settings and environ-ments. As the experimental validation is ambi-guous due to the limitations of the experimen-tal techniques for the basic systems, recom-mended literature as well as first approachvalidation experiments are used to generatebasic input. The status of the development ofthe hybrid model algorithm, first implementa-tion of parameter settings so far, resultinginput requirements and the next steps to betaken are discussed. In this regard the impor-tance of the working procedure for rheologi-cal slip characterisation, developed within theproject for functional characterisation of thesystems is highlighted.

IntroductionThis article highlights the status of the interdi-sciplinary project, in which the impact of che-mical and physical characteristics of the mine-ralogical surfaces of the mineral raw materialson the process will be exemplary clarified forhighly automated sanitary ceramic pressure-casting systems. The primary goal of this pro-ject is to enable the raw material suppliers andthe ceramics industry to develop adapted sta-ble raw material mixtures and casting slurriesand establishing stable and robust processesby way of quantitative predictions from simu-lation calculations. These are validated by sta-tistical experiments including the effectivenessof additives and process-defining parameters.It provides a novel approach for extracting andprocessing technology of the raw materials,engineering of performance characteristicsduring the production by additive modificationof the surfaces, and adequate process controlduring pressure-casting in the ceramic produc-tion. Due to the high complexity of the subject,the project relies on a strong interaction bet-ween the different partners from the proces-sing chain, represented by the partner fromsanitary ware industry, a raw materials supplier,

a supplier of additives for the ceramic industry,expertise in the field of colloid chemistry andanalytical characterisation and modelling com-petence. This cooperation already resulted inthe specification of a working procedure for afor unambiguous capable and transferablerheological slip characterisation, useable forthe whole process chain, which has alreadybeen presented to the relevant ceramic indu-stry at the 2010 annual Meeting of the Ger -man Ceramic Society [1].

Experimental results The fact that common techniques used to con-trol the raw materials fail to detect differencesin raw materials characteristics which correlateto known process variations requires the inve-stigation of important intrinsic material charac-teristics as well as their influences on the rheo-logy and the process ability. The compositionof the joint project group offers a unique basisto perform the conventional ceramic productand process characterisation as well as measu-rement of colloidal parameters (rheology, ca -tion exchange capacity zeta potential, flowpotential), state of the art mineralogical analy-sis and submicron particle size measurement,which cannot be achieved by regular produc-tion laboratory equipment, but constitute theessential basis for the input in the numeric cal-culations as well as process and product deve-lopment. The results presented here build up onthe results previously presented in precedingGEOTECHNOLGIEN Science Reports No. 12.

Characterisation of the aplied raw ma -terials Based upon the knowledge of sanitary pressu-re casting bodies present in the consortium, aspecified mixture of plastic clay components,feldspars, kaolin, quartz as well as alternativekaolin and clay components for a production-relevant casting slip have been specified, usingthe relevant bandwidth of mineralogical andparticle size composition in regard to functio-nal characteristics. In the preceding publica-tions the initial characterisation of the body


128

composition as sum of the components bychemical analysis, particle size measurementusing sieve analysis, sedimentation and laserlight scattering, specific surface measurementusing BET, as well as thermal analysis has beenpresented. Zeta potential measurements at dif-ferent pH settings of the system, supported bythe measurement of eluted cation and anioniccomponents using ICP, have been performed.This composition has been used as referencebasis for the development of a sodium silicate-based deflocculant system by Z&S, aimed atoptimal deflocculation and stabilisation of theslip. In the following the results of the separa-te raw materials characterisation, essential asbasis for the development of novel body for-mulations, additional deflocculant systems andreference systems for experimental validationof the numeric calculations are presented.

As described in the prececding articles patriclesize analysis plays a significant role in the defi-nition of the specification of the particle modelfor simulation. The results from the chemical,mineralogical and colloidal charactersiation ofthe different patricle size fractions of the bodycomposition (0.63 µm, 0.63 – 2 µm, 2 – 6.3µm, and 6.3 – 63 µm), which had to be obtai-ned using a special fractioning technique con-ceived by KIT/CMM, due to the high amountof the fine fraction of the clay mixture as spe-cified by sedigraph analysis (50 % < 0.63 µm),were used for first assessment of the particlecharge distribution for the clay fraction.Particle size and morphology data, eminent forthe description of rheological characteristicsand microstructural development of the bodyduring slip casting formulation, have beenderived from a comparison and evaluation ofthe available particle size analysis techniques.For the submicron range alternative measure-ments to sedimentation and laser light scatte-ring like acoustic spectrometry, dynamic lightscattering techniques as well as new laser scat-tering equipment using different light beamsettings (and using Mie theory based evalua-tion) for edge diffraction effects have beenimplemented by the project partners.To gene-rate a fist assessment of the morphology and

an »actual« size distribution of the fractions,samples were freeze-dried for electron micro-scopy image analysis. Most important conclu-sion from this previous investigation is that inthe chosen composition there is a noticeabletransition at the 2 µm fraction: the lower frac-tions, mainly below 0.63 µm, consist of main-ly plate-like particles with a general averagelength about 300 nm and a stack thickness of50 – 70 nm (with aspect ratio 1:6.8) with aclay-based chemical composition, whereasabove 2 micron a grain type size with high SiO2 contents, indicating quartz and feldsparto be main components, prevails. This impliesa different validity of different measurementtechniques for different fractions, the newlaser scattering techniques and acoustic spec-trometry providing also a validated possibilityfor the measurement of the fraction below 2 µm, as this fraction, in this case mainly below0,63 µm, cannot adequatly be measured byregular light scattering and sedigraph Theseresults have been presented at the annualmeeting of the German Ceramic Society tosupport further implementation for the essen-tial focus on the submicron fraction character-isation in clay minerals [2]. Regarding the rawmaterials characterisation as presented herelaser light scattering results have been used forthe general classification of the particle sizedistribution of the raw materials, whereas forthe specification of the particles for the frac-tions reference is made to the previous resultsof the fractioned composition.

In Table 1 the characterisation results of theclay raw materials, including chemical analysisby X-ray fluorescence, mineralogical phaseanalysis by X-Ray diffraction using the Auto -quan software for Rietveld analysis [3], CECmeasurements and the amount of solublesalts, also a strongly influence the behaviour ofthe raw material, due to surface interactionsand influences in the ionic concentration, aresummarized. For a simplified description of thesurface and edge charge and the charge distri-bution of the particles the previously imple-mented estimation, based upon pH dependentCEC measurement after Dohrmann [4], is used.


129

Table 1. Characterisation results for the raw materials components.


130

Regarding the possible meta-stable characterof the material within the pH range investiga-ted (pH 4 to pH 8) a comparison with referen-ce data from literature [5] is being implemen-ted as orientation for the validation experi-ments and the numerical calculations. For theestimation of surface potentials, necessary forthe simulation approach, the influence of pH,ionic strength and volume concentration forselected clays is being investigated and as faras possible experimentally validated. Firstresults are presented in the chapter on thereference systems for the experimental valida-tion. To investigate the possible influence ofdelamination effects within the pH range thepH dependence of particle size measurementof the fraction < 0,63 µm was investigatedusing acoustic spectrometry (QuantachromeDT1200 Zeta Potential, Table 2). Within thepH range used no significant effects could bedetrmined, lower and higher pH values howe-ver indicate larger deviations, due to the mea-surement limitations (see fit error, Table 2).

Development of reference systems forexperimental validation of the simulationAs a basic reference system for the validationexperiments different approaches have beeninvestigated. A first approach, based upon usinga well-defined Ca-bentonite with a high mont-morrillonite content (up to 96 %, d50 around 3µm) to provide a defined amount of swellablematerial in a basis reference system, had to beabandoned due to the strong gelation of thematerial, impeding initial characterisation aswell as experimental implementation.

Based upon the characterisation results of theclay raw material and the kaolin (Table 1), feld-spar and quartz components the choice wasmade to convert to the specified clay rawmaterials and their mixtures with specific diffe-rences in the amount kaolinite, quartz andamount of swellable material. For a first appro-ach the clays D (37 % kaolinite - 7 % quartz -18 % swellable material - total clay content 86 %), E (56 % kaolinite, the content of quartz,

Table 2. pH dependence of the particle size measurement using acoustic spectrometry for the frac-tion < 0.63 µm of the mixture.

Figure 1: pH dependent zeta potential measurement for Clay D and H.


131

swellable material and total clay content beingcomparable to clay D) and H (21% kaolinite -54 % quartz, 8 % swellable material and totalclay content comparable of 42 %) have beenchosen. Clay D and E should indicate how theamount of kaolinite and other clay compo-nents (illite, muscovite) could be treated in themodelling; clay H introduces the dilutioneffetct by the quartz influence.

For each these clays corresponding defloccula-tion curves have been measured by Z&S usingthe developed additive systems. For laterimplementation first assessments of the ionicconcentration due to the additive dissociation,influencing the electrolyte environment of theparticle system, have been established. In thisregard the possible approach of the descrip-tion of the permittivity, which in the context ofthe project cannot be unambiguously defined,

is under investigation [6]. For the estimation ofthe surface potentials, necessary for the simu-lation approach, the influence of pH, ionicstrength and volume concentration for theseselected clays is being investigated. In figure 1the difference between the zeta potential ofclay D (high amount of swellable metarial) andH (high amount of quartz) is displayed.Measurement limitations however are reachedmeasuring the clay materials without the useof additives at higher volume fractions and atthe specified pH values.

Implementation of different ionic concentra-tions by using KCl did not generate reliableresults by exeedingly high zeta potential rea-dings. Especially the determination of a reliableparticle size by acoustic spectrometry provesto be difficult in these settings. Further litera-ture research regarding measurement settings

Figure 2: Rheological characterisation ofClay D and H at different pH values andionic concentrations.


132

and validation of the results is therefore beingperformed. For the experimental validation ofthe functional behaviour of the systems underthese settings rotational viscosimetry measu-rements are being performed. Figure 2 pre-sents first results. Clay D seems to stabilise ata higher viscosity at pH increase. Clay Hworks in an opposite direction, however at ahigher viscosity level than clay D. Mixturesequences are being experimentally proces-sed and will be adapted to the progress inmodel development.

Development of body compositions forlaboratory scale pilot testingA first investigation into the effect of the pla-stic clay components in the body formulationby standard characterisation of the rheologicaland physical characteristics (like cast forma-tion, dry bending strength, deflection and dryshrinkage) of a mixture sequence by succesivereplacement of the plastic content by kaolinconfirmed the significant influence of the pla-stic components, especially by significant shiftsin rheology (figure 3.). As the exchange of kao-lin might not be a feasible route for productionslip compositions, alternative systems have tobe investigated.

Figure 3: Body development andcharacterisation by succesive repla-cement of the plastic clay fractionby kaolin content.


133

Based upon these results a second approachwas conceived, based on modification of theamount of swellable material within the plasticfraction by clay exchange. For this reason theclays D and E, identified as main contributantsof swellable minerals, have succesively beenreplaced by the other clay components. Figure4 specifies these mixture compositions andidentifies quantitatively the significant influen-ce of the swellable metarials. From this mixtu-re range the composition without clay D and E(M1) is chosen as first basis for the statisticalexperiments to identify the key parametersand their interactions in comparison to thereference system. In figure 5 the significant dif-ferences in rheological behaviour of both mix-tures at different practicable specific densities,as derived from the developped rheology mea-surement routine, are presented. Especially atlower specific densities a significant viscosityincrease in the case of the mixture M1 -without swellable material- is detected, confir-ming the necessity of these components foradequate pressure casting slips. The next stepis to quantify these influences within the con-text of production slip compositions, to betested on a laboratory scale.

Statistical Design of Experiments fort he de - velopment of laboratory scale casting slipTo support de development of the laboratory

scale casting slips, a screening of the relevantinfluencing factors, the magnitude and signifi-cance of their influence and their interactionsis of importance, especially in regard to theproject goal of developing and maintainingrobust processes based upon improvement ofrelevant material and definition of processparameters with adequate control possibilities.In this regard the use of statistical design ofexperiment techniques [7], as implemented byV&B to support the development of the rheo-logy measurement routine, is perceived as anecessary mean of screening these interactionswithin a manageable effort.

For this purpose the previously performedFMEA-based parameter analysis was used tospecify dependent and independent variablesor factors, background variables to be conside-red and levels settings for the relevant factors.As target value a stable rheological behaviourin time is specified, measured by the key datafrom the rheology measurement routine deve-loped within the project. This routine, inclu-ding a defined sample preparation, has provento provide an essential basis for measurementcomparabilty within the consortium. For thispurpose the jump test routine with succesivechanges in shear rates from 0.5 s-1 to 100 s-1

and back is chosen for rheology measurementsat specified time intervals. The proven compa-rability renders a unique possibility to split the

Figure 4: Body development by suc-cesive replacement of the clay frac-tion with swellable material contentand initial functional characterisation


134

efforts for the designed experiments within theconsortium, thus providing the possibility usethe available capacity to expand the amount ofmeasurements and so enhance the reliability ofinterpretation of the results, inherent to stati-stical experimental design.

From the FMEA analysis of parameters influen-cing the rheology, a total of 13 influencing fac-tors was defined, from which 5 were identifiedas main factors to be adressed for the identifi-cation of the interaction of the mineralogicalsurfaces with additives on rheology on a labo-ratory scale (see table 3). The factor mineralo-gy (specified by the measured mineralogicalphases including the amount of swellablematerials) is introduced by the forementionedbody compositions. As main factor also thekaolin quality is specified, based upon thepractical experience that different amounts ofsoluble salts, inherent to the kaolin quality,may influence the rheology and its stability.Therefore an alternative kaolin quality is intro-duced (see »kaolin B« in table 1).

For the specification of the additive combina-tion as well as the specification of the levels,like the amount added, the developed additi-ves, also including polyacrylates and phospho-nates, which addtionaly to the electrostaticfunctionality implement sterical stabilisation,were pre-tested using the chosen body com-positions. In this case for each additive type the

qualitative levels for the dosage are quantitati-vely specified for experimental implementa-tion. Regarding the use of filtration aids at thisstage 2 alternatives are defined. Specific densi-ty of the slip as well as temperature and waterquality have been defined as background vari-ables, here to be controlled but kept constantwithin specified tolerances. The other factorsderived from the assessment, mainly proces-sing parameters derived from production prac-tice, adressing recycling of casting slip, ageingin production and upscaling influences onmixing and slip transport, will be adressed insubsequent test plans.

For the screening of the identified main para-meters at first a so called 25-1 screening plan isbeing implemented, developed using the stati-stical software »statistica«, already sucessfullyimplemented at V&B. This plan is builds upona base design of 5 experimental factors witheach 2 levels, leading to 16 combinations to betested, which are randomized and divided overthe participating project partners. Each partnerwill prepare and measure one specified addi-tional combination in duplo from his selectionfor reproducability and repeatability control, aswell as 1 combination, specified to be measu-red by each of the partners separately to vali-date the anticipated comparabilty of results.The experimental design takes into accountthat 2 factor interactions can be adressed, but3 factor interactions are neglected in this

Figure 5: Rotational visco-simetry at shear rate 100s-1 for the reference andthe M1 mixture at diffe-rent specific densities ofthe slip (LG = specific den-sity in g/L)


135

stage. Furthermore mixed interactions ofdependent variables might not be detected,which is inherent to the statistical experimentaldesign. To increase the resolution of the resultsthis plan can be »folded« to 32 combinations,adding a second »block« of measurements toadress possible differences by repetition. As anext step a plan based upon 6 factors, with anincreased number of additives and amounts ofadditive by a 3 level approach (adressing thenon-linear deflocculation behaviour more reali-stically) has been designed, leading to a totalof 72 combinations to be evaluated by expan-sion of the original set of experimental data.

The results as generated by rheological analy-sis will be evaluated within the tolerances asspecified by the industrial partners. Basedupon these results, which are supported byadditional measurements by the partners(additional oscillation measurements, conduc-tivity, pH development, casting performance)additional influences and their interactions tocreate optimal and stable casting performancewill be specified to be included in followingexperiments. The results of the experimentalvalidation will be presented and discussed infollowing publications.

ModellingThe performed model development so far hasbeen based upon the simulation of a colloidalsystem, consisting of spherical particles in anelectrolyte, motivated by previous studies [8, 9]on aluminum oxide particles, and showingrepoducible results. To exemplify the capabili-ties of this approach the phase diagram wasampled as defined by two parameters: pHvalue and ionic strength. Varying these para-meters enabled the simulation of the transitionfrom a stable suspension to a state of agglo-meration, by increasing ionic strength and pHvalue, as presented in the previous publica-tions. Based upon the results the single com-ponent spheric particle model has been exten-ded to a binary version, where the two com-ponents can have different material nature,sizes, densities etc. The algorithm now simula-tes two main components, derived from theraw material characterisation: small clay parti-cles below 2 µm to be further specified andlarge particles (> 5 µm), specified as quartzpatricles. The description of the particles regar-ding charge distribution and surface potentialis, as mentioned before, based on assumptionsfrom experimental validation where possible,otherwise from reference values from literatu-re [5, 10].The simplified binary model alreadyallows a study of the influence of the chemi-cal/minerlogical nature and the size of colloidal

Table 3: The statistical experimental setup for the evaluation from basic mineralogical interactionsto processing influences on laboratory scale in the development of body composition concepts


136

particles on the rheological properties of theinvestigated system, being an essential para-meter for the production process. At the leastbasic trends in behaviour will be derived fromthe simplified description, guiding furtherexplorations of more complex and realistic mo -dels (like the transition to a ternary system).

For the simulations, a combination of two dif-ferent techniques has been used: MolecularDynamics (MD) [11] and Stochastic RotationalDynamics (SRD) [12] (also known as Multi par -ticle Collision Dynamics). These techniques al -low various thermodynamic and rheologicalproperties of a system, with the use of statisti-cal mechanics, starting with the microscopicand mesoscopic components, which in thiscase are the fluid and the colloidal particlesrespectively to be modelled. With the MDtechnique, the following interactions betweencolloidal particles are included in the simula-tion: screened Coulomb repulsion, van derWaals attraction, contact repulsion and short-range lubrication forces. The SRD, on the otherhand, is used in combination with MD in ahybrid manner, in order to include the long-range hydrodynamic interactions between col-loids and thermal fluctuations due to randomcollision between fluid particles and colloids.

To incorporate the dependence of the stabilityof colloidal systems on pH value now the so-called charge regulation model [14] is imple-mented in the program. It manages to modelthe adsorption-desorption reaction of electro-lyte ions on the surface of colloidal particles.An expression of the effective surface potenti-al of colloidal particles is worked out as thefunction of the pH value and the ionic strengthof electrolytes. Experimental data therefore areneeded to calibrate the model to adjust itsparameters to the used material.

To show the tendency of the behavior of thebinary component model including this regula-tion model a small system cell of the size 8.88µm by 4.44 µm by 4.44 µm is modelled. Thevolume fraction of colloidal particles is 40% (asspecified from industrial partice) and the tem-perature is 303K. 1182 Alumina particles ofthe radius 0.185 µm and 591 quartz particlesof the radius 0.25 µm are considered. The col-loidal dispersions are now sheared with therate 100 s-1, comparable to the testing proce-dure implemented for the clay characterisa-tion. Ionic strength and pH value have beenvaried to investigate the stability of the colloi-dal dispersions and transitions between diffe-rent states. In Figure 6 snapshots of three sta-tes with pH value 6.5 and different ionic

Table 4: The 25-1 screening plan for the specification of the mineralogical composition effectsand the interaction of the mineralogical surfaces with additives on rheology


137

strengths are presented. At a high ionic strengthcorresponding to salt concentration of 1*10-2

mol per liter, the repulsive coulomb interactionis strongly screened. The colloidal particlesagglomerate to form clusters due to the domi-nation of the van der Waals attraction, whichunder shear attach themselves to one of theboundary planes and change shape with timedue to the motion of the surrounding fluids.With decreasing ionic strength to 1*10-3 molper liter a stable suspension is ob served in resultfrom the balance of the screened Coulombrepulsion and the van der Waals attraction, witha nearly linear velocity profile of colloidal parti-cles (see Fig. 7), like a New to nian fluid.

Further decrease of ionic strength to 1*10-4

mol per liter leads to a repulsive state wherethe colloidal particles rearrange their configu-rations under shear. Except in the boundarylayers colloidal particles can move in bothdirections irrespective the velocity of the sur-rounding fluid. The particle velocity profiles inthe direction perpendicular to the shear flowas presented in Fig. 7 show the different natu-re of these states. The variation of the shearviscosity with the ionic strength as shown inthe left diagram in Fig. 7 can be used to furt-her explore the difference between these sta-tes. As can be seen from the plot, the shearviscosity of the repulsive state is much higherthan for the stable suspension state and thecluster state. The strong increase of the visco-

sity would originate from the increase of theeffective particle size in the repulsive state.Similar transitions are observed as varying thepH value from 8, 10 to 12 with a fixed ionicstrength 1*10-4 mol per liter. It has to be kept inmind that the parameters used in above simula-tions are up to now simple modifications of thesingle component system [8, 9] to include thesecond component of quartz particles, thesystem kaolinite/quartz at the moment beingregarded as more suitable for the modelling.

The difficulty in the definition of the particlefor the binary and ternary system originatesfrom the description of the surface types,which can be allocated due to the mineralogi-cal characterisation:

Kaolinite as 1:1 layer silicate contains threetypes of surfaces: siloxanplanes (Q4), gibbsiticplanes (Al-OH bonds) and edges (brokenbonds from silanol and aluminol groups). Hereno permanent surface charge can be determi-ned and edge charges are variable by adsorp-tion and desorption of protons at the edges.

Illite/Muscovite as 2:1 layer silicates containstwo siloxanplanes and edges. In this case theparticles are specified with high permanentnegative surface charge (0.26 to 0.34 Cm-2)due to substitution in the 2:1 layers. Strongspecific sorption of counter ions (K+) mightcause much lower zeta potentials compared to

Figure 6: Snapshots of a clusterstate (left), a stable colloidal sus -pension (middle) and a repulsivestate (right) with the pH value 6.5and the ionic strength 1*10-2,1*10-3 and 1*10-4 mol per literrespectively. Lighter color indicateshigher speed than darker color.

Figure 7: Velocity profi-les of colloidal particlesin direction perpendicu-lar to the shear flow forthe states shown in Fig.6.(left) and shear viscosityversus the ionic strengthwith the pH value 6.5.(right).


138

swellable clay minerals like smectites in gene-ral or montmorillonite in special. Edge chargesvary by adsorption and desorption of protonsat the edges.

Smectites like montmorillonite as 2:1 layer sili-cate also contain siloxanplanes and edges,showing intermediate permanent negative sur-face charges (0.10 to 0.14 Cm-2) due to sub-stitution in the 2:1 layers, subject to specificsorption of large mono valent (e.g. Cs+, whichare not common ceramic clays) or »commonsized« tri and quadro valent cations (Fe3+, Al3+,Si4+) due to oxidation and/or dissolution ofsoluble salts in the clay material, with variableedge charges by adsorption and desorption ofprotons at the edges.

On this basis the following assumptions arediscussed for modelling: the siloxan surface ofkaolinite can be treated similar to quartz, theedge charge being estimated as most impor-tant. Quartz and feldspars could be regardedas similar for modeling. For the approximationof a binary system the raw clay materials selec-ted could be treated simply as a binary mixtu-re of quartz and kaolinite resulting in a massratio 90:10 for Clay D and 60:40 for Clay H.For the transition to a ternary system the edgezeta-potential of the different silicate systemscan be modeled as the linear combination ofthose of quartz and alumina. The ternarymodelling will be based upon kaolonite, quartzand swellable minerals, using different edgezeta-potential settings. The effect of the solu-ble salt ions (like Al3+) ands the liquefier inter-actions can be included later on in the model-ling by the charge regulation model to modelthe adsorption-desorption reaction of electro-lyte ions on the surface of colloidal particles,influencing the zeta potential. For the firststeps these factors are neglected. To derivenecessary input for the effective surface poten-tial and the isolelectric point of the used mate-rial the experimantal results and reference data[10] are investigated, the literature howeverdelivering distinctly different values within apartially unspecified context for the mineralo-gical components.

An important task in the next step towards thesimulation of clays is the introduction of plate-like particles into the system. The main chal-lenge to be overcome for this task results fromthe fact that –in contrast to the case of spheri-cal particles, where symmetry allows great sim-plification of the problem– the theoretical andcomputational framework needed to obtainthe inter-particle interactions are not yet fullydeveloped [15]. As mentioned in prevoiuspublications two different, likely complementa-ry, approaches have been defined. The firstoption, »glueing« of point-like particles (orspheres) to a plate-like structure [15, 16], obtai-ning the interaction between platelets byadding the interaction between the correspon-ding point-like particles, can only be achievedat relatively high computational cost, arisingfrom the many interactions between pointlikeparticles that need to be taken into account inorder to obtain the total force and torque oneach platelet at each step of the simulation.The second approach is to obtain the solutionto the Poisson-Boltzmann equation of a discimmersed in an electrolyte [17]. This methodhas the advantage of computational economy,but must be treated carefully regarding theconstraints on the charge distribution on theplatelet. In the spirit of the charge renormali-zation theory it is expected that suitable para-meterizations of the model would extend thevalidity of the used approximations beyondtheir original theoretical constraints. In thenext project period the plate-like clay particleswill be implemented in the program by usingthis approach.

Outlook The research during the next project phase willconcentrate on development of the numericalsimulation of binary and ternary systems andthe development of the numerical computa-tion of platelike particles. The developmentwill be based upon characterisation and expe-rimental validation using the selected claymaterials and their mixtures, as well as litera-ture evaluation regarding reference data onsurface potential dependency on pH, ionic


139

concentration and volume fraction. The stati-stically designed experimentation will be evalu-ated and expanded to provide the basis fornew material mixture concepts to be tested ona laboratory pilot scale, based upon the deve-loped rheological evaluation method. From theevaluation of the interactions of the mineralo-gical composition with the developed additivesnew additive concepts will be optimized anddeveloped.

References[1] Prüfmittelfähigkeit der rheologischenCharakterisierung keramischer Schlicker mitHilfe moderner Rheometer, Engels M. (2010),presentation at the Annual Meeting of theGerman Ceramic Society 2010, Hermsdorf/ -Thüringen, 23.03.2010

[2] Vergleich der Korngrößenanalyse – Sedi -graph, Laserbeugung, Zetasizer und REM beiTonmineralen, Latief O. (2010), presentation atthe Annual Meeting of the German Ceramic So -ciety 2010, Hermsdorf/Thüringen, 23.03.2010

[3] Quantitative phase analysis using theRietveld method and a fundamental parameterapproach, Kleeberg R., Bergmann J. (2002),Proceedings of the II International School onPowder Diffraction, pp. 63-76.

[4] Kationenaustauschkapazität von Tonen - Be -wertung bisheriger Analysenverfahren und Vor -stellung einer neuen und exakten Silber-Thio -harnstoff-Methode, Dohrmann R. (11997), Diss.RWTH Aachen, AGB-Verlag Nr. 26, 234 S. ISBN:3-86073-605-1

[5] Surface Charging and Points of zeroCharge, Kosmulski, M. (2009), Surfactant Sci -ence series Volume 145,CRC Press

[6] Characterization of water in bacterial cellu-lose using dielectric spectroscopy and electronmicroscopy, Gelin K et al. (2007), Polymer 48,pp. 7623 – 7631

[7] Taschenbuch Versuchsplanung, KleppmannW. (2003), Carl Hanser Verlag Wien, ISBN: 3-446-22319-3

[8] Simulation of claylike colloids, Hecht M. etal. (2005), Phys. Rev. E 72, 011408.

[9] Shear viscosity of claylike colloids in com-puter simulations and experiments. Hecht etal. (2006), Phys. Rev. E 74 021403

[10] On the zeta potential and surface chargeofmontmorillonite in aqueous electrolyte solu-tions, Delgado A. et al. (1986), J. of Colloidand Interface Science Vol 113 p.,203-211.

[11] Computer Simulations of Liquids, Allen M.P. and Tildesley D. J. (1987) Oxford SciencePublications, ISBN: 0 19 855 645 4 (Pbk)

[12] Mesoscopic model for solvent dynamics,Malevanets A. and Kapral R. (1999), J. Chem.Phys. 110, 8605

[13] Solute molecular dynamics in a mesoscalesolvent, Malevanets A. and Kapral R., (2000), J.Chem. Phys. 112, 7260

[14] Electrical double layer interactions underregulation by surface ionization equilibria---dissimilar amphoteric surfaces,. Chan D.Y.C etal. (1976), J. Chem. Soc. Furuduy Trans. 1, 72,2844

[15] Interaction of Nanometric Clay Platelets,Jonsson B. et al. (2008) Langmuir, 24 (20).

[16] Aging in a Laponite colloidal suspension:A Brownian dynamics simulation study, MossaS. et al. (2007), J. Chem. Phys. 126, 014905

[17] Electrostatic Properties of Membranes:The Poisson-Boltzmann Theory, Andelman D.(1995), Handbook of Biological Physics.Volume 1. Elsevier Science.


140

Control mechanisms of clays and their specificsurface area in growing media – assessmentof clay properties and their parametrization forthe optimization of plant quality

SummaryThe addition of clay to growing media aims atconstant supply of potassium, phosphorus andmicro nutrients, pH-buffering, improvement ofrewettability, and cohesion of growing media.The identification of suitable clays and theirclassification is a prerequisite for product andcultivation safety. A range of different clays hig -hly variable in their mineral parameters wereselected for experiments on nutrient buffering(P) and Mn toxicity, their ability to improve therewettability and binding capacity of growingmedia. From batch experiments and growthtrials it was derived that a threshold value forthe sum of exchangeable and easy reducibleMn in clays for growing media is not justified,as even very high Mn contents in clay were notphytotoxic. The P binding capacity of clays wasstrongly correlated with the oxalate extractableFe and Al content. A newly developed capilla-ry rise method (WOK) was used to characteri-ze the rewettability of growing media. Thespeed of rewetting mainly depends on thefineness of the amended clay. Surface freeenergy (SFE) data of the growing media indica-

te that those with a good rewettability showalso high values for SFE. Compared to the ka o -linitic and illitic clay amendments, bentonitesshow no significant increase in the SFE. Sur -faces of clay minerals exhibiting originally polarand hydrophilic surfaces, can render hydro-phobic when coated with weakly or non-polarorganic matter moieties. Dissolved organic car-bon (DOC) sorption was found to be positivelycorrelated with the specific surface area (SSA),cation exchange capacity (CEC) and amount ofdithionite extractable Al and Fe. Clays contai-ning expandable clay minerals with high CECand SSA (e.g. smectites) and those rich in Al-and Fe-oxides seem to be less effective forimproving rewettability, whereas an additionof non-expandable clays with lower SSA, CEC(e. g. kaolinitic and illitic clays) and amorphousoxide content appears more promising. Newinsights on the adsorption of DOM on claymineral`s surfaces will be obtained in the thirdyear by chemical analysis of the surface with X-ray photoelectron spectroscopy (XPS), surfacetopography analysis and contact angle measu-rements. A new method for the determination

Dultz S. (1)*, Schellhorn M. (2), Schmilewski G. (3), Schenk M.K. (4), Dombrowski, I., (4), Schmidt, E. (2),

Walsch, J. (1), Below, M. (1)

(1) Institute of Soil Science, Leibniz University Hanover, e-mail: [email protected],

[email protected]

(2) Stephan Schmidt KG, Langendernbach, e-mail: [email protected],

[email protected]

(3) Klasmann-Deilmann GmbH, Geeste-Groß Hesepe, e-mail: [email protected]

(4) Institute of Plant Nutrition, Leibniz University Hanover, e-mail: [email protected],

[email protected]



141

of the binding capacity of clays in blockingmedia is currently being tested and will bestudied together with determinations on K-dynamics.

Introduction

Concept on function und effect of clays ingrowing media related to practiceA suitable rooting medium is fundamental forplant growth. Horticultural crops have certainrequirements which the grower needs to fulfilwith the help of individually tailored growingtechniques and cultivation measures. For theselection of growing media constituents eachshould possess optimum characteristics for thespecific culture. Modern horticulture withcomputer-controlled irrigation and fertilisationprogrammes, potting machines, pricking ro -bots, climate-controlled greenhouses and just- in-time production requires dependable, quali-ty-assured growing media. Specialist compa-nies rely on ready-made growing media whichare either part of the manufacturer's standardrange or special mixtures produced at the gro-wer's request. For the development of formu-lations and the production of growing mediasuitable for this market a large number of che-mical, physical, biological and economic cha-racteristics of the constituents must be takeninto account (Table 1).

Due to its characteristics Sphagnum peat hasbeen the most important constituent of orga-nic growing media for several decades. A highwater and air capacity in combination withother favourable features make peat ideal forthis purpose. After fertilising and liming, peatis the sole constituent of many growing media.Nonetheless, peat has also some drawbacks

arising from the relatively poor wettabilitywhen dry, limited nutrient buffering capacity,and structural stability. Mineral amendments,especially those rich in clays, are commonlyincorporated into the organic media in orderto improve physical and chemical conditions. In horticultural practice, clays are amended topeat in amounts of 10-80 kg/m³. The amend-ment aims mainly to improve the followingthree different properties of the growingmedia:1. Buffering of nutrients. Clay amendment

shall ensure that optimum conditions adju-sted at start of the crop remain constantduring cultivation. Relative large amounts ofclay are needed.

2. Rewettability during irrigation. Due to thelarge specific surface area and charged surfa-ce sites, hydrophilic mineral surfaces of clayminerals coating the surfaces of peat shouldhave a positive effect on the wettability ofinherently water repellent growing mediaand the amount of plant available water.Relative low amounts of clay are needed.

3. Binding capacity. For the use of prickingrobots when transplanting young plants andalso for the production of peat blocks (e.g.4 x 4 x 4 cm) for mechanical sowing the co -hesion of the growing media shall be impro-ved by the amendment of a suitable clay.Relative low amounts of clay are needed.

As a certain clay can hardly support all of therequired properties, also blends of differentclays are used in growing media industry. Claysavailable show great differences in their mine-ral parameters, e. g. texture, mineralogical com -position, layer charge, and oxide content.Chemical as well as physical characteristics ofthe growing media strongly depend on thekind of clay amendment. Currently no proven

Table 1: Properties of growing media and their constituents that pertain to »quality«(adapted from Schmilewski, 2008).


142

standards are available for evaluation of clayfor use in growing media and to design gro-wing media perfectly for specific crops andproduction methods. Here improved knowled-ge of the mechanisms and processes of theseclays in growing media helps to identify suita-ble clays and to make the prognosis of clayproperties more reliable.

Surface chemistry of growing media con-stituentsThe amendment of clays or clay blends to peattogether with the addition of fertilizers andlime in a growing medium has a strong effecton the surface chemistry of its constituents.Clay and peat stem from very different envi-ronments and a distinct number of exchangereactions can be assumed. E. g. the clays usedin this project from Mesozoic-Tertiary weathe-ring mantle of the Rhenish Massif are undersa-turated for the binding of dissolved organicmatter (DOM), whereas peat is known to be asource for DOM. For this reason the externalsurfaces of clay minerals might act as a sink forDOM, whereby in turn the adsorption of DOMcan modify the surface properties of the clayminerals. Here the interaction of water withpeat and clay minerals defines the physical andchemical properties of the growing medium,as virtually all chemical and biological reactionstake place at the solid-water interface. A sche-

me of the interactions of water with the gro-wing media constituents is shown in Fig. 1.

The solid-water interface is important becauseit controls the retention and transport ofnutrients and pollutants and provides physicalsupport for plants. Also contained in the poresolution besides inorganic solutes is DOM, itsfate and transport highly depending on interac-tions of the solid-water interface. On addition,many chemical species in the growing mediasolution can interact with each other. Solutebehaviour is controlled by multiple solid phases,different clay minerals, oxides and organic mat-ter, which have a complex and heterogeneousnature. The type of surface reaction has a deci-sive role for the potential of the solid phase forthe desired properties listed above.

Water repellent surfaces of growingmediaWater repellent (hydrophobic) growing mediacan cause problems as, after drying out, theyrequire a long time to rewet (Michel et al.,2001). For that reason the rewettability couldbecome a property with major repercussionsfor plant growth (Doerr et al., 2000). Organicmaterial in peats used for growing media com-prises many substances which are only weaklyor non-polar. As a consequence they are notable to form hydrogen bonds with water mole-

Figure 1: Scheme for the interactions growingmedia constituents withsolution.


143

cules and hence exhibit often distinct waterrepellence (hydrophobicity). On the otherhand, charged mineral surfaces are usuallyhydrophilic as they are capable of forminghydrogen-bonds. Thus, when incorporatedinto a water repellent sandy soil or a growingmedium, it is expected that charged clay mine-rals show a positive influence on the wettabili-ty, by increasing the surface area, maskinghydrophobic organic surfaces and exposinghydrophilic clay surfaces (McKissock et al., 2002;Lichner et al., 2006). The effect of mineralamend ments on the wettability of soils wassubject to several studies. Nevertheless, theresults were somewhat inconsistent. Indicationwas obtained that the variations observeddepend to a great extent on the formation oforganic coatings on mineral surfaces. Theadsorption of weakly or non-polar organicmatter moieties can render a hydrophilic mine-ral surface hydrophobic (Chenu et al., 2000;Doerr et al., 2000). The hydrophilic reactivity ofmineral surfaces and accordingly their positiveinfluence on wettability can be reduced oreven lost, when the minerals tend to accumu-late hydrophobic organic compounds derivedfrom soil organic matter or peat on their exter-nal surfaces. For soils sorption of dissolved organic matteronto mineral surfaces was found to be affec-ted by the parameters CEC, SSA and contentof amorphous oxides (e. g. Kaiser and Gug gen -berger, 2000; Kahle et al., 2004). The stabilityof organic coatings depends on the type ofbond, which is not only dependent on the che-mical composition of DOM, but also on mineralsurface properties. Basically it has to be assu-med that wettability is largely influenced by thesorption of DOM and, as the wettability is adynamic property, also desorption processes. In the study on rewettability of growing media,the effect of certain clay amendments, highlydifferent in their mineral parameters (mineralo-gical composition, specific surface area, oxala-te and dithionite soluble Fe- and Al-oxides, tex-ture) was determined. Processes and reactionsresponsible for the variations in growing mediawettability are assessed and used to rate theeffectiveness of the amendment. The results

will be used to find reliable criteria for the selec-tion of suitable clays for improving wettability.

Clays for improving P buffering capacityThe amendment of clay to peat-based growingmedia is supposed to improve both physicaland chemical properties of growing media.Pure white peat is nearly non-buffered and thebuffering capacity for the nutrients potassiumand phosphorus can be increased by the addi-tion of clay. A buffered growing media canabsorb nutrients in large amounts by bindingthem and is able to release them again in thesolution of the growing media when plantsdeplete the nutrient concentration. By this,fluctuations of plant demand or fertigation canbe balanced and optimal growing conditionscan be maintained. The buffering capacity of a clay in growingmedia is mainly caused by its contents ofamorphous Fe- and Al-oxides. Therefore P fixa-tion varies widely between clay minerals andthe choice of the »right« clay is very important.In horticultural practice the basic fertilizationof a growing media is generally not sufficientto supply the plants over the whole vegetationperiod with nutrients and subsequent fertiga-tion is necessary. The nutrient concentration inthe fertigation solution should reflect thenutrient demand of plants. Growth rate andnutrient uptake of plants differs widely bet-ween species and cultivars and is dependenton environmental conditions. Therefore opti-mum fertigation concentration can be deter-mined only with uncertainties. Thus, i) clay characteristics affecting P buffe-ring capacity were determined and ii) the in -fluence of the buffering capacity of peat/clay-growing media on the safety of plant cultiva-tion at varied concentrations of P fertigationwas investigated.

Progress within the projectThe work in the first two years focused on thetwo sections »Buffering of nutrients« (1) and»Rewettability during irrigation« (2). In prelimi-nary studies in the section »Binding capacity«(3) different methods were tried out and ex -tensive experiments on blocked growing me -


144

dia by mechanical agitation will be carried outin the third year.Determinations on Manganese toxicity (section1) on clays with different contents of active Mn(sum of exchangeable and easy reducable Mn)reveal that a threshold value for Mn in clays forgrowing media is not justified (Dombrowskiand Schenk, 2010). Even very high Mn con-tents in clays only resulted in small increase ofMn concentration in the solution of the gro-wing medium. Additionally plants toleratedhigh Mn-concentrations in the growing mediasolution because, i) Mn was complexed by DOMand thus is less phytotoxic ii) silicic acid dissolvedin growing media solution alleviated the harm-ful Mn effects. Results on the P binding capaci-ty of clays, strongly correlated with the oxalateextractable Fe and Al content of clays, are des-cribed in detail in the result section.For the determination of the rewettability ofgrowing media the capillary rise method wasused. For this purpose a WOK-apparatus allo-wing 8 simultaneous measurements was in -stal led at Klasmann-Deilmann GmbH. The ef -fectiveness in improving the wettability of gro-wing media was found to depend on the sur-face properties of the clay minerals amendedwhich govern the binding of DOM. Amend -ments with non-expandable clay minerals withlow SSA, CEC (e.g. kaolinite and illite) and clayswith a low amorphous oxide content appearmost promising (Walsch and Dultz, 2010). Thefineness of the clay was observed to be a deci-sive factor for improving the rewettability.

Material and Methods

RewettabilitySeven clay samples with distinct differences inmineral parameters (Tab. 1) and a blend of dif-ferent Sphagnum peats with a moderatedegree of decomposition were used.

The growing media was limed (6.0 kg/m3) andfertilized (1.1 kg NPK 14-16-18 standard ferti-lizer per m3 growing media. The rewettabilityof the growing media mixtures was assessed interms of capillary water uptake (WOK, DutchRHP foundation) and contact angle (CA) mea-surements (OCA 20, Data Physics). In the WOKapparatus dry samples are placed in cylindricalrings (500 cm3) on a layer of water and theincrease in water content is logged over timefor 24 h (Fig. 2).

Hydration of particles in the growing mediawas determined in an environmental scanningelectron microscope (ESEM; Quanta 200, Fei).As peat surfaces consisting mainly of C arecovered with clay minerals rich in Si, thedegree of coverage of the peat surface can bedescribed by the C/Si ratio. By energy dispersi-ve X-ray spectroscopy five different sectionseach of 6 mm² size were determined.The adsorption of peat derived DOM on mine-ral surfaces was determined in batch adsorp-tion experiments. DOM solutions for adsorptionexperiments were extracted from limed and fer-tilized peat. Batch adsorption experiments werecarried out for 24 h at 20 °C in the dark. DOCconcentrations were measured with a totalorganic carbon analyser (liquiTOC, elementar).Aromaticity of the adsorbed DOC and DOM-fractionation phenomena due to adsorption

Table 1: Physical, chemical and mineralogical properties of seven clay samples from the Westerwald area, representingsaprolithic clay (S), bentonite (B), translocated clays (U) and blends (M).


145

were determined with a Cary 50 UV-Vis spec-trometer (Varian). According to Chin et al. (1994)absorbance values were recorded at a wave-length of 280 nm, where electron transitionsoccur for a number of aromatic substances.

P buffering capacityThe influence of P buffering of peat/clay gro-wing media on the safety of plant cultivationwas characterized in a plant experiment. Atfirst nine clays with different P fixation capaci-ty were selected for the determination of plantavailability of P in peat/clay growing media. Toassess the P fixation clays and growing mediawere shaken in a P solution (1000 mg P L-1) ata ratio of 1:10 for 24h. Oxalate soluble Fe andAl were extracted using 0.2 M oxalate solution.Based on these data two clays with a low P fixa-tion (04S and 01S) and two with a high P fixa-tion (06B and 39W) were selected for the plantexperiment and mixed with peat (80 vol-%peat, 20 vol-% clay). For each growing media a calibration curvewas established to determine the amount of Pneeded to obtain a CAT extractable P concen-tration of 25 mg P/L growing media, which isusually adjusted in horticultural practice. CAT(0.01 M CaCl2 + 0.002 M DTPA) is commonlyused for the extraction of potentially plantavailable P in horticultural growing media (Alt& Peters, 1992). The growing media wereequilibrated in an oven at 50°C for 48 h, thenat room temperature for another 48 h prior todetermining the CAT extractable P concentra-tion. Previous work showed that this incuba-

tion procedure is highly correlated with CATsoluble P after 9 weeks of storage. Seedlings of Impatiens walleriana F1 »Candy ®

Coral Bee« were planted in 600 ml plastic pots(∅ 12 cm) filled with the different growing me -dia and pure white peat as control. Volumeweight of growing media was determinedaccording to standard method of VDLUFA(1991). All growing media were fertilized with1.5 g L-1 of a P free compound fertilizer (Ferti8® – N : P2O5 : K2O = 20 : 0 : 16 + micronu-trients) and pH was adjusted with CaCO3 topH 6 (0.01 M CaCl2). The CAT soluble P con-tent of 25 mg P/L growing media was obtainedby addition of Ca(H2PO4)2 according to thecalibration curves. The fertigation solution con-tained in mg/L solution: 120 N as KNO3 andNH4NO3, 130 K as KNO3, K2SO4 and KH2PO4,10 Mg as MgSO4*7 H2O and 200 Flori® 10.The P concentration of the fertigation solutionwas varied (in mg P/L solution): 0 (= no P), 17(= sufficient P) and 35 (= excess P). Plants wereharvested after 10 weeks cultivation after mea -suring plant height and diameter and length ofinternodes. The P concentration in plant d.m.was measured after wet digestion and the Pconcentration in the solutions was determinedwith the ammonium-vanadate-molybdate me -thod. Treatments were replicated five times(each replicate consisted of 5 plants) in a com-pletely randomized design and statistical ana-lysis was performed with the program R 2.8.1.Means were compared between treatments ata = 0.05 using Tukey-Test.

Figure 2: WOK apparatusused for the determinationof the time dependentwater uptake capacity.


146

Results

Rewettability

ESEM observationsGrowing media submitted to hydration showclearly that the condensation of water startson the hydrophilic clay particles coating theexternal surfaces of the peat compounds (Fig.3a, b). The formation of water drops on thesurface of peat indicates high contact angelsand strongly hydrophobic properties.

Saprolithic and translocated clays have thestrong est effect on surface coverage of peatcompounds with clay minerals (Fig. 4). Herealready 20 kg/m³ are sufficient to reach lowestC/Si ratios indicating maximum coveragedegree. Bentonites and clay blends still increa-se the coverage degree of the peat`s surface bythe addition of 30 kg/m³. Bentonites showsmall effects even at the largest amendedamount. Here C/Si-ratios are comparable withgrowing media amended with saprolithic andtranslocated clays in amounts of 10 kg/m³.

Water uptake is improved for all samples bythe amendment of clay (Fig. 5), strongly de -pen ding on the clay parameters. Clays consi-sting mainly of illite and kaolinite (saprolithicand translocated clays) show fastest wateruptake (50 vol.% within 10 min).

At C/Si ratios <20, where the surfaces of peatare most completely coated with clay minerals,all clay-peat systems show the highest wateruptake rate. Highest variability is observed for asaprolitic clay rich in illite and kaolinite, wherethe water uptake rate ranges from 2.6 to 15.5(%v/v)/min. Bentonites show only minor effects,water uptake ranges from 1.47 to 3.63 (%v/v)/min. The translocated clay is most effective toimprove water uptake rate with small amen-ded amounts. Larger added amounts showonly slight improvements. For the clay blendand bentonite, the absolute amount of amen-ded clay showed minor effects, but wateruptake rate is quite higher than that of originalpeat (1.3 (%v/v)/min). The results show a good

Figure 3a, b: Environmental scanning electron microscope(ESEM) observation of a peat-based growing media amen -ded with a translocated clay (27 U) submitted to hydra-tion conditions. The image shows a leave of Sphagnummoss peat covered with clay platelets in the dry state (a)and directly after beginning of condensation (b). Notethat condensation starts at sites where clay minerals coatthe peat´s surface. Water droplets are marked with arrows.

Figure 4: Coverage degree of peat surfaces with clayminerals expressed by the C/Si ratio (n=5). Low valuesindicate highly covered surfaces.


147

correlation of the surface coverage degree ex -pressed by the C/Si ratio and rewettability deter-mined by the capillary rise »WOK« method. Thedetermination of the C/Si ratio in a relativelyshort procedure is a suitable method to identifysuitable clays and the amount of clay needed.

Effect of clay particle size on water uptakecharacteristics Time dependent water uptake measurementson growing media with clays of different parti-cle size fractions reveal that the initial particlesize of the amended clay has a strong influen-ce on the water uptake rate. For fine aggrega-te sizes <0.5 mm the fastest water uptake wasobserved (Fig. 6). Note that in horticulturalpractice irrigation is often performed in theebb and flow mode. If the uptake rate is highthe time needed for irrigation can be shorte-ned. After the total duration of the experimentof 1440 min (24 h) in the WOK apparatus thedifferences in water uptake rate are low. This isnot relelevant for horticultural practice, wherethe differences at the beginning of irrigationare most important. Here the growing mediawhere the smallest aggregate size (< 0,5 mm)and the highest amount of amended clay (20kg/m³) shows the fastest water uptake rate inthe time period up to 15 min.

Effect of different clays on water uptake cha-racteristics From the water uptake curves, it becomes evi-dent that the growing media differ significant-ly in their rewetting behaviour (Fig. 7). Purepeat showed the slowest water uptake, whe-reas the increase in the volumetric water con-tent of mineral amended growing media wasgenerally faster. Up to 50% water saturation,also marked variations in the rewetting beha-viour among the different mineral amend-ments were observed. Amendments of smecti-te rich clays (06B and 07B) caused only a minorincrease in water uptake velocity, while kaolini-tic and illitic mineral amendments notablyenhanced the rewetting. It can be concludedthat the wettability of a growing media is adynamic property, which is to a great extentinfluenced by the clay mineralogy and alsoother mineral parameters.

Formation of hydrophobic coatings onclay mineral surfacesMineral surfaces provide a variety of reactivefunctional groups on their surface, whichallows strong interaction with dissolved orga-nic and inorganic substances. Consequently,batch adsorption experiments with peat deri-ved DOM revealed that both, the bulk samples

Figure 5: Effect of four different clays on the water upta-ke rate. Growing media are amended in amounts of 10,20 and 30 kg/m³ clay. Water uptake rate is determined bythe time (min) needed to 50% saturation. The dashedline is the 95% confidence interval of the linear regres-sion (solid line).

Figure 6: Time dependent water uptake of a growing me -dia (fineness of peat: 0-10 mm) to which translocated clay36M was added in two different amounts (10 and 20 kg/m³)und two particle size fractions (<0.5 and 0.5-2.0 mm).


148

as well as the clay fractions adsorb relativelyhigh amounts of dissolved organic substances(Fig. 8). For the samples under investigation,the adsorbed C content varies between 20 and180 mmol C kg-1, with the highest amount be -ing adsorbed by the samples rich in smectite(06B and 07B).

The composition of the organic coatings onthe mineral phase was determined by UV-Visanalyses of the equilibrium solutions after thebatch adsorption experiments (Fig. 9). Here, adecrease in light absorbance values (Abs280)indicates a high removal of aromatic organicsubstances from the solution due to the sorp-tion onto mineral surfaces.The extent of sorption was found to be depen-dent on the mineral parameters of the samples(Fig. 9, a, b, c). Linear regression analysis sho-wed that sorption of aromatic compounds cor-relates well with the SSA, CEC and amount ofoxalate soluble Al and Fe, which is consistentwith findings reported elsewhere (e.g. Kaiserand Guggenberger, 2000; Kahle et al., 2004).Absorption at 280 nm shows strongest corre-lation to the SSA (r2 = 0.86), which is most like-ly due to the higher surface area available for

DOC sorption. Considering liming and fertiliza-tion, it can be suggested that DOM sorption ismost likely also mediated by the solution che-mistry. For instance Feng et al. (2005) foundthat sorption of peat derived humic acid ontokaolinite and montmorillonite increased withincreasing ionic strength and decreasing pH,and that the presence of Ca2+ largely enhan-ced the sorption. Subsequently, DOC linkageto negatively charged mineral surfaces viacation-bridges appears to be an important bin-ding mechanism, which is also indicated by thepositive correlation (r2 = 0.66) between adsor-bed DOC and the CEC of the clay amendment.The correlation with the sum of oxalate extrac-table Al and Fe (r2 = 0.71) is indicative for DOCbinding onto poorly crystalline Fe and Al mine-rals (Mikutta et al., 2005). It can be assumedthat the amount and composition of adsorbedDOM as well as the mineral parameters are thekey to explain the observed variations in wet-tability and accordingly the differences in theeffectiveness of the mineral amendments.

Effect of hydrophobic mineral coatings onwater uptake rate Adsorption of peat derived DOM onto mineral

Figure 7: Time dependent water uptake of peat-based growing media amended with differentclays in an amount of 30 kg/m3. For comparison, the water uptake of the pure peat is shown.


149

surfaces influences the wettability of the gro-wing media in a negative way (Fig. 10). Fromthe non-linear regression it becomes evidentthat variations in wettability of the growingmedia can be attributed to a high extent todifferences in the affinities of the mineral sur-faces for aromatic DOM compounds. Mostprobably the hydrophilic behaviour of themineral surfaces is affected by the adsorbedorganic compounds forming hydrophobicsurface coatings, diminishing their positiveeffect on the wettability.

Conclusions on rewettabilityIt was shown that the absorption of aromaticorganic compounds changes the hydrophilicityof clay mineral surfaces. The intermolecular in -ter actions involved in the formation of hydro-phobic coatings can be very complex. However,sig nificant effects can be expected if specificpro perties of the dissolved organic mat ter isvaried (molecular size, configuration and pola-rity or hydrophobicity) and also specific surfaceproperties of the adsorbents (texture, minera-logical composition and the distribution ofreactive functional groups). Additio na lly, theproperties of the organic and inorganic reac-tants can be affected by aqueous phase proces-ses (e. g. hydrolysis, hydration, dissolution, andprecipitation). Here the composition and pro -perties of the aqueous solution (e. g. pH, ionicstrength, presence of bridging-ca t ions or com-petitive anions) have to be considered as well.

For instance, Tarchitzky et al. (2000) explainedthe role of polyvalent cations for water repel-lency with the enhanced formation of coatingson mineral surfaces, resulting in a reduction ofthe hydrophilic reactivity. Additionally it has

Figure 9: UV-Vis light absorbance values of the equili-brium solutions after batch adsorption experiment. Note,a decrease in the Abs280-values indicates a high removalof peat derived aromatic organic substances due to sorp-tion onto mineral surfaces. Abs280-values are shown inrelation to mineral properties a) CEC, b) SSA and c) oxala-te soluble Fe and Al. The dashed line is the 95% confi-dence interval of the linear regression (solid line).

Figure 8: Sorption of peat derived DOM on bulk clay sam-ples and the clay fractions.


150

been suggested by Hurraß and Schaumann(2006) that bridges by polyvalent cations maycause an aggregation of the humic substancestht would also result in a lower wettability.

The results of our study show that the obser-ved differences in rewettability can be mainlyattributed to the extent and kind of organiccoatings on the mineral surfaces. Obviouslythe formation of these coatings relies on diffe-rences in the sorption properties of the mine-rals and, in turn on distinct mineral properties.Mineral amendments containing expandableclay minerals with high CEC and SSA (e. g.smectites) and amendments rich in amorphousFe and Al compounds seem to be less effecti-ve, due to their high affinity for peat derivedDOM, especially for hydrophobic DOM consti-tuents (Fig. 11). Accordingly, an addition ofnon-expandable clays with lower SSA, CECand amorphous oxide content (e.g. kaoliniteand illite) seems more promising.

P buffering capacityThe nine selected clays differed clearly in theirP binding capacity and content of ∑Feox+Alox

(Fig. 12). The P binding of the clays correlatedpositively with the ∑Feox+Alox, but could notfully explain the whole variation in P fixation. Itis assumed that the surface of the oxides dif-fers widely, which will be investigated in furt-her experiments.

After ten weeks cultivation the varied P ferti-gation resulted in clear differences in plantgrowth and quality among the treatments andgrowing media. Without P fertigation the elongation of shootswas reduced in pure white peat and the twoclays with low P binding capacity 04S and 01S,compared to peat/clay blends containing clay06B and 39W. Also P fertigation increasedinternode length in these blends (Fig. 13). Thesame trend was observed in plant diameterand height (data not shown).

Figure 10: Water uptake of growing media amended with different clays in relation to the adsorption of aro-matic compounds from peat derived DOM. The measure for the absorption of aromatic organic compoundsfrom the equilibrium solution is the absorbance at 280 nm.


151

In all growing media the dry matter yield of theplants not fertigated with P was reduced com-pared to those receiving P. Without P fertiga-tion yield of treatments in pure white peat andin the peat /clay blend 04S was lower than inthe other three peat/clay mixtures (Fig. 14).

The increasing P supply via fertigation resultedin increased P conc. in plant d.m. in all growingmedia. Without P fertigation, P concentration inshoot d.m. was the lowest in treatment purewhite peat and the peat/clay blend 04S and thehighest P concentration was obtained in the tre-atment 06B. The same differences between gro-wing media were observed with fertigation con-centration 17 mg P L-1 solution, but no longer atthe highest P fertigation level (Fig. 15).

Figure 11: Water uptake of peat-based growing media mix -tures amended with different clays with an amount of 30kg m-3 and its relation to mineral properties a) CEC, b) SSAand c) oxalate soluble Fe and Al. Dashed line is the 95%confidence interval of the linear regression (solid line).

Figure 12: Relationship between the content of oxalatesoluble Fe and Al and the P binding of nine different claysin a batch experiment with a P supply of 10000 mg kg-1.Marked clays were used in the plant experiment.

Figure 13: Length of internodes of Impatiens wallerianadepending on growing media type and P conc. in fertiga-tion solution. Different capital letters indicate significantdifferences between growing media at the same P ferti-gation level and different small letters indicate significantdifferences between fertigation treatments (p<0.05).


152

To characterize the capacity of different claysto supply P to plants the P amount containedin shoot d.m. was calculated for the treatmentwithout P fertigation (Fig. 16). The clays 01S,06B and 39W supplied about twice as much Pto plants compared to pure peat and peat/clayblend 04S. In the latter treatments plants didnot exhaust the P amount determined by CATextraction whereas in the other treatmentseven more P was taken up than characterizedas plant available by the extraction. In all gro-wing media more P was fertilized than deter-mined as plant available or taken up in shoot

d.m., which was more pronounced in blendswith clays having a high binding capacity (01S,06B and 39W). At the same CAT level, in thiscase 25 mg/L growing medium, peat/clayblends containing clay with a high P bindingcapacity supplied more P to plants indicating thecapacity to buffer fluctuating P supply. In horti-cultural practice P availability in growing mediais characterized by CAT extraction procedure.

Conclusions on P buffering capacityThe P binding capacity of clays was stronglycorrelated with the oxalate extractable Fe andAl content. However, this factor did not fullyexplain the observed variability. Low availabilityof P reduced internode length and supportedcompaction of plants. But at the same timed.m. production was decreased indicating thatshortage of P is not a successful strategy to pro-duce more compact and better shaped plants.Peat-clay blends containing clay with a high Pbinding capacity provided up to twice as muchP to plants than those with a lower bindingcapacity, thus buffering fluctuations in P uptakeand P fertigation. The common extraction pro-cedure for horticultural growing media tendedto underestimate the plant available P in peat-clay blends with high binding capacity. Toimprove the understanding of P binding of claysspecific surface of oxides will be investigated.

Figure 14: Dry matter of Impatiens walleriana dependingon growing media type and P conc. in fertigation solu-tion. Different capital letters indicate significant differen-ces between growing media at the same P fertigationlevel and different small letters indicate significant diffe-rences between fertigation treatments (p<0.05).

Figure 16: Amounts of fertilized and CAT-extractable Pper plant container (0.6 L growing media) and P contai-ned in shoot d.m. of Impatiens walleriana depending ongrowing media type.

Figure 15: Phosphorus conc. in d.m. of Impatiens walleri-ana depending on growing media type and P conc. in fer-tigation solution. Different capital letters indicate signifi-cant differences between growing media at the same Pfertigation level and different small letters indicate signifi-cant differences between fertigation treatments (p<0.05).


153

ReferencesAlt, D. & Peters, I. (1992) Die CaCl2/DTPA- Me -thode zur Untersuchung gärtnerischer Erden aufMengen- und Spurenelemente. Agribiol. Res.45, 204-214.

Chenu, C., Le Bissonnais, Y. & Arrouays, D.(2000) Organic matter influence on clay wetta-bility and soil aggregate stability. Soil Sci. Soc.Am. J. 64, 1479-1486.

Chin, Y.P., Aiken, G. & O’Loughlin E. (1994)Molecular weight, polydispersity, and spectro-scopic properties of aquatic humic substances.Environ. Sci.Technol. 28, 1853-1858.

Doerr, S.H., Shakesby, R.A. & Walsh, R.P.D.(2000) Soil water repellency: its causes, cha-racteristics and hydro-geomorphological signi-ficance. Earth-Science Reviews 51, 33-65.

Dombrowski, I. & Schenk, M.K. (2010) Mang -anese toxicity in peat/clay blends used as gro-wing media? Conf. »Genetics of Plant MineralNutrition«, Deutsche Gesellschaft für Pflanzen -ernährung, Abstract Book p. 62.

Feng, X., Simpson, A.J. & Simpson, M.J. (2005)Chemical and mineralogical controls on humicacid sorption to clay mineral surfaces. OrganicGeochemistry 36, 1553–1566.

Hurraß, J. & Schaumann, G.E. (2006) Proper tiesof soil organic matter and aqueous extracts ofactually water repellent and wettable soil sam-ples. Geoderma 132, 222-139.

Kahle, M., Kleber, M. & Jahn, R. (2004) Re ten -tion of dissolved organic matter by phyllosilica-te and soil clay fractions in relation to mineralproperties. Organic Geochemistry 35, 269-276.

Kaiser, K. & Guggenberger, G. (2000) The roleof DOM sorption to mineral surfaces in thepreservation or organic matter in soils. OrganicGeochemistry 31, 711-725.

Lichner, L., Dlapa, P., Doerr, S.H. & Mataix-So -lera, J. (2006) Evaluation of different clay mine-

rals as additives for soil water repellency allevi-ation. Appl. Clay Sci. 31, 238-248.

Marschner, H. (1995) Mineral nutrition of higherplants. 2nd ed. Academic Press, London, UK.

McKissock, I., Gilkes, R.J. & Walker E.L. (2002)The reduction of water repellency by addedclay is influenced by clay and soil properties.Appl. Clay Sci. 20, 225-241.

Michel, J.C., Riviere, L.M. & Bellon-Fontaine,M.N. (2001) Measurement of the wettability oforganic materials in relation to water contentby the capillary rise method. European Journalof Soil Science 52, 459-467.

Mikutta, R., Kleber, M. & Jahn, R. (2005) Poorlycrystalline minerals protect organic carbon inclay subfractions from acid subsoil horizons.Geoderma 128, 106-115.

Schmilewski, G. (2008) The role of peat inassuring the quality of growing media. Miresand Peat 3, 1-8.

Tarchitzky, J., Hatcher, P.G. & Chen, Y. (2000)Properties and distribution of humic substan-ces and inorganic structure-stabilizing compo-nents in particle-size fractions of cultivatedMediterranean soils. Soil Sci. 165, 328– 342.

VDLUFA, 1991. Bestimmung der Rohdichte(Volumengewicht) von gärtnerischen Erdenund Substraten ohne sperrige Komponenten.VDLUFA Methodenbuch, Band 1. DieUntersuchung von Böden. 4. Auflage, VDLUFAVerlag, Darmstadt.

Verhagen, J.B.G.M. (2004) Effectiveness of clayin peat based growing media. Acta Hort. 644,115-122.

Walsch, J.& Dultz, S. (2010) Improving thewettability of peat based growing media bythe amendment of clay. Acta Mineralogica-Petrographica Abstract Series 6, 5th Mid-European Clay Conference, Budapest, Hun -gary, p. 85.


154

SURFTRAP - Development and Optimisation of a Process to Biosynthesize Reactive IronMineral Surfaces for Water Treatment Purposes

1. IntroductionIn this project we aim to develop a low-costtechnology to remove ionic constituents fromraw waters such as arsenic species. The propo-sed technology is based on the reactivity ofschwertmannite, an oxyhydroxosulfate of themean stochiometry Fe8O8(OH)6SO4 (molarmass 772.89 g/mol). This mineral typicallyforms in acidic and sulfate rich mine waters asa secondary mineral upon oxidation of Fe(II) ina biologically mediated process. Schwert man -nite can be generated in a biotechnologicalpro cess after aeration of mining process wa -ters. It forms surface-rich aggregates of need-le-like nanocrystals. It rapidly transforms intoferric hydroxides of high specific surface areaonce exposed to water containing at leastsome alkalinity. Our rationale follows the con-

cept to make use of this transformation reac-tion by adding biosynthesized schwertmanniteto contaminated raw waters where it genera-tes a large sorption capacity to remove the pol-lutants (Peiffer et al., 2008).

2. Materials and Methods

2.1. Microbial Investigations

2.1.1. Depth profile of activity of microorga-nisms in schwertmanniteA Schwertmannite core from the carrier mate-rial of the pilot plant was collected with a hol-low drill and cutted into 0.5 cm layers (volumeof each layer=1.7 cm³). The mineral was dis-solved with 0.2 M oxalic acid and the cells

Peiffer S.*, Paikaray S., Damian Ch. (1), Janneck E., Ehinger S., Martin M. (2), Schlömann M., Wiacek C.,

Kipry J., (3), Schmahl W., Pentcheva R., Otte K., Götz A., Wang Z., Hsieh K. (4), Meyer J., Schöne G. (5),

Koch T. (6), Ziegler A. (7), Burghardt D. (8)

(1) Department of Hydrology, University of Bayreuth, Universitätsstraße 30, D-95440 Bayreuth

(2) GEOS Ingenieurgesellschaft mbH, Gewerbepark »Schwarze Kiefern«, D-09633 Halsbrücke

(3) Department of Environmental Microbiology, Tech. University of Freiberg, Leipziger Str. 29, D-09599 Freiberg

(4) Department of Earth and Environmental Science, Section Crystallography, University of Munich, D-80333 Munich

(5) Wismut GmbH, 09117 Chemnitz

(6) Vattenfall Europe Mining AG, 03050 Cottbus

(7) Central Facility for Electron Microscopy, University of Ulm, 89069 Ulm, Germany

(8) Institute for Groundwater Management, Technical University of Dresden, 01062 Dresden


Table1: 16S rRNA probes used for FISH analyses


155

were washed with 1x PBS buffer. The resuspen-ded cells were analysed with the LIVE/DEAD®

BacLightTM Bacterial Viability Kit (Invitrogen,L13152) and FISH (fluorescence in situ hybridi-sation). The FISH analyses were performed afterHallberg et al. (2006) with probes for Ferrovummyxofaciens and Gallionella TrefC4 (tab. 1).

2.1.2 Biologic Analysis / Interface Bacteria –MineralWe investigated samples from cultures of thespecies Leptospirillum ferrooxidans. To investi-gate the interface between mineral and orga-nic matter the organic matter had to be fixa-ted. This was done by physical (high pressurecryo-fixation) and chemical methods (2.5 % glu -tardialdehyde/2.0 % paraformaldehyde mixtu-re in 0.2 M cacodylate butter (pH 4.5)). Se ve ralsteps for drying and contrasting followed. Thefixated samples were embedded in Epon andcut into 70 nm thick sections with a DIATOMdiamond knife on a microtome for the TEM.For the SEM work, the samples were criticalpoint dried and coated with platinum. Sampleswere taken from two positions: from the sedi-ment and from a bioactive TERMINOX foil.

2.2 Schwertmannit generationThe designed pilot plant consists of an oxida-tion basin with removable growth carriers aswell as an aeration and a precipitation tank.The overall volume is about 10.5 m³ and theoxidation basin has a capacity of 8.14 m³. Dueto the high flow rate of the circulation pump(approximately 30 m³/h) low gradients of pro-cess parameters can be guaranteed. The waterinflow is intensively aerated. A chain cleanerassembled at the bottom of the oxidation ba -sin removes dropping schwertmannite to thesludge colllecting channel. From there thesludge is pumped to a storage tank where thesludge can sediment and thicken. The thicke-ned sludge can be recycled to the oxidationbasin or pumped into big bags for dewateringby gravity.From the chemical engineering point of view,the designed pilot plant is a hybrid type of bio-reactor. It has physical characteristics of afixed-bed reactor in parallel with a circulation

reactor. The advantages of this special reactordesign are avoiding plugs in the fixed-bed(growth carrier) and a free circulation insidethe oxidation basin. The pilot plant was operated continuouslysince the start of the project.

2.3. Mineralogical and Structural Analysis

2.3.1. Mineral analysisThe crystal structure of the prepared precipita-tes (see below) was examined by XRD. XRDmeasurements were performed on a Stoe pow -der diffractometer and on an Oxford Dif frac -tion area detector diffractometer, both usingMo-ka1 radiation. Scanning electron microsco-pe (SEM) analyses of the biologic samples wereperformed with a Hitachi S-5200 field emissionscanning electron microscope equipped withan EDX and a STEM detector. For the SE-pictu-res an accelerating voltage of 4kV was usedand for the EDX and STEM analyses 20kV wasemployed. SEM pictures of the mineral sam-ples were taken on a JEOL 6500 F SEM with5kV accelerating voltage. Specific surface areas were measured by Bru -nauer-Emmett-Teller (BET) method (Bru nau eret al., 1938) using a FlowSorb II 2300 (Fa. Mic - romeritics) and N2/Ar (80%/20%) as ad sor bate.Different methods from literature to produceschwertmannite were compared. The firstmethod was published by Bigham et al. 1990:2L of MilliQ water were heated to 60 °C and10.8 g of FeCl3*6H2O and 3 g of Na2SO4 wereadded. The solution was kept at 60 °C for 12mi nutes. The orange suspension was then pla-ced in a dialysis bag and dialyzed against MilliQwater. The water was changed daily during aperiod of 33 days. The precipitate was then fil-tered and vacuum dried. This sample will bereferred to as dialyzed schwertmannite. A se -cond recipe originates from Pentinghouse andwas first published by Regenspurg et al. 2004:5 g of FeSO4 was dissolved in 1L of distilled wa -ter. Then 5 mL of H2O2 (32%) were ad ded. Thesolution became brown-red, and after onlyfew minutes, a rust-colored precipitate couldbe observed. After 24 h of ripening in solutionat 25 °C the obtained material was vacuum-


156

dried. We will refer to that material as fast-H2O2-schwertmannite. The method of Pen ting - house was also applied in a modified way.H2O2 (0.03%) was pumped over 24 h continu-ously (flow rate: 5 mL/h) to 100 mL culture m -edia iFe containing 25 mM ferrous iron (John -son and Hallberg, 2007). The synthesis wasperformed at 30 °C. After the ferrous iron oxi-dation and the precipitation the precipitatewas harvested by centrifugation (6000 g, 15 min)and freeze dried. The samples will be calledslow H2O2-schwertmannite. A third recipe wasdescribed by Loan et al. (2004) and sampleswere synthesized in a slightly modified waywith 2.5 g of Fe2(SO4)3•xH2O that was dissol-ved in 500 ml of MilliQ water in glass bottleand stored at 85 °C for 24 hours. The precipi-tate was vacuum dried and this sample will bereferred to as 85 °C-schwertmannite.

2.3.2. XASXAS experiments were conducted at the SUL-Xbeamline at ANKA synchrotron facility inKarlsruhe, Germany. The investigated sampleswere synthesized by the dialyses bag methoddescribed above but initial solutions weremodified by addition of arsenate, chromate orvanadate to replace sulfate. Samples were pre-pared with concentrations of 1.65 mmol (16.5 %), 2.5 mmol (25 %), 5 mmol (50 %)and 10 mmol (100 %), respectively, the per-centage indicates the ratio of heavy metal tosulfate. Goethite and lepidocrocite sampleswere synthesized as references and exposed tosolutions of arsenate, chromate or vanadaterespectively. The solutions for the chromatestandards contained 0.7 mM, 2 mM and 10mM chromate, 1 mM and 5 mM arsenate and1 mM vanadate, respectively.

2.3.3. Density functional theory (DFT) calcula-tionsUsing systematic density functional theory(DFT) calculations we study the structural, e ner -getic, and sorption properties of schwertman-nite as well as goethite(101), akaganeite(100),and lepidocrocite(010) surfaces with differentterminations. The exchange and correlationpotential is treated within the LDA/GGA+U

approach which is explicitly suitable for thestrongly correlated FeOOH system.

2.4. Schwertmannite Transformation andArsenic Adsorption

2.4.1. SchwertmanniteTranformation testsStability experiments were carried out for fourmonths using 5 g/L (equivalent to 7.75 mmol/LSO4

2-) sediment load in milli-Q water using poly -propylene reactor vials with regular shaking (4-5 times per day) at room temperature and at -mospheric pressure. The system was investiga-ted at four different pH conditions, i.e., 5.0,6.0, 7.0 and 8.0. The pH adjustments weredone with NaOH/HCl prior to addition ofschwert mannite samples. Due to release ofprotons in the course of ageing processes(equation 2), additional pH adjustments weredone daily using 1.0 N NaOH and the amountof NaOH consumed was noted. Samples werecollected at regular intervals for four months.Similar stabilization experiments were conduc-ted at pH 3.0, i.e., within the stability range ofschwertmannite (pH 3-4.5) to distinguish therole of higher pH on the fate of sulfate. Thesolid phase was separated by filtration andoven dried at 60 °C.

2.4.2. As(III)-Schwertmannite interactionsThree different schwertmannite (SHM) speci-mens produced through chemical and biologi-cal processes were used in the current study.SHM_MS was synthesized in a mine water tre-atment plant (GEOS, Freiberg, Germany) bymicrobial oxidation of Fe(II) between pH 2.9-3.2 (Glombitza et al., 2007) from SO4

2- richmine waters. One schwertmannite sample wassynthesized by fast synthesis method called»oxidative synthesis« as described by Regens -purg and Peiffer (2005). 10 g of FeSO4.*7H2Owas dissolved in deionized water and ~ 5 ml of32% H2O2 was added drop wise to acceleratethe oxidation of Fe2+ to Fe3+. The reaction waspreceded for 24 hours and the pH remainedstable at 2.4. The precipitated orange colouredsolids were filtered and oven dried at 35-40 °C.The samples were used directly for further expe-rimental purposes without any treatment, cal-


157

led here after SHM_FS. Schwertmannite wasalso synthesised as described by Bigham et al.(1990) The specimen is called SHM_DS.The sorption studies were conducted as batchexperiments. The suspension was allowed toequilibrate for 5 days with a sediment load of0.25 g/25 ml at pH 3.0 and initial As(III) con-centrations ranging from 0.13 to 1.33 mmol/L.Preliminary kinetic experiments demonstratedthat equilibrium is achieved after this time. Thesuspension was continuously stirred during thewhole reaction time. The pH of the As(III) con-taining solutions were adjusted to pH3.0±0.05 before addition of schwertmanniteto avoid possible transformation of SHM atchanging pH conditions (Jönsson et al., 2005).The ionic strength was maintained at 0.01mol/L by NaNO3. Experiments were performedin the dark in the presence of oxygen in poly-propylene reactor vials that were preconditio-ned by 10 % nitric acid overnight. After theequilibration time samples were filteredthrough < 0.45 µm cellulose filter papers andthe aqueous phase was analyzed for pH, Fe(II),Fe(t), SO4

2- and As(t) concentrations. Solid pha -se was characterized by XRD, FTIR, SEM andXANES techniques. Solid phase arsenic loadingwas determined by mass balance (CS=(Ci-Ceq)*V/M), where Ci and Ceq are initial and equi-librium As concentrations (mol/L), M is mass ofsorbent (g), L is volume of solution (L). SHMsamples were stored in crimp sealed serum vialsin O2 free glove box to eliminate the effect ofatmospheric O2 in As(III) oxidation during stora-ge for As species measurement by XANES.Two commonly used statistical isotherm mo -dels were fitted with experimental data usingIsoFit v1.2 (Matott and Rabideau, 2008) toevaluate the sorption behavior. Langmuir mo -del (Eq. 1) demonstrates a monolayer sorptionmechanism with homogeneous sorption ener-gies, while Freundlich model is an empiricalmodel demonstrating multilayer sorption sites

and heterogeneous sorption energies (Eq. 2)(Weber and DiGiano, 1996).

2.4.3. Dumping Experiments with Arsenic-loa-ded Schwertmannite/Transformation productsLong-term experiments were set up to studythe development of the chemical bonding bet-ween the species and schwertmannite orschwert mannite transformation products uponageing in order to test their deposition proper-ties. Periodically, the desorption of arsenic bysynthetic rainwater (pH 5, elution 1d/month,B-tests) or groundwater (pH 7, continuous elu-tion, C-tests) was analysed. The »groundwater«was effluent water (< 0.1 mg As/L, <0.2 mgU/L) of the water treatment plant in Schlema-Alberoda (table 2). As summarized in table 5,five different »dumping sceniaros« were inve-stigated. Each test contains 9g mineral preci-pitate and was eluted periodically or continu-ously by 10 mL water. The experiments will beac com panied by XAS (X-Ray-Absorption-Spec - tros copy) and EXAFS studies, results werealready open.

2.4.4. Laboratory experiments for pilot testpreparationFollowing to SHM transformation and As-Ad -sorption experiments from the last year, a ot -her test was performed in duplicate. Therein,brown coal filter ash (FA) (co. the neutraliza-tion was used instead of the conventional do -sage of lime milk. 1 L influent water from thewater treatment plant in Schlema-Alberoda(average composition see table 2) was acidifiedby 0.1M HCl and air-stripped for 30 minutes inorder to eliminate inorganic carbon. After fil-tration (0,2 µm cellulose acetate), FA (about0.15 g/L as 0.5% FA-water suspension) andSHM (42.6 mg/L according to 20 mg Fe/L as 2% SHM-water suspension) was added subse-quently in order to distinguish between arse-nic, uranium and radium immobilization to FA


158

and SHM. Samples were filtered by 0,2 µm cel-lulose acetate and acidified before AAS, ICP-OES as well as γ-spectrometry analysis.

2.5. Pilot test for removal of arsenic frommine waterThe schwertmannite produced in the pilotplant Tzschelln was used to enhance the arse-nic removal from mine water in the abandoneduranium mine of Schlema-Alberoda (East Ger -many). The uranium mine water has to be puri-fied from iron, arsenic, uranium and radium,iron and manganese. The installed treatmenttechnology is described by Meyer, et al. (2009).The average composition of mine water inflowis shown in table 2.

The required limit of arsenic for discharginginto the river Mulde is ≤ 0.1-0.3 mg/L (depen-dent on flow rate of the Zwickauer Mulderiver). The naturally occurring iron concentra-tion is not sufficient for complete arsenicremoval or to meet the prescribed maximum

value. Thus it was investigated, whether it ispossible to enhance the arsenic removal bydosing SHM as a diluted aqueous suspension. The pilot plant was constructed in a building ofthe mine water treatment plant of the Wismutat Schlema and consists of a precipitation tank,a flocculation tank and an inclined clarifier. Aflow chart of the pilot plant is shown in figure1. A data logger records the pH in the precipi-tation tank and enables adjustment of pH bycontrolling a peristaltic pump to dose limemilk. Table 3 summarized main parameters ofthe last seven pilot campaigns.

The water to be treated in the pilot plant istaken from the inflow of the mine water treat-ment plant after the acidification and CO2

stripping stage.Until now three schwertmannite samples wereapplied: two wet filtered aggregate samplesand one dried sample from the pilot plant atTzschelln. The schwertmannite is suspended inprocess water and stored in a tank with agita-

Table 2: Selected parameters of the Schlema-Alberoda treatment plant’s inflow (May 2010)

Table 3: Overview of previously performed pilot test campaigns


159

tor. In the first pilot experiments the schwert-mannite was pre-hydrolysized by adding limemilk to the storage tank until pH 8. Schwert -mannite suspension with 5 wt% of solids ledto plugging of the dosing tube. Dilution to 1wt% solved the problem.Lime milk and a flocculant are taken from thestorage tanks of the mine water treatmentplant. With a peristaltic pump the schwert-mannite suspension is dosed into the precipi-tation tank. Lime milk is dosed simultanouslyto adjust pH to 7.5. Subsequently the waterflows to the flocculation tank where the poly-mer flocculant is dosed. In this tank precipita-tes are transformed to flocks. In the followinginclined clarifier these flocks are sedimented.The overflow of the clarifier goes back into themain treatment plant.To monitor the treatment process inflow andoutflow of the pilot plant were analyzed fortotal arsenic, iron and uranium. In the outflow,dissolved concentrations were determined,too. The sludge from the clarifier was analysedfor uranium and arsenic and investigated by asedimentation test.

3. Results and Discussion

3.1. Microbial Investigations

3.1.1. Bacterial activity and diversity in theschwertmannite deposits

For a stable process of schwertmannite gene-ration the number and activity of iron oxidizingbacteria play a key role in the pilot plant.Because bacteria are present in the water aswell as in the schwertmannite that is depositedon the carrier material a recirculation of preci-pitated schwertmannite could stabilize theprocess and increase the oxidation rate if thebacterial cell numbers are high in the mineralphase.To obtain detailed information about the acti-vity and diversity of bacteria in the schwert-mannite a depth profile was prepared and themicrobial cells were analysed with a BacterialViability Kit and FISH analyses. A 4-week old 2cm thick sample core was cut into 4 layers,where layer 0-0.5 cm represented the youn-gest (water facing) layer. The first data indicate a decreasing total cellnumber with increasing depth (Fig.2). 3x106

cells were detected in the first layer (0-0.5 cm)and 1.4x106 cells in the deepest layer (1.5-2.0cm). More than 40 % of the cells in the firstcentimetre were found to be viable. In the dee-pest layer more than 30% of the cells were stillalive. This trend could be observed in materialfrom different water depth.

To obtain information about the microbialcommunity structure in the different layers thesamples were investigated with FISH probes forFerrovum myxofaciens and Gallionella TrefC4.Figure 3 shows the dominance of Ferrovummyxofaciens which reaches 25 to 36 % of the

Figure 1: Flow chart of themine water treatmentpilot plant


160

microbial community. In contrast to this appro-ximately only 1% of the microbial cells weredetected as Gallionella TrefC4 cells.Previous investigations of the microbial com-munity from water of the pilot plant demon-strated, that relatives of Ferrovum myxofaciensand Gallionella are the main species with apercentage of the total microbial communityup to 90% (Heinzel et al. 2009).The differen-ces may be due to the use of differentmethods for the analyses. For example T-RFLP,used by Heinzel et al. (2009), analyses the rela-tive composition of PCR-amplified 16S rRNAgene fragments. In contrast to this, the micro-scopic FISH approach to quantify cells by fluo-rescent DNA probe is a direct method that avo-ids the PCR amplification. PCR amplificationhas been shown to produce artifacts in micro-bial diversity (Acinas et al. 2005).This problem of comparison between FISH-based and PCR-based analyses is also suppor-

ted by our initial results from first TRFLP analy-ses of the same samples that indicate thatFerrovum myxofaciens comprises up to 90%of the total bacterial cells (data not shown).More analyses have to been done to quantifythe microbial diversity and activity and to givea suggestion which part of the schwertmanni-te deposit should recirculate.

3.1.2 Biologic Analysis / Interface Bacteria –MineralFerris et al. (2004) put forward the theory thatschwertmannite hedge hogs (globular aggre-gates with concentrically surrounded byschwertmannite whiskers) are overgrown bac-teria. Our TEM analyses of microtome sectionsshowed that these hedge hogs are in fact mas-sive and that no sign of overgrown cells are ininside these aggregations (Fig. 4). Furthermoreno direct connection of bacteria and mineralcould be found. Instead traces of EPS (extra-

Figure 2: A cell number of total, living and dead cell per mineral layer, B relative abundance of living and dead cells incorrelation to the depth of schwertmannite deposit on carrier material

Figure 3: relative abundance of Ferrovummyxofaciens and Gallionella TrefC4 in correla-tion to the depth of schwertmannite depositon carrier material

A B


161

cellular polymeric substance) are present in thesamples that connect mineral and cell. Weassume that the EPS is a seed for the precipita-tion of schwertmannite to prevent lethal cellovergrowth. This hypothesis is underlined bythe results of section 3.3.2. where we showthat schwertmannite whiskers form if the oxi-dation of iron happens slowly, irrespective of abiotic or abiotic environment.

3.2. Schwertmannite Generation: Estima -tion of the oxidation capacity The microbially enhanced oxidation process inthe pilot plant occurs both on the surface ofthe biofilm carriers and in the free volume ofthe oxidation basin. The oxidation rate is in flu -enced by various parameters, e.g. oxygen sup-ply, pH, temperature, sludge circulation, com-position of the inflowing mine water and surfa-

ce area of the biofilm carrier. At technical scaleonly the sludge circulation and the a mounts ofcarriers can be used to accelerate the oxidationrate. So in a first step the oxidation rate wasdetermined in the pilot plant at different surfa-ce areas of the growth carrier. In these experi-ments the oxidation rate was calculated accor-ding to Eq. (3) from the total amount of ferriciron (Fe(III)ges) formed during microbial oxida-tion and the residence time τ. The results arepresented in figure 5. By drawing the oxidationrate as a function of AF /VR ratio it was possibleto distinguish νOx-F and νOx-V and to estimatethese values.

Fig. 6 shows the oxidation rate of the pilotplant according eq. (1). Oxidation rates in therange of 10 – 40 g/(m3h) were achieved. The o -verall value was in the range of 20 – 25 g/(m3h).

Figure 4: TEM images from cultures of Leptospirillum ferrooxidans. A Two cells that are surrounded by EPS. The schwertmannite whiskers in that sample are not connected directly to the cell but to the EPS. B Sections through schwertmannite hedge hogs. They are of massive nature and no signs of an overgrown cellcan be found in the center

(Eq.3)


162

The influence of the flow-rate on the oxidationperformance of the pilot plant was discussedin detail in the Science Report 2009 (Peiffer etal., 2009).

3.3. Mineralogical and Structural Analysis

3.3.1. Schwertmannite structureThe morphology of the schwertmannite de -pends on the way of synthesis. Samples that

were synthesized with H2O2 consist of roundspheres whereas samples that were producedby bacterial oxidation showed the characteri-stic whiskers. Samples synthesized in an inor-ganic but slow way showed a third variety.Exemplary SEM pictures are shown in Fig. 7.Samples synthetised by a slow oxidation withdiluted H2O2 also showed whiskers indicatingthat the whiskers form in environments withslower oxidation processes and not specifically

Figure 6: Oxidation rate and throughput of the pilot plant at Tzschelln until April 2010

Figure 7: Left: Exemplary SEM pictures of schwertmannite precipitated by different methods. The typical schwertmanni-te hedge hog morphology is present in the biogenic schwertmannite (produced in the pilot plant) and the schwertman-nite synthesized abiotically at 85 °C. Right: diffractograms of synthesized samples. All show the typical schwertmannitereflections, however, the products differ regarding the intensity ratio between several peaks (e.g. 1.5 Å vs. 1.45 Å,intensity of shoulder at 2.3 Å) as well as regarding the peak width.


163

due to biotic/organic influence. The XRD mea-surement on the 2D detector allowed preciseanalyses of the different synthesized schwert-mannites in finite time. Even though the speci-mens show similar diffraction patterns thereare distinct variations in intensity and peakwidth. Also additional reflections appear insome samples (Fig. 1). This indicates that the»schwertmannites« are not one homogeneousphase but a mixture. This raises the demandfor a reproducible method to synthesize pureschwertmannite.

3.3.2. XASThe evaluation of XAS data is ongoing work.Preliminary results on the As K-edge spectra ofarsenic adsorbed to lepidocrocite and goethite

are essentially identical independent of As con-centration and of the adsorbant. Also the AsK-edge spectra of schwertmannite are identi-cal for different concentrations of As. Anyhow,a comparison of the spectra and the radialdistribution function of both materials, thestandards and the co-precipitated schwert-mannites, shows that arsenate absorbed inschwertmannite is bonded more closely, andtherefore more strongly, to the iron than thearsenate adsorbed on goethite and lepidocro-cite surfaces (Fig. 8).

3.3.3. Results DFT: Bulk FeOOH polymorphsand bulk schwertmanniteOur work on the bulk properties of the FeOOHpolymorphs goethite (a), akaganeite (b), lepi-

Figure 8: Representative examples of As K-edge EXAFS spectra of representative examples of aresenate adsorbed ongoethite (grey) and incorporated in schwertmannite (black). The left graph shows a magnification of the XANESregion. Please note the differences between goethite and schwertmannite in the marked box. The graphs on the rightshow the radial distribution function. Whereas the first coordination sphere (As-O) is the same in both samples thesecond (shoulder between 2 and 3 Å) and third (between 3 and 4 Å) coordination sphere is closer to the central atomindicating a stronger bond.

Figure 9: Akaganeite with interstitial ions


164

docrocite (γ), and the high pressure phase (e)was recently published in Physical Review B(Otte 2009). The energetic relations among thephases reveal that the framework structures(a, b, e) are more favorable than the layeredone (γ ). The schwertmannite structure is con-sidered akin to akaganeite where additionalinterstitial sulfate ions are added in the chan-nels, see Fig. 9. Because schwertmannite issynthesized in Cl- containing solution, we con-sider as a starting point Cl atoms in the akag-aneite channels. We find that the Cl-ionswiden the channels. As Cl is singly negativelycharged, we have explored the role of intro-ducing a background charge of the wholesimulation cell to model correctly the electro-nic properties.

Furthermore, we explore the bonding mecha-nisms of SO4

2-. We find that the channelsadapt to the ions in the middle of the chan-nels. However, this adjustment of the frame-work is connected with a high energy cost. Thesorption energy of the system can be determi-

ned per formula unit and unit cell (f.u./u.c.) as

(Equ.4)

where negative values correspond to an exo-thermic reaction. The absolute value sorptionenergy decreases with increasing sulfate con-centration: The lowest absolute value ofESorption =-0.092 eV is obtained fort the highestsulfate concentration (every akaganeite chan-nel is occupied by sulfate ions) where thestrongest distortion of the akaganeite frame-work occurs (table 4). The sum formula ofschwertmannite Fe16O16(OH)16-2n(SO4)n(n=1-3) according to (Bigham 1994) allows foran even higher sulfate concentration inexchange with hydroxyl groups. Therefore,further bonding mechanisms (mono- andbidentate) are current ly under investigation.A further task under investigation is the sorp-tion of As(V). As an even larger molecule, it isfound to cause a stronger distortion of the fra-mework. The sulfate-arsenate exchange is afurther aspect we are currently working on.

Sorption on the FeOOH surfacesThe FeOOH polymorphs have high surfaceareas and high sorption affinities for aqueoussolutes (Cornell 2001). The transformation pro-duct after arsenate incorporation in schwert -man nite is expected to be a FeOOH polymorph.Various terminations of the most prominent andrelevant surfaces in sorption reactions, goethi-te(101), akaganeite(100), and lepidocrocite(010), are modeled. Prior to studying adsorp-

Figure 10: Density of states (a-c) and spin density plot (d) of b-FeOOH (100)-surface

Table 4:


165

tion processes it is important to understand theproperties of the clean surfaces. Oxygen, hydro-xyl and water terminations are considered tomodel different pH conditions. Indeed, we find that the surface terminationimpacts strongly the oxidation state of the Feat the surface as displayed in Fig. 10 for akag-aneite(100). The deprotonated oxygen termina-tion (basic condition, Fig. 10a) leads to a reduc-tion of the surface iron, FeSF, and the respecti-ve magnetic moments, MFe, are reduced. Thehydroxyl termination with all iron in Fe3+ and aband gap of 1.7eV (Fig. 10b) is closest to bulkbehavior. For a water termination which repre-sents an acid environment, (Fig. 10c), the FeSF

are reduced to Fe2+ and a breaking of FeO6

octahedra at the surface occurs. The spin density plot (Fig. 10d) integratedaround the Fermi level for the water termina-tion reveals that the Fe1SF is reduced with a 6th

partially occupied dxy orbital, while for Fe2SF theone spin channel is fully and a 6th electronappears in the opposite spin channel with par-tially filled eg-states at the Fermi level. Similarproperties are obtained for goethite(101) andlepidocrocite(010), but are not displayed here.Ongoing work focuses on the adsorption ofarsenate on these relaxed surface terminations.

3.4. Schwertmannite Transformation andArsenic Adsorption

3.4.1. Transformation of schwertmanniteLong term exposure of schwertmannite to pHvalues between 5 and 8 for 4 months at nor-mal atmospheric condition leads a sulfaterelease, which increased with time and pH upto roughly 1.40 mmol/g solid phase, corre-sponding to 90 % of total initial sulfate, after4 months of exposure at pH 8 (Fig.11). Thehigher SO4

2- release at elevated pH is probablydue to the direct exchange of OH- for theSO4

2- ion at the schwertmannite surface, whilefaster SO4

2- release during initial days may bedue to release of surface-adsorbed SO4

2-.

3.4.2. As(III)-schwertmannite interactionsA detailed understanding on As(III)-schwert-mannite interaction in acidic medium whichwill foster attempts to remove As(III) with thismineral was obtained. Schwertmannites tur-ned out to be able to remove significant frac-tions of As(III) (> 97 %) from contaminatedwater, the efficiency of which strongly dependson surface area and synthesis pathway. Thedifferent synthesis products, produced throughbiotical and abiotical synthesis pathways differin their physical and chemical properties inclu-ding morphology, dissolution rate, surface areaand crystallinity which strongly affects the

Figure 11: PHREEQC modelled (lines) SO42- adsorption profile by schwertmannite at diffe-

rent pH va lues. The experimentally observed values at pH 5, 6, 7 and 8 after 120 days areshown as filled circles.


166

sorption behaviour of As(III) (Fig.12).Our findings suggests the occurrence of tworemoval mechanisms: 1) ligand exchange withsurface adsorbed SO4

2- which appears to be ofminor relevance as indicated by the absence ofany particular trend in SO4

2- release in case ofSHM_FS, and the weak SO4

2- release in case ofSHM_DS inspite the highest As(III) uptake andSSA of the three specimens and 2) formationof amorphous As(III)-Fe(III)-SO4

2- precipitates athigher aqueous As concentrations as indicated

by increased diffraction intensity proportional-ly with the As(III) loading and the appearance ofnew IR bands. These amorphous surface preci-pitates may stabilize sorbed As(III) towards sur-face oxidation. Most of the aqueous As(III) ap -peared to sorb through the second mechanism.

3.4.3. Dumping experiments with As-loadedSHM – samplesThe relative amounts (%) of mobilized arsenicby periodical eluation with rain water (pH 5, B-

Figure12: Freundlich (a) andLangmuir (b) sorption model fits tothe experimental data. Solid linesrepre-sent model predicted sorbedAs(III), while experimental data areshown by filled and open symbols.

Table 5: Arsenic mobilization under different dumping scenarios


167

tets) or continuous elution with groundwater(pH 7) were summarized in table 5. Actually(after 140 days), a significant arsenic mobiliza-tion was only detectable in test 3B, a periodi-cal elution of As-loaded Schwertmannit withrain water.

3.4.4. Laboratory experiments for pilot testpreparationFig. 13 shows the promising results of the firstbatch tests, where brown coal filter ash (FA) wasused instead of lime milk for the neutralisation ofthe acidic process water of the water treatmentplant in Schlema-Alberoda. Six hours after theaddition of 0.21 g/L (test 1) and 0.16 g/L (test 2)FA addition, 0.22 mg As/L and 0.96 mg U/L

could be immobilized in test 1 (pH6,13), 0.17mg As/L and 1.46 mg U/L in SHM/L (20 mg Fe/L)was added to each test. Six hours later, a total of0.49 mg As/L, 1.01 mg U/L and about 1000 mBq226Ra/L were eliminated in test 1 (pH 5.01), atotal of 0.64 mg As/L, 1.57 mg U/L and about930 mBq 226Ra/L in test 2 (pH 6.42).

3.5. Pilot test for removal of arsenic frommine waterAs mentioned above the required limit of Asfor discharging into the river Mulde is ≤ 0.1 0.3mg/L (dependent on flow rate of the Zwi ckau erMulde river). The naturally occurring iron con-centration is not sufficient for complete ar senicremoval or to meet the prescribed maximum

Figure 13: Arsenic and uranium elimination in batch tests, where FA was used as neutralizing agent

Figure 14: Comparison of As concentration of inflow andoutflow (pre-hydrolyzed SHM, 62 and 41 mg Fe/L)

Figure 15: Residual As concentration depending fromSHM dosage (pre-hydrolyzed SHM)


168

value. The main results of the pilot tests areshown in fig. 14 to 17. Fig. 14 shows the resultsof the first pilot experiment with pre-hydrolyzedSHM. Pre-hydrolysis was applied because labscale investigations of the revealed that theSHM transformation at high sulfate concentra-tions, as in the Schlema-Alberoda mine water, isdelayed. In order to directly compare the per-formance of the two SHM varieties in thethird pilot experiment untreated SHM wasapplied (fig. 16).

Surprisingly in the campaign with untreatedSHM at comparable dosage a better As elimi-nation was obtained. A dosage >30 mg Fe/Lpre-hydrolyzed SHM was necessary to reach

the limit for discharge of 0.1 mg As/L (Fig.15).With untreated SHM dosage of 21mg Fe/L wassufficient to reach this concentration. Fig 17shows that, compared to As removal by theiron content of the mine water alone, it waspossible to improve the arsenic removal from60 % to nearly 90 % by dosing 20-40 mg/L Fe,and 35-70 mg/L SHM (dry mass) respectively.Following to the promising results of the firstbatch tests with brown coal filter ash (FA) asneutralizing agent (chapter 3.4.4),pilot testcampaign 7 was performed with FA dosageinstead of lime milk. A suspension of 5 % FAand water of the river Mulde was used. Thisleads to an average, effective dosage of 0,15 gFA/L process water. The SHM-addition was

Figure 16: Comparison of As concentration of inflow andoutflow (untreated SHM, 21 mg Fe/L)

Figure 17: Enhancement of arsenic removal by dosage ofschwertmannite suspension

Figure 18:Arsenic andUranium elimi-nation duringthe pilot testcampaign 7


169

comparable with the foregoing campaigns(17-26 mg Fe/L). As shown in Fig.18, the requi-red limit for discharging into the river Muldecould always be hold for arsenic. The averageduranium elimination was 1,2 mg/L, probablydue to a precipitation of Calciumuranatetogether with eluated Calcium from the FA.

4. ReferencesAcinas, S. G., Sarma-Rupavtarm, R., Klepac-Ceraj, V. and Martin F. Polz (2005): PCR-InducedSequence Artifacts and Bias: Insights fromComparison of Two 16S rRNA Clone LibrariesConstructed from the Same Sample. Appl.Environ. Mircobiol 71:8966-8969.

Bigham et al., (1994) Mineral. Mag. 58: 641 Bigham, J. M., Schwertmann, U., Carlson, L.,Murad, E. (1990): A poorly crystallized oxyhy-droxysulfate of iron formed by bacterial oxida-tion of Fe(II) in acid mine water. Geochimica etCosmochimica Acta, 54, 2743-2758.

Brunauer, S., P.H. Emmett and E. Teller (1938):Adsorption of gases in multimolecular layers. J.Am. Chem. Soc. 60: 309-19

Cornell and Schwertmann, (2001): Wiley-VCHGmBH & Co. KGaA, Weinheim, ISBN 3-527-30274-3

Ferris F. G., Hallbeck, L., Kennedy, C.B.,Pedersen, K. (2004): Geochemistry of acidicRio Tinto headwaters and role of bacteria insolid phase metal partitioning. ChemicalGeology, 212, 291-300.

Glombitza, F., Janneck, E., Arnold, I., Rolland, W.,Uhlmann, W.(2007): Eisenhydroxisulfate aus derBergbauwasserbehandlung als Roh stoff. In:Heft 110 der Schriftenreihe der GDMB, S.31-40ISBN 3-935797-35-4

Hallberg, K.B., K. Coupland, S. Kimura andD.B. Johnson (2006): Macroscopic streamergrowths in acidic, metal-rich mine waters innorth Wales consist of novel and remarkablysimple bacterial communities. Appl. Environ.

Mircobiol. 72(3): 2022-2030

Heinzel, E., E. Janneck, F. Glombitza, M. Schlö -mann and J. Seifert (2009): Population dyna-mics of iron-oxidizing communities in pilotplants for the treatment of acid mine waters.Environ. Sci. Technol. 43(16): 6138-6144

Janneck, E., (2008) Abschlussbericht FKZ: 01RI05013; Teilprojekt 1: Koordination sowieAnlagenbetrieb und Produktherstellung

Loan, M., Cowley, J. M., Hart, R., Parkinson, G.M. (2004): Evidence on the structure of syn-thetic schwertmannite. American Mineralogist,89, 1735-1742.

Otte, K., Pentcheva, R., Schmahl, W. W. andRus tad, J. R.; (2009): Pressure induced structu-ral and electronic transitions in FeOOH fromfirst principles, Phys. Rev. B 85, 205116

Peiffer, S., Burghardt, D., Janneck, E., Pinka, J.,Schlömann, M., Wiacek, C., Seifert, J., Schmahl,W., Pentcheva, R., Meyer, J., Rolland, W. (2008)in: Geotechnologien Science Report No. 12,pp. 34-45; ISSN 1619-7399

Regenspurg, S., Peifer, S. (2005): Arsenate andchromate incorporation in schwertmannite.Applied Geochemistry, 20, 1226-1339.


170

MicroActiv - Optimization of water purificationtechnology for arsenic and antimony scaven-ging by microbially-activated Fe-oxide minerals

AbstractThe main research activity of this collaborativeresearch effort is the optimization of water tre-atment technology on basis of granulated Fehydroxides (= GFH, or in German GEH). GEH®

is applied by the SME-Partner, the GEH Was ser -chemie GmbH & Co. KG established in 1997and now one of the leading providers of iron-based high-capacity adsorbents for use in fixedbed filter process. Over 2000 plants have yetbeen delivered into more than 20 countriesworldwide. Main purpose is treatment of As-and Sb-tainted waters, providing e.g. in Indiamore than half million people with treated tapwater. GEH is based on pure synthetic ironhydroxide (b-FeOOH, akaganeite) with a largeeffective surface area (>250 m²/g) and anadsorption capacity of up to 50 g/kg As if app-lied on waters containing only As. This enablesa simple, low maintenance removal procedurewith capacities of up to 300,000 bed volumesover years without producing hazardous As-loaded sludge to be costly deposited. The mainpractical problem is the yet less well under-stood efficiency degradation with differentgroundwater milieus; with more or less redu-ced bed volume capacity. Our experimentsrevealed that these effects are mainly due tocompeting inorganic oxyanion species (e.g.,

silicate). An open question to be solved for theSME partner was whether a specific biogenicfunctionalization of the surface may help inoptimizing towards a more versatile and smoothfilter technology, but our preliminary experi-ment results were not thus promising towardsthis task. Instead, we found a severe inhomoge-neity in adsorption efficiency and competingeffects by inorganic oxyanion species on an indi-vidual grain scale which may become unexpec-ted and new key information how to improveon the GEH production process.

Introduction Access to safe drinking water is now a basichuman right and a critical component of effec-tive policy for health protection. Due to its hightoxicity at even low concentrations (in the µg/Lrange), arsenic is a cause for concern in manyregions of the world. Millions of people inSoutheast Asia are severely suffering from drin-king As-tainted tap water which has become anational health disaster in some of these coun-tries. Recently, arsenic contamination of shal-low aquifers have been reported, e.g., for thenorthwestern Hetao Plain of Inner Mongolia,China, an arid region with slow groundwaterflow and arsenic concentration up to 1000 µg/L.

Kersten M. (1)*, Bahr C. (5), Daus B. (4), Driehaus W. (5), Kappler A. (2), Karabacheva S. (1), Kolbe F. (4),

Posth N. (2), Reich T.Y. (1), Schurk K. (3), Stanjek H. (3), Wennrich R. (4)

(1) Geosciences Institute, Johannes Gutenberg-University Mainz, [email protected]

(2) Center for Applied Geosciences, University of Tübingen, [email protected]

(3) Ton- und Grenzflächenmineralogie, RWTH Aachen, [email protected]

(4) Helmholtz Centre for Environmental Research – UFZ Leipzig, [email protected]

(5) GEH Wasserchemie GmbH & Co. KG, Osnabrück, [email protected]



171

This is particularly important as the risk of con-tamination to human health will increasebecause of increasing groundwater demandupon decreasing quality and availability of sur-face water in such arid regions. According tothe World Health Organization the maximumconcentration of arsenic allowed in drinkingwater is 10 µg/L. This is because long term ex -posure to arsenic can affect human health andmental development of children, among otherpossible adverse effects. The skin seems to bequite susceptible to the effects of chronic ex -posure of As via drinking water. Arsenic-indu-ced skin lesions seem to be the most commonand initial symptoms of arsenicosis (»blackfooddisease«). Other manifestations include neuro-logical effects, obstetric problems, high bloodpressure, diabetes mellitus, diseases of therespiratory system and of blood vessels inclu-ding cardiovascular, and cancers typically invol-ving the skin, lung, and bladder (Rahman et al.2009). If the contaminated groundwater is alsoused for irrigation, a second health risk via thefood chain may be present. The global impactnow makes it a top priority water quality issue,second in order to microbiological contamina-tion. In SE Asia and China, many tube wellshave been drilled into aquifers to avoid for thelatter hazard, but as realized later at the priceof the former problem. The magnitude of thishuman tragedy increasingly encountered willdepend on the rate at which monitoring andmitigation programs can be implemented inthe affected regions. There are multiple factors that need to beaccessed before choosing a solution for a sui-

table mitigation strategy once arsenic contami-nated groundwater are found. Two categoriesof processes largely control arsenic mobility inaquifers: (1) solid-phase oxidation and dissolu-tion reactions, and (2) adsorption and desorp-tion reactions, mainly governed by the aquiferphysicochemical conditions. For an option tomitigate safe drinking water in case of alreadyinstalled tubewells, a variety of methods havebeen developed in the last years. Existing tech-nologies for the removal of arsenic from con-taminated water such as oxidation/precipita-tion, coagulation/coprecipitation, ion exchan-ge, adsorption, etc. are well established, andhave their merits and limitations. Efficientoptions are activated carbon and in particulariron oxide coated media (Schmidt et al. 2008,Kersten and Vlasova 2009a). Oxidation/ad sor -ption reactions are influenced by physicoche-mical milieu conditions such as pH and redoxstate, but also presence of competing anionssuch as silicate and phosphate, and solid phasestructural changes at the atomic level. Anunderstanding of factors controlling the arse-nic mobility and remediation options requiresknowledge of the natural groundwater charac-teristics as well as abiotic and biotic surfacereactions on a molecular or species scale. Anyoption has then to be evaluated and adoptedto the physicochemical characteristics of thelocal groundwater sources in order to enhancetheir efficiency in field practice and to buildtrust for their wider acceptance and use. Thishas to be done for arsenic, but also otheremerging oxyanion contaminants like antimo-nate and vanadate (Figure 1).

Figure 1: Major contaminants in groundwater used for producing tap water (modified after Von Gunten 2008).


172

Materials and HypothesesThe groundwater treatment by various ironoxide media filter techniques achieve capaci-ties which in practice depend on varying ope-ration conditions, e.g. intermittent operation,but also if not predominantly on quality varia-tions in the water treated. Breakthrough mayoccur much earlier than predicted, presumablyas an effect of interfering water constituents.During a long-lasting practice with naturalgroundwater, the SME partner GEH Was ser -chemie GmbH has observed complex competi-tion relationships, such as even an antagonisticeffect by Ca which may significantly mitigatethe competing effect by silicate (Smith andEdwards, 2005; Sperlich and Werner, 2005). Inprinciple, the adsorber columns are capable toprovide for very long service lives - up to300,000 bed volumes - before the maximumpermissible arsenic concentration of 10 µg/L isreached in the effluent stream. Evaluation oflaboratory and field monitoring data reveals,however, that actual live cycles can be as shortas 50,000 bed volumes at some locations with -out immediately evident causes. In most casesthis appears to be the result of a combinationof numerous factors detrimental to the As ad -sorption process (e.g. pH, silicate, water hard-ness, inappropriate operating conditions, Caconcentration, etc.) which are not yet under-stood in detail.It is of particular importance to clarify thedynamic behavior of silica species and theirdepositional behavior on the adsorbent media.Silica undergoes polymerization, precipitation,dissolution or adsorption depending on the pH

and/or temperature (Kersten and Vlasova2009b), which becomes greatly complicated inthe presence of other ions. A general para-digm is that the addition of electrolytes to asupersaturated solution accelerates the aggre-gation and precipitation of polymeric species.However, experimental results recently showedthat polysilicic acid in the presence of Ca ions(1 mM/L) is apparently stable in solution overextended time, compared with that under aCa-free condition (Chida et al. 2007). On theother hand, the concentration of soluble mo -no meric silicate in the presence of Ca ionsimmediately became metastable, that is, slight-ly higher than the solubility of silica. Its dyna-mic behavior was similar to that in the Ca-freecondition. One of our hypotheses is therefore,that the antagonistic behavior of increased Caconcentrations is based on its stabilization ofthe Si polymer in groundwater. Dissolved poly-meric Si is much less adsorbed and thus com-petitive than monomer silicic acid with respectto arsenate and other toxic oxyanions (Swed -lund and Webster 1999, Swedlund et al. 2009).The SME partner expects a major break-through by the scientific deliverables of thiscollaborative project with potential for optimi-zing their water treatment technology. Theproject is therefore aimed at resolving theseopen questions and hypotheses by a collabora-tive activity according to the complementarymethodology and expertise of the participants.These activities are namely concentrating onexperiments with water purification unitsbased on Fe oxide phases in laboratory co lum -ns and on a field scale with different ground

Table 1: Trace metal contents of used GEH adsorbent samples taken from the top of the filter bed (XRF measurements)


173

water regimes (WP 1, GEH WasserchemieGmbH), spectroscopy of the oxyanion sorbatesinvolved in the surface reactions on a molecu-lar scale (WP 2, Geoscience Institute of Univ.Mainz), characterization of adsorption reac-tions on microbiologically activated iron mine-ral surfaces (WP 3, ZAG at Univ. Tübingen),characterization of the GEH sorbent structureson a nanoscale by applying X-ray techniques(WP 4, RWTH Aachen), and adsorption batchexperiments and analytical characterization ofpotentially formed organic metalloid species(WP 5, UFZ-Helmholtz Leipzig).

For the investigations, both fresh and usedGEH adsorbent samples from several water-works were collected and analyzed. The tracemetal contents determined in the used GEHsamples were compared with those of freshGEH adsorbent taken from current production.The results from five German waterworks aresummarized in Table 1. They show high arsenicloading in the used GEH adsorbent materials,ranging from approx. 3 g/kg up to over 13 g/kg.These arsenic levels correlate with the As con-centrations in the inlet raw water (12 - 40 µg/LAs) and with the duration of operation of theadsorber system. The analyses also revealedhigh loading levels of phosphate and vana-dium along with high silicate loading, the lat-ter up to 6 wt.-%. The inlet raw water enteringthese waterworks typically contains 0.1 - 0.2mg/L PO4 and approx. 5 - 10 mg/L Si. Inlet con-centrations of both vanadium and antimonyare typically below 10 µg/L, e.g. 1.6 µg/L Sband 1.7 µg/L V at the »Boss« water works. Sin -ce only negligible antimony loading was foundin the used adsorbents, it can be concludedthat vanadium is much better adsorbed by theGEH adsorbent than is antimony. The highvanadium loading found confirms the highadsorption capacity for this element onto ironhydroxides as reported in literature (e.g.Naeem et al. 2007).The following further findings can be summari-zed from the used GEH bulk chemical analysis:– During operation, chloride and sulfate were

eluted by the raw water from the GEHmaterials. The content of both anions in the

adsorbent thereby decreased to approx. 0.5 % chloride and 0.1 % sulfate. New GEHadsorbent contains approx. 2 % chloride and0.6 % sulfate resulting from the manufactu-ring process. The chloride ion is an impor-tant constituent of the akaganeite crystalstructure.

– An increase in alkaline earth metal contentwas found in the used adsorbent, presumab-ly due to co-sorption of cations (Mg, Ca, Sr)or precipitation of insoluble compounds (Ba).

– Enrichment with other trace metals (e.g. Cuand Zn) in the used adsorbent was alsofound but only in some water works.

– The filter bed shows an inhomogeneouscontaminant loading distribution: At the topof the filter, the GEH material is found to behighly loaded whereas at the bottom of thefilter, the GEH adsorbent is less loaded.

Preliminary Results of the LaboratoryExperimentsWhile at room temperature, Mössbauer spec-tra indicate akaganeite as the major phase ofthe active GEH filter medium, a deviation in fitat 4.9 K revealed a more complex composition.Quantitative analysis of the spectra suggestsequal proportions in ferrihydrite (Fh) and akag-aneite (b-FeOOH) composition. Therefore theresearch project partner from RWTH Aachenperformed a thorough investigation with helpof X-ray diffraction analysis (XRD). To optimizethe XRD analysis of the GEH material, thepotential ferrihydrite content had to be quan-tified with a Rietveld program due to its lowcrystallinity. For this task, a synthetic 2-line fer-rihydrite was prepared as a standard and ana-lyzed by XRD. After subtracting the back -ground of the diffractometer, the net intensi-ties of the ferrihydrite were converted intoobserved structure factors by using a hexago-nal cell with a primitive space group. The peakswere refined with isotropic peak widths. Withthis new structure a better detection of 2-lineferrihydrite was possible. In five known mixtu-res of this ferrihydrite with hematite an excel-lent correlation of r2 = 1 was obtained (Figure2). While in previous investigations the akag-


174

aneite seemed to be a major phase of the bulkGEH, the new ferrihydrite structure resulted nowin a higher content of as much as 37 - 47 wt.-%of Fh in fresh GEH. These amounts correspondbetter to the Mössbauer data provided by theresearch partner ZAG Tübingen. More over, theferrihydrite in GEH seems to differ from thesynthetic 2-line ferrihydrite. It appears to be aneven poorer crystallized Fe-oxihydroxide pha -se because the significant two lines of ferrihy-drite are not clearly visible in the XRD patterns.Further investigations on this issue are in pro-gress to refine the quantitative XRD analysis.

The fresh GEH material was tested for theremoval of inorganic and methylated arsenicand antimony species from water by the pro-ject partner UFZ Leipzig. Batch experimentswith GEH suspensions (2 g/L) were done witharsenate (As(V)), arsenite (As(III)), dimethylarsenic acid (DMA(V)), antimonate (Sb(V)), andtrimethyl antimonate (TMSb(V)). The ionicstrength was adjusted with sodium nitrate tofinal concentrations of 10 mM, 50 mM, and100 mM, respectively. The samples were sha-ken for 24 h at 150 rpm at room temperature,which was found previously to be sufficient forequilibration (Daus et al., 2004). The total con-

Figure 2: Comparing weigh-ted-in with BGMNfitted Fh content.The deviations fromthe regression lineare less than 1%absolute.

Figure 3: Sorption loadings of DMA(V) onto GEH material; variation of initial concentrations (5.7 and12.0 mg/L), ionic strengths (10, 50, and 100 mM) and pH.


175

centrations of antimony and arsenic in thesupernatant liquids were analysed by ICP-MS.As an example, the results of DMA(V) are sum-marized in Figure 3.

The adsorption efficiency for DMA(V) decrea-ses at pH values higher than 6. This effect wasobserved for the adsorption on both goethiteand ferrihydrite in other studies (Lafferty et al.,2005). The influence of the ionic strength isnegligible. These experimental data are provi-ded by the UFZ group for modelling to be doneby the University of Mainz partner. Anotherimportant factor influencing the performanceof the water treatment filters are competitiveadsorption reactions by other water consti-tuents. This effect is well known for phospha-te which can compete for sorption sites duringthe adsorption process as well as remobilizealready adsorbed arsenate (Zhang et al. 2007).The used water works material contains notonly arsenic but also vanadium (Table 1). More -over, recent literature data indicate effects ofvanadium on As adsorption efficiency by GEHadsorber (Speitel et al. 2010). There fore, vana-dium was included in competition experimentsbesides phosphate, antimonate, arsenite, andarsenate. For the competition experiments, theconcentration of arsenate, antimonite, and va -n adate as primary oxyanion was chosen to be10 mg/L. The correspondent sorption capacityfor arsenate at this initial concentration is ap -proximately 5 g/kg, which is a typical loadingin water works material after 2 years in use.The concentrations of the competing oxyanions(arsenate, arsenite, antimonate, phospha te, or

vanadate) were varied from 1 - 100 mg/L. Avolume of 50 mL of the mixed solution wasadded to 0.1 g GEH material (2 g/L suspension).The material was shaken under identical condi-tions such as pH = 7. In the supernatant aftercentrifugation (3000 rpm, 10 min), arsenic,antimony, and vanadium concentrations wereanalysed.

The results shown in Figure 4 revealed that thesorption of antimonate onto GEH was mostaffected by the other oxyanions. Its sorptioncapacity decreased in presence of vanadate (a -bout 9 %) and even more in the presence ofphosphate, arsenite, and arsenate (up to 30 %)for a ratio of competitive oxyanion/antimonateof 10:1. This ratio is quite realistic for rawwaters of water works as found at the test site.Therefore, antimonate is apparently most sen-sitive for sorption disturbance by other waterconstituents. Vanadate as primary oxyanionseems to be relative robust to competitivereactions in the investigated concentrationratios. The sorption of arsenate was less sensi-tive than the sorption of antimonate, but vana-date can decrease the capacity by 23 % at asurplus of factor 10. The adsorption affinitytowards GEH therefore increases in the orderantimonate – arsenate – vanadate. Therefore,not only silicate but also vanadate concentra-tion in the raw water should be taken intoaccount to predict a premature break-throughof arsenate.

Additional experiments on adsorption kineticsof silicate onto GEH were performed in a

Figure 4: Overview of competition experiments performed at pH 7 and 24 h shaking.


176

mixed-flow reactor (MFR) by the project part-ner at RWTH Aachen. The material in the reac-tor was flushed by a 0.05 M KCl solution. ThepH was adjusted to different values in varioussorption cycles to observe the influence of pHonto the sorption behaviour. After equilibration,solutions with selected inlet concentrations ofsilicate were pumped through the MFR by aHPLC pump which ensured exact flow rates.Silicate was measured photometrically as mono-meric Si with the molybdenum-blue methodafter Köster (1979). From the balance of silicatein the input output solutions, sorption rateswere calculated (Figure 5). Higher sorption rateswere observed at lower the pH value.

Rapid small scale column tests (RSSCT method:Westerhoff et al. 2005) were carried out toinvestigate the breakthrough behavior of thecompeting anions As, V, phosphate and silica-te at ambient concentrations. Unlike the com-petitive adsorption in batch experiments des-cribed above, these tests aimed at investiga-tion of the dynamic adsorption properties infilter columns. The RSSCT breakthrough curvesshow concentration in the effluent as a func-tion of water throughput expressed in bedvolumes. Ideally, the RSSCT results are directlyapplicable to the corresponding full-scale GEHadsorber beds at the water works. An artifici-ally formulated ground water was used, con-

taining As, V, phosphate, and silicate in a typi-cal background water matrix (pH 6.8; I = 270µS/cm). GEH adsorbent (grain size: 63-75 µm)was filled into the filter column, providing abed volume of 5.3 cm3. The RSSCT data indi-cates an arsenic breakthrough at 80,000 BVrelative to the threshold value of 10 µg/L As(Figure 6). The arsenic adsorption capacity forthis particular water is significantly lower thanthose determined from typical breakthroughcurves for water without competing anions.Vanadium shows a similar breakthrough beha-vior, although the breakthrough curve is shif-ted significantly to higher BV. Here, theeffluent V concentration exceeds the 10 µg/Lvalue at approx. 140,000 BV. Considering thefact that the V concentration in the influent isapprox. half of the As concentration, it can beconcluded that adsorption capacity of GEHadsorbent is similar for both elements. Ana -lyses showed loading levels in the used GEHadsorbent of 2.8 g/kg As and 2.0 g/kg V (dryweight). Moreover, the breakthrough curvesshow several peaks which correlate with tem-porary pH value variations in the inlet rawwater (increases from pH 6.8 up to 7.2). Thisreflects the strong influence of pH on arsenicadsorption by GEH absorbent. The same effectis also observed for the vanadium curve, albeitless pronounced.

Figure 5: Sorption rates of Si onto GEH by variation of pH.


177

As detailed already in the last report (Kerstenet al. 2009), Fe(II)-oxidizing bacteria werefound on the 2-year old GEH filter material.Initial experiments with the related bacteriastem Acidovorax sp. BoFeN1 showed efficientremoval of As(V) and As(III) by co-precipitationof the arsenic with biogenic Fe(III) hydroxidesformed by these bacteria (Hohmann et al.2010). For this reason, laboratory experimentswere carried out to first determine whichminerals and cell-mineral aggregates are pro-duced by Fe(II)-oxidizing bacteria (Posth et al.2010). In a second step, the influence of Fe(II)-oxidizing bacteria on sorption and co-precipi-tation of arsenic was investigated. For this,sorp tion and co-precipitation batch experi-ments with As(V) and As(III), and both biogenicand abiogenic Fe(III) minerals were set up. Thebiogenic minerals were produced by Acido vo -

rax sp. BoFeN1, an anaerobic, nitrate-reducingFe(II)-oxidizing bacteria found in arsenic conta-minated aquifers. Freshly prepared GEH mate-rial was used for the abiogenic substrate. Theminerals were suspended in carbonate-buffe-red (pH 7) water with a low phosphate con-centration (<10 µM) and 8 - 10 mM dissolvedFe(II). Experiments with GEH contained 250mg solid material/25 mL water (10 g/L suspen-sion or about 2800 m2/L surface). Arsenite orarsenate was added at a final concentration of0.025 - 10 mM. All sorption experiments weresampled at the start and after 48 h of equilibra-tion time. Co-precipitation experiments weresampled at the start and at the end of Fe(II) oxi-dation. The dissolved As(V) and As(III) concen-trations of the samples were quantified by thecollaboration partner at UFZ-Leipzig. On a massbasis, more As(V) and As(III) sorbed per g Fe to

Figure 6: Arsenic andvanadium breakthroughcurves, determined inrapid small scale columntesting (RSSCT) using amodel ground watermatrix

Figure 7: Sorption of arsenic to GEH material, with (A) sorption of As(V), and (B) sorption of As(III).


178

GEH material than to the biogenic minerals(Figures 7 and 8). Nearly all As(V) up to an initi-al As concentration of 10 mM (98.3 - 100.0 %of arsenate added) was immobilized by GEHvia sorption, yielding up to 2 mM As(V) sorbedper g Fe in GEH (Fig. 7). Likewise, As(III) is near-ly entirely sorbed to GEH material (94.8 - 100.0 % of the amount added yielding up to1.5 mM As(III) per gram Fe).

Nonetheless, in GEH filter systems, in whichnitrate-dependent Fe(II)-oxidizing bacteriahave been found, sorption to GEH, sorption tobiogenic minerals, and co-precipitation with

biogenic minerals are all occurring simultane-ously. To approach the complexity of thesesorption continuum systems, step-wise experi-ments were set up. First, the sorption of As(V)to a mixture of GEH and biogenic minerals(biogenic minerals first precipitated thenadded to the GEH material) revealed that 98.1– 100 % of the added As(V) was removed (Fig.9). In this experiment, sorption to the GEH islikely the more important mechanism, and theresults are similar to those of sorption of As(V)to GEH alone (up to 2 mM As(V) sorbed per gFe). Interestingly, in case of precipitation of thebiogenic minerals in the presence of the GEH

Figure 8: Sorption and co-precipitation of arsenic to and with biogenic minerals. A) sorptionof As(III) to biogenic minerals, B) sorption of As(V) to biogenic minerals, C) co-precipitation ofAs(III) with biogenic minerals, D) co-precipitation of As(V) with biogenic minerals.

Figure 9: Sorption of arsenic to a mixture of GEH material and biogenic minerals. A) As(V) sorption toGEH material and biogenic minerals (biogenic minerals first precipitated and then added to GEHmaterial before As(V) addition). B) As(III) sorption to GEH material and biogenic minerals (biogenicminerals precipitated in the presence of GEH material before As(III) addition).


179

material, only 39.4 % of As(III) (at the highestAs(III) concentration) and up to 100% As(III) atlower As(III) concentrations were removedfrom solution indicating that the biogenicminerals precipitated in the presence of theGEH material are obviously coating the GEHmaterial and thus lowering its sorption capaci-ty or even preventing sorption of As to theGEH material. Additionally, the cell organicmatter may block sorption cites in the biogenicminerals.

In a final experimental set-up, the processes ofsorption and co-precipitation are occurringsimultaneously as would be the case in a GEHfilter system with Fe(II)-oxidizing bacteria. As aresult, 96.4 - 99.9 % of As(V) and 97.1 - 99.9% of As(III) were removed, corresponding to aload of up to 1.8 mM As(V) per g Fe andapproximately 1.5 mM As(III) per g Fe. The mix-ture is still very efficient (Figure 10), but itseems that the participation of microorganismsthat form biogenic minerals again reduces theefficiency of the GEH material. Compared tothe abiotic systems where As sorbs to GEH,this system removes less arsenic. Our prelimi-nary hypothesis is again that the cell organicmatter in the biogenic systems probably blockssorption sites or changes the surface chargesfrom positive to negative. In order to investiga-te the influence of organic compounds in thesorption capacity and co-precipitation capacityof As species to GEH and biogenic minerals,further experiments are warranted which focuson the influence of humic substances on sorp-tion capacity of the GEH material. These expe-

riments, as well as column experiments withbiogenic minerals and GEH filter material, arecurrently in progress. As a preliminary conclu-sion, however, it can be stated from all thoseresults that microbiological activation is not asuitable way to enhance the performance ofthe GEH material as originally expected.

While all XRF and XRD measurements, adsorp-tion experiments, and microbiological investi-gations yet mentioned were based on bulkGEH material, contaminant load has beenmeasured by microprobe analysis also on indi-vidual used filter grains by the project partnerat Mainz University. This analysis revealed thatthere is a strong inhomogeneity between indi-vidual grains in scavenging efficiency for thetoxic elements As and V. There are nearly inac-tive GEH grains with almost no As and V sca-venged (e.g., the grain in the upper left cornerin Figure 11), and very active grains with a highload in both elements (central grain in Figure11) but in the same sample batch from the tophorizon of a filter bed used for 2 years at thewater works station »Schö« (Table 1). The ratiobetween active and inactive grains is about 1:1as indicated by a scatterplot of As vs. Si forabout 200 analyzed grains (Figure 12). Boththe inactive and active grains have a similar Siload but with extremely different effect ontheir As/V affinity. Differing phase composition(akaganeite/ferrihydrite) as indicated by diffe-rent Cl concentrations between both grain po -pulations could be a probable reason. Thiswould also explain the difficulties and devia-tions in receiving representative bulk phase

Figure 10: Sorption and co-precipitation of arsenic in a mixed system of GEH and bacteria that produ-ce biogenic minerals. A) As(V) in the mixed system B) As(III) in the mixed system.


180

composition analysis as mentioned above.Verification of this hypothesis by micro-spec-troscopic phase analysis on the individual grainscale could potentially become new and unex-pected key information how to improve theGEH production process.

LiteratureChida T., Niibori Y., Tochiyama O., Mimura H.,Tanaka K. (2007): Deposition rates of polysilicicacid with up to 10-3 M calcium ions. Appl.Geochem. 22, 2810–2816.

Daus B., Wennrich R., Weiss H. (2004): Sor p tionmaterials for arsenic removal from water: acomparative study. Water Res. 38, 2948-2954.

Hohmann C., Winkler E., Morin G., Kappler A.(2010): Anaerobic Fe(II)-oxidizing bacteria show

As resistance and co-precipitate As duringFe(III) mineral precipitation. EnvironmentalScience and Technology, 44, 94-101.

Kersten M., Vlasova N. (2009a): Arsenite ad -sorp tion on goethite at elevated temperatures.Appl. Geochem. 24, 32-43.

Kersten M., Vlasova N. (2009b): Silicate ad -sorp tion by goethite at elevated temperatures.Chem. Geol. 262, 372-379.

Kersten M., Vlasova N., Posth N., Kappler A.,Schurk K., Stanjek H., Daus B., Kolbe F., Wenn -rich R., Bahr C., Driehaus W. (2009): Micro Ac -tiv - Optimization of water purification techno-logy for arsenic and antimony scavenging bymicrobially-activated Fe-oxide minerals. Reportfor the 1st Statusseminar Bayreuth, 10 pp.

Figure 11: Microprobe element distribution mapsfor two typical GEH grains sampled in the tophorizon of a filter used for 2 years at the waterworks station »Schö« (Table 1). The lighter arethe pixels (or more yellow-red in color), the hig-her was the element concentration.

Figure 12: Scatterplot of As vs. Si for 200 analy-zed grains sampled in the top horizon of a filterused for 2 years at the water works station »Schö«(arbitrary units of the qualitative microprobe ana-lysis). Note the steep slope of the negative As vs.Si regression curve in between 3.5 and 5.5 Siunits, and another less steep but with low to vir-tually zero As concentrations in between 3.5 and6.5 Si units. The number of grains in both datapopulations is about equal.


181

Köster H.M. (1979): Die chemische Silikat ana -lyse - Spektralphotometrische, komplexometri-sche und flammenspektrometrische Analysen -methoden. Springer-Verlag, Berlin, Heidelberg,New York.

Lafferty B.J., Loeppert R.H. (2005): Methyl Ar se -nic Adsorption and Desorption Behavior on IronOxides. Environ. Sci. Technol. 39, 2120-2127.

Miot J., Benzerara K., Morin G., Kappler A.,Ber nard S., Obst M., Férard C., Skouri-Panet F.,Guigner J.M., Posth N., Galvez M., Brown G.E.Jr, Guyot F. (2009): Iron biomineralization byanaerobic neutrophilic iron-oxidizing bacteria.Geochim. Cosmochim. Acta 73, 696-711.

Posth N.R., Huelin S., Konhauser K.O., KapplerA. (2010): Size, density and mineralogy of cell-mineral aggregates formed during anoxygenicphototrophic Fe(II) oxidation. GeochimicaCosmochimica Acta, 74, 3476-3493.

Rahman M.M., Ng J.C., Naidu R. (2009):Chronic exposure of arsenic via drinking waterand its adverse health impacts on humans.Environ. Geochem. Health 31, 189-200.

Schmidt G.T., Vlasova N., Zuzaan D., KerstenM., Daus B. (2008): Adsorption mechanism ofarsenate by zirconyl-functionalized activatedcarbon. J. Colloid Interf. Sci. 317, 228-234.

Smith S., Edwards M (2005): The influence ofsilica and calcium on arsenate sorption tooxide surfaces. J. Water Suppl. AQUA 54, 201-211.Speitel G.E.Jr., Katz L.E., Chen C.-C., Stokes S.,Westerhoff P., Shafieian P. (2010): Surfacecomplexation and dynamic transport modelingof arsenic removal on adsorptive media. AWTPProgram Report, WERC Water Research Foun -dation, Denver CO, USA, 140 pp. (available at:http://www.arsenicpartners.org/3098.pdf).

Sperlich A, Werner A. (2005): Breakthrough be -haviour of granular ferric hydroxide (GFH) fixed–bed adsorption filters: Modeling and experi-mental approaches. Water Res. 39, 1190-1198.

Swedlund P.J., Webster J. (1999): Adsorptionand polymerisation of silicic acid on ferrihydri-te, and its effect on arsenic adsorption. WaterRes. 33, 3413-3422.

Swedlund P.J., Miskelly G.M., McQuillan A.J.(2009): An attenuated total reflectance IRstudy of silicic acid adsorbed onto a ferric oxy-hydroxides surface. Geochim. Cosmochim.Acta 73, 4199-4214.

Von Gunten U. (2008): Can the quality of drin-king water be taken for granted? EAWAGNews 65e, 4-7.

Zhang J.S., Stanforth R.S., Pehkonen S.O.(2007): Effect of replacing a hydroxyl groupwith a methyl group on arsenic (V) speciesadsorption on goethite (a-FeOOH). ColloidInterface Sci. 306, 16-21.


182

Microstructural Controls on MonosulfideWeathering and Heavy Metal Release (MIMOS)

1. Introduction and summaryThe iron monosulfides pyrrhotite (Fe1-xS) andmackinawite (tetragonal FeS) are both environ-mentally important phases linking the iron andsulfide elemental cycle. This report summarizesstudies that aim at understanding on the effectof microstructure, mineral chemistry and cry-stallography of these minerals on the reactivity(Pollok et al., 2008) and provides insights to -wards possible applications.Pyrrhotite comes a close second after pyriteamong the most abundant iron sulfides in cru-stal rocks and ore deposits including Ni-Cu, Pb-Zn, Au, and platinum-group elements (PGE).The weathering of pyrrhotites contributes tothe release of iron and other metals as well assulfate to surface waters as a part of the acidmine drainage problem. The relevance of thisreaction lies in the high reactivity of pyrrhotite(20 to 100 times faster compared to the disul-fide pyrite; Nicholson & Scharer, 1994 ) whichlimits the neu tralization of produced sulfidicacid by silicate weathering reactions. Further -more, many po ten tially important implicationsto the fields of geomagnetics, petrology, envi-ronmental mineralogy, and technical mineralprocessing are to be expected if detailed know -ledge about structures and phase relations isavailable. Lately, the flotation behavior of pyr-rhotite is attracting more at ten tion as this mine-ral is an important host for Ni and PGEs (Becker

et al., 2010; Ekmekçi et al., 2010).In the first section the coupling between cry-stallographic structure and composition of pyr-rhotite has been studied by the use of trans-mission electron microscopy (TEM) which hasbeen used for the first time to directly visualizedetails of the (non-integral) stacking sequencesusing superstructure diffraction spots. Theobservations result in a new and versatile des-cription of pyrrhotite superstructures whichprovide a basis to understand variation in phy-sical properties and reactivity of pyrrhotites. Inthe second section the relevance of differentsuperstructures in terms of reactivity is directlysubstantiated for the biologically-enhanced dis-solution of a pyrrhotite intergrowth. The diffe-rence in dissolution rates betwee the inter-grown phases is quantified by using topogra-phic data produced by confocal microscopy pro-viding a new view on the reactivity and weathe-ring patterns of this mineral. The third part dealswith the evolution of surface roughness for suchreacted and heterogeneous surfaces. To extendour possibilities to generate meaningful data forreactive surface area and, hence, dissolutionrates, a new program is introduced, whichallows a non-biased statistical evaluation of sur-face roughness distributions. The concept pres-ented here can be utilized for any kind of topo-graphic data and might be well applicable toother areas of surface science.

Pollok K. (1)*, Harries D. (1), Hopf J. (1,2), Etzel K. (1), Chust T. (1), Hochella M.F., Jr. (3), Hellige K. (4),

Peiffer S. (4), Langenhorst F. (1)

(1) Universität Bayreuth, Bayerisches Geoinstitut, D-95440 Bayreuth, [email protected]

(2) Friedrich-Schiller-Universität Jena, Institut für Mikrobiologie, Neugasse 25, D-07743 Jena

(3) Center for NanoBioEarth, Department of Geosciences, Virginia Polytechnic Institute and State University,

Blacksburg, Virginia, USA

(4) Universität Bayreuth, Lehrstuhl für Hydrologie, D-95440 Bayreuth



183

Mackinawite is a metastable, nanoparticularphase that forms by precipitation from aqu e -ous solutions at low temperatures. It is a majorcomponent of acid volatile sulfide (AVS) com-pounds, causing black-colored layers in recentanoxic sediments. Mackinawite is therefore animportant precursor of sedimentary pyrite andcan facilitate in-situ remediation of metals inconnection with (biogenic) sulfate reduction inacidic pit lakes and sediments (Bilek, 2006,Church et al. 2007). It is known as an effectiveabsorber due to its large specific surface area(Ohfuji & Rickard, 2006) but detailed reactionmechanisms important to biogeochemistry arestill missing.The last section provides experimental eviden-ce for the formation mechanisms of sedimen-tary sulfides via mackinawite under anoxicconditions. The results highlight the importan-ce of iron hydroxide surfaces as the locationsof electron transfer for mackinawite forma-tion, but also show the spatial decoupling topyrite formation. In conclusion, the study pro-vides fundamental data that enables to modelbasic reactions at geochemical interfaces.

2. Towards a new structural model forNC-pyrrhotites: Structural and chemicalcharacterization of selected samples byelectron microscopyThe physical properties of a mineral and thereactivity of a mineral surface are determinedby its chemical composition and internal ato-mic structure. In this respect, non-stoichiome-tric pyrrhotite is a striking phase because itsstructure and the physical (e.g. magnetic) pro-perties change dramatically within a small com -positional range. A detailed knowledge aboutthe composition-structure relationship as wellas the structure-property relationship is nee-ded to gain a fundamental understanding ofthis environmentally and technically impor-tant mineral. Pyrrhotite has a NiAs-derivative structure thatcontains sequences of hexagonal close-packedS atom layers stacked in alternation with layersof Fe atoms in octahedral coordination. Non-stoichiometric compositions arise from the pre-

sence of vacancies within Fe layers and spanthe range in-between ~47 to 50 at% Fe (Fe7S8

to FeS), whereas most structural complexityoccurs in the narrow compositional range bet-ween 47 to 48 at% Fe. At temperatures below300 °C various superstructures are observed inthis range due to vacancy ordering among andwithin Fe layers, resulting in a rather complica-ted subsolidus phase diagram (e.g., Nakazawa& Morimoto, 1970). Based on a NiAs-type structure of FeS, the wellunderstood monoclinic 4C structure is derivedby removing 1/4 of Fe atoms from every se condFe layer, leading to the composition Fe7S8. Fourvacancy layers are mutually discernible in thestacked arrangement and the resulting sequen-ce of partially vacant (V) and filled (F) Fe lay-ers is VAFVBFVCFVDF. The spacing between Vlayers of the same type is strictly four timesthe c-axis repeat of the NiAs substructure (oreight Fe layers). Structural descriptions of integral NC-pyrrhoti-te structures can be roughly divided into twoapproaches using either discrete (0 or 1) orpartial Fe site occupancies. The former relatesto an idealized extension of the 4C stackingscheme by introducing additional F layers,which increases both the periodicities of thesuperstructures as well as the Fe/S ratios, whilethe latter proposes in addition to extended lay -er stacking a considerable atomic scale dis or -der of Fe vacancies. Yamamoto & Nakazawa (1982) developed asuperstructure description for an Nc = 5.54pyrrhotite based on a continuous modulationof (partial) Fe site occupancies. In this model ofan aperiodic crystal structure the probability offinding a Fe atom at a given lattice site is go -verned by a continuous, sinusoid-like modula-tion function embedded in a four-dimensionalsu perspace formalism. Aperiodicity resultsfrom the occupation modulation being incom-mensurately related to the substructure peri-odicity. Based on this work Izaola et al. (2007)introduced a generalized superspace approachby replacing the continuous occupancy modu-lation with a step-like function governing dis-crete site occupancies. By virtue of a »close-ness« condition, the model strictly obeys the


184

sequence of ABCD layering as well as the rulethat two V layers must not follow consecutive-ly without an interjacent F layer, resulting inthe fixed relation 1/Nc = 2x between super-structure Nc value and composition expressedas x in Fe1-xS.A primary motivation for this study was thatthe reported electron diffraction pattern ofnon-integral NC-pyrrhotites strongly resemblethose seen in certain translation interfacemodulated binary alloys such as Au3+xZn1-x

(Amelinckx & Van Dyck, 1992). In these alloysystems integral superstructures arise in stoi-chiometric compounds as Au3Zn from orderingatoms on certain sites of the fcc lattice. Non-stoichiometric compositions and further super-structures are related to variably ordered, non-conservative (»composition-changing«) anti-pha se boundaries (APBs) of these simple su -perstructures. Such translation interface mo -du lation (TIM) is characterized by replacement(or splitting) of the simple superstructure dif-fractions spots (e.g., those of 4C-pyrrhotite)with arrays of satellite reflections – as seen inNC-pyrrhotites.

2.1. Samples and experimental methodsIn this study, two natural pyrrhotite samplesfrom geologically different locations were cha-racterized in detail by electron microscopy. Sam -ples were obtained from the Sta. Eulalia mi -ning district in Chihuahua, Mexico (EUL) andfrom the Nyseter mining area near Grua, Nor -way (NYS). Ar-ion thinning for transmission elec -

tron microscopy (TEM) was carried out usingliquid nitrogen cooling or a low angle thinningsystem. In order to check for preparation arte-facts, powdered and ultramicrotomed sampleswere prepared as well, but no systematic diffe-rences in SAED pattern could be detected. Nc

values were obtained from selected area elec-tron diffraction (SAED) patterns with typicaluncertainties of ±0.01 (2s). TEM superstructu-re dark-field imaging (SDF) was accomplishedby placing an objective aperture of 2.3 nm-1

diameter on strong satellite reflections.

2.2. Results - Electron microprobe analysis(EMPA)Back scatter electron (BSE) imaging of EUL andNYS shows abundant exsolution lamellae inboth samples. In sections the lamellae formtwo sets of lens shaped or sigmoidal bodiesinclined towards the [001] direction of hostpyrrhotite. Lamellae typically measure <200µm in length and up to 30 µm in width and aredarker in BSE images (Fig. 1a). Quantitativeanalyses show the lamellae to have composi-tions near Fe0.875S (or Fe7S8) indicative of 4C-pyrrhotite, while the host pyrrhotites’ Fe con-tents are slightly below Fe0.900S (or Fe9S10) asindicated by x values of 0.106±0.001 (EUL)and 0.102±0.001 (NYS). Contrary to the EULsample NYS shows a small Ni content of 0.05to 0.11 wt%. Exsolution of 4C-pyrrhotite fromNC-pyrrhotite appears to be quite abundant innature and is well documented by petrogra-phic methods.

Figure 1: (a) BSE image of 4C/NC intergrowths in sample EUL. (b) TEM-SDF image of the 4C/NC interface in EUL. APBsterminate in eight-fold node structures. (c) TEM-SDF image of the 4C/NC interface in sample NYS. Nearly identical nodestructures occur at the boundary but additionally changes in APB configurations (arrow) in the NC phase can be seen


185

2.3. Results - TEM-SAEDThe indexing of superstructure satellite reflec-tions is done by applying a (3+1)-dimensional(hklm) scheme analogous to the one adoptedby Yamamoto & Nakazawa (1982) and com-monly used for the description of one-dimen-sionally modulated structures (Fig. 2). A dif-fraction vector g is thus:

g = h a* + k b* + l c* + m q

m is the satellite order of a superstructure ref -lection and q is the modulation vector relatingeach satellite to a (hkl) substructure reflection.q is generally parallel or subparallel to c*. TheNc value of a superstructure is obtained as thelength ratio c*/q with c* referring to the NiAs-type subcell (c* ≈ 1.72 nm-1). For the EUL andNYS samples coincident Nc values of 4.81-4.87and 4.78-4.96, respectively, have been obtai-ned. In general, the Nc values compare well tothe x values obtained from EMPA assuming therelation 1/Nc = 2x. Based on the hypothesis of an APB modula-tion, the splitting of reflections in SAED pat-terns obtained from zone axes perpendicularto c allows to calculate average spacings anddisplacement vectors of APBs (Amelinckx &Van Dyck, 1992). The geometrical evaluationof pattern obtained from EUL and NYS suggestspacings of 1.58 to 1.74 nm and a displace-ment vector R = 1/8[001]4H (the 4H indexingrefers to a hexagonal cell in which the NiAs aand c dimensions are doubled and quadrupled,respectively).

2.4. Results - TEM-SDFIn order to obtain TEM-SDF images from elec-tron beams diffracted by the superstructure, itwas found to be most effective to place theobjective aperture on the (0004) and (008-4)satellite reflections accessed from near the[210]4H zone axis. The sample foil was tiltedsuch that in the ideal case only diffracted beamsin the (00lm) reciprocal lattice row were excited.In the NC portions of both the EUL and NYSsamples, SDF images reveal dense arrays ofdark stripes being aligned parallel to the (001)planes and solely visible in dark-field images

using strong superstructure reflections. In EULthe stripes appear more ordered and show lesswiggles and irregularities compared to NYS,which displays some weak diffuse diffraction inthe area from which the images were obtai-ned. Further, in NYS the configuration of stri-pes appears to change along certain domainboundaries where the thin stripes, as seen inEUL, change into thicker stripes. The latterdominate the NYS sample and are likely dou-blets of two stripes being too close to be resol-ved under the SDF imaging conditions. Theaverage spacing of stripes in EUL is 1.63±0.05nm and the observed (half-)width of a singlestripe is approx. 0.80 nm. Assuming the thickstripes in NYS to be non-resolvable doublets,the average spacing between individual stripesis 1.94±0.10 nm. The average distance for EUL compares verywell with the expected APB spacing as obtai-ned from the SAED pattern geometries and wetherefore interpret the dark stripes to be APBs.Based on contrast in dark-field images we canexclude them to be stacking faults or twinboundaries, as the former would be seen withsubstructure reflections as well and the latterwould show intensity differences of domainson either site of the boundary. The larger ave-rage spacing in NYS, despite very similar Nc

values, is likely due to the higher disorder withlarger and more irregular spacings betweenstripes which bias the mean distance to highervalues. The average of 1.94 nm corresponds toan Nc value of approx. 4.7, which is compati-ble with diffuse satellite reflections seen insome SAED patterns. Most striking features in SDF images of EULand NYS are the phase boundaries between4C- and NC-pyrrhotite where the stripes formcomplicated but remarkably similar node struc-tures in which the stripes terminate. When thephase interface is oriented mostly parallel tothe stripes, no nodes occur directly at the inter-face, but occasionally appear nearby withinthe NC phase, often associated with crystallo-graphic edge dislocations having the Burgersvector b = [001] of the NiAs cell. In all cases ofnode structures we observed eight stripes toterminate in a single node.


186

The behavior of stripes in the observed nodestructures and adjacent 4C-pyrrhotite stronglyendorses our interpretation of them beingnon-conservative APBs of the 4C structure.Their orientation is fully compatible with theinferred displacement vector R = 1/8[001]4H

which creates double layers of completely filledFe positions as opposed to single layers beingpresent in the 4C structure. The eight-fold no -des facilitate the reconstruction of the 4C pha-ses from the APB modulated NC phase anddirectly support the proposed displacementvector 1/8[001]4H as its eight-fold multiple willbe the c dimension of the hexagonal 4C unitcell (see discussion section). Apparently thehighly similar node configurations represent aself-organization towards an energeticallyoptimal phase interface.

2.5. DiscussionOur TEM observations provide compelling evi-dence that the structural diversity of non-inte-gral NC-pyrrhotites can be explained in termsof a TIM structure model. The translation inter-faces are APBs based on the Fe sublattice of 4C-pyrrhotite. This model bears resemblance to pre -viously suggested »out-of-step« and anti-phasedefect models (Pierce & Buseck, 1974) but ismore precise on the actual nature of the invol-ved translation interfaces. On the other hand,the superspace approach of Izaola et al. (2007)is fully compatible with our TIM model. In es -sence, this superspace model produces discre-tely modulated superstructures by insertingadditional F layers into the basic 4C layer se -

quence, producing either periodic or aperiodic(but uniform) stacking sequences parallel to(001). Because the so produced double F layerscarry an Fe excess, the resulting phases willhave higher Fe/S ratios than 4C-pyrrhotite andfollow strictly the relation 1/Nc = 2x as spacingsbetween double layers are evenly distributed. For a structure with Nc = 4.8, which is a simplerepresentation for the host NC-pyrrhotite inour EUL and NYS samples, it is possible to deri-ve two well ordered stacking schemes:(DSSDS)4 and (DDSSS)4, where D representsdouble F layers and S single F layers, intercala-ted with the four vacancy bearing layers. The(DSSDS)4 sequence can be derived via thesuperspace model of Izaola et al. (2007) and isuniform, while the (DDSSS)4 sequence cannotbe obtained from the model and has lessequally spaced D layers, leading to slightly lessorder while maintaining the same averagevacancy-vacancy distance. The observed D layer or APB arrangements inEUL are alike the well ordered (DSSDS)4 se -quence, while in NYS domains relating to bothstacking schemes coexist as indicated by thechanges in stripe configuration. Here the do -mains containing the thicker, non-resolvablestripes may correspond to the (DDSSS)4 stak-king in which the closest spacing between Dlayers or APBs is approx. 0.86 nm and thusmore or less equal to the half-width of a singlestripe as observed in EUL. The average ~1.72nm D layer spacing of the idealized (DSSDS)4sequence compares well with the average stri-pe and APB spacings in EUL as obtained from

Figure 2: Diffraction patterns of non-integral NC-pyrrhotite. (a) Schematic representation of zone axis [210]for Nc = 4.85. Grey open circles/indices represent 4C reflections, crosses mark the (generally absent) 4Creflections used for superstructure indexing. (b) Experimental SAED pattern of NYS with Nc = 4.84±0.01


187

SDF images and SAED patterns, respectively,and consistently reproduces the observedslightly non-equal spacings in the bulk NCphase. It appears reasonable to assume that incase of NYS the energetic difference betweenboth arrangements of D layers in the (DSSDS)4and (DDSSS)4 stacking variants are very smalland either the structure is observed on a tran-sition to a more ordered state as in EUL or theelevated Ni content exerts some control onvacancy arrangements. So far the discussion focused on an idealizedlayer stacking involving parallel layer arrange-ments. SDF images clearly shows that this idealsituation is virtually never fulfilled in the stu-died samples as stripes and therefore APBs areslightly wavy, corrugated and often non-paral-lel. Changes in stacking schemes at given Ncvalues as well as disorder introduced by bentAPBs will not produce strong effects in SAEDpatterns. For the (DSSDS)4 and (DDSSS)4 stak-king schemes the diffraction pattern geome-tries will be the same with only moderate dif-ferences in spot intensities, likely being hiddenby dynamical diffraction effects. Increasingmesoscale disorder of APBs will first lead tovanishing higher order satellite reflections andgradually lead to streaky diffraction along thec* direction when a quasi-periodicity is no lon-ger maintained. While a D layer in the ideallayer model is always an APB of the Fe sublat-tice the converse is not strictly true as isolatedAPBs are capable of moving more or less free-

ly through the crystal and can have even in theconfinement of a modulated structure a muchmore dynamical behavior as represented by astatic layer model. Therefore, describing thelayer stacking as a TIM structure of denselyarranged APBs offers much better prospects ofunderstanding the behavior and properties ofnon-integral NC-pyrrhotites. The need of a more dynamical model clearlybecomes apparent when trying to understandthe node structures observed at NC/4C phaseinterfaces. The termination of eight stripesequivalent to eight APBs in a single node isnecessary to reconstruct the 4C structure froman NC structure with well ordered V layer stak-king. Because the stacking of V layers on theleft side strictly obeys the ABCD sequence, thetermination creates a sequence fault where twoVF layers follow on each other. Conse quent ly,two APBs have to terminate either at a perfectdislocation with b = [001]NiAs = 1/4[001]4H, whichremoves a pair of VF layers, or 4 pairs of APBshave to terminate close-by such that the eight1/8[001]4H displacement vectors add up to the4C lattice repeat. Both variants and combina-tions thereof are seen in our samples and wesuggest that the observed node configura-tions facilitate an energetically most favorablestate when there are no or few dislocationsinvolved. Based on the TIM model we suggest that thereis no significant difference between integraland non-integral pyrrhotites and both types

Figure 3: a) BSE image of the TYS sample. Bright lamellae contain 2 wt% more iron than the matrix. b) SEimage of the surface of the abiotically reacted pyrrhotite cubes after 40 days which shows almost no disso-lution. c) SE image of the pyrrhotite surface reacted with A. ferrooxidans after 40 days. Deep trenches for-med at the position of the stoichiometric FeS lamellae


188

can be understood as part of a continuum ofstructures in-between Fe7S8 and (at least)Fe11S12, in which variable Fe contents are go -ver ned by changes in densities and orderingstates of APBs. Based on the presented dataand preliminary data on other natural pyrrhoti-te samples we currently do not see evidencefor preferred stability of integral NC-pyrrhoti-tes in exsolution association with 4C-pyrrhoti-te, but the true phase relations may be obscu-red by slow diffusion kinetics below 220 °C,where NC-pyrrhotites first appear (Nakazawa& Morimoto, 1970). However, the very sharpcompositional gradients at the 4C/NC interfa-ces in our EUL and NYS samples point to wellequilibrated assemblages. Therefore, the asso-ciation of Nc ≈ 4.85 pyrrhotite with exsolved4C-pyrrhotite in two of our samples, as well asthe coexistence of Nc = 4.88 pyrrhotite and 4C- pyrrhotite observed by Morimoto et al. (1975),points to a favored stability of NC-pyrrhotiteswithin the range of Nc = 4.8 to 4.9 – at leastwith regards to typical temperatures of theuppermost crust, where the samples residedfor geological timescales. The insight into complexity in the bulk structu-re and the applied TIM model can lead to abetter understanding of physical propertiesand the reactivity of surfaces. Due to the rathertwo dimensional, layer-like structure of pyrrho-tites one clear expectation is a pronouncedanisotropy of surface properties. Preliminaryexperimental results indicate that reactionspeeds on basal (001) faces are considerablydifferent than on prismatic (hk0) surfaces.Owing to pyrrhotite’s excessively variable mag-netic and physicochemical properties andwidespread occurrence in rocks and ore depo-sits, many potentially important implications tothe fields of geomagnetics, petrology, environ-mental mineralogy, and technical mineral pro-cessing (e.g. pyrrhotite flotation) are to be ex pec -ted when the presented detailed characteriza-tion of structures and phase relations is applied.

3. Structural controls on the alteration ofpyrrhotite by Acidithiobacillus ferrooxi-dans under acidic conditonsBased on the detailed characterization proce-dures described above, a well-directed studyon the reactivity of pyrrhotite intergrowth inexperiments with acidophilic microorganismswas conducted. Acidophilic microorganismsplay a dominant role in acid mine drainage(AMD) and for the metal recovery from low-grade ores by bioleaching applications. A num-ber of studies focused on the reaction paths ofchemical and biological metal sulfide oxidation(Rawlings, 2002; Schippers, 2004). However,experimental studies on the biologically-en hanc -ed pyrrhotite oxidation are still scarce (seeBelzile et al. 2004 for review) or focus on bio-leaching applications of low grade ores (Ke &Li, 2006; Jiang et al., 2007), but clearly indica-te the catalytic effect of microorgansmswithout reporting a reliable characterization ofthe starting material and dissolution rates.Schippers et al. (2007) provides a potential pyr-rhotite oxidation rate measured in a mine tai-ling using microcalorimetry, which suggeststhat natural rates can exceed laboratory ratesby up to 4 orders of magnitude. Beside theenhancing effect of microbial consortia, pyr-rhotite dissolution may be also influenced bythe reactivity and intergrowth texture of diffe-rent superstructures. Janzen et al. (2000) stu-died a number of pyrrhotite samples from diffe-rent localities and determined the relative pro -portions of so-called monoclinic (mainly 4C) andhexagonal (NC) pyrrhotite. No systematic trendin dissolution rates were found, however, infor-mation about the intergrowth texture is missingfor these samples.Acidithiobacillus ferrooxidans, an acidophilicproteobacteria that catalyses the oxidation offerrous iron, elemental sulfur and reduced sul-fur compounds is by far the best characterizedbioleaching microorganism and has been usedin the presented study. Cells of A. ferrooxidansproduce an exopolysaccharide (EPS) layer toattach to the sulfide mineral surface (Gehrke etal., 1998) and this attachment is viewed as aprocess by which the bacterial membranecomponents interact directly with metal and


189

sulfur of the mineral by an enzymatic type ofmechanism (Sand et al. 1995).In this study X-ray photoelectron spectroscopy(XPS), scanning electron microscopy (SEM),and confocal microscopy have been used totrace the chemical and morphological changesat the surface of a well characterized pyrrhoti-te sample. Surface topography has employedto quantify a differential dissolution rate bet-ween intergrown pyrrhotites.

3.1. Samples and experimental methodsA pyrrhotite from Tysfjord region, Norway(TYS) was used in this study. The TYS sampleshows a dense population of 1-3 µm wideexsolution lamellae with a common orientationparallel to (001). They appear brighter in BSEimages compared to the matrix indicating ahigher Fe/S ratio (Fig. 3a). Quantitative analy-ses place them close to stoichiometric FeS (tro-ilite). The matrix pyrrhotite of the TYS samplehas a composition close to Fe0.900S (or Fe9S10)with x values of 0.096±0.001 and thus, thelamellae contain about 2 wt% more iron thenthe matrix.Polished cubes (3x3x3 mm) of TYS pyrrhotitewere dissolved in presence of Acidithiobacillusferrooxidans at pH 2 over 40 days with an a bi -otic control in parallel. A. ferrooxidans ATCC19859, obtained from the American Type Cul -ture Collection (ATCC, Manassas, VA), wasrou tinely cultivated as a pre-culture in ATCC

medium 2039 at 28°C for five days on a rota-ry shaker (300 rpm). For the inoculation of theexperiments 10 µl pre-culture (approx. 2×103

cells/ml) was directly used. Cultures for theexperiments were grown in 50 ml mineral saltsolution with 0.4 g each of (NH4)2SO4, K2HPO4,and MgSO4×7H2O per liter. Cubes were remo-ved from the batch experiments after 1, 4, 7,14, 17, 21, 28 and 40 days, respectively. XPS was performed on a PHI Quantera SXM-03 Scanning XPS Microprobe using an Al Ka X-ray source with a highly focused beam (<9microns). High resolution scans of the Fe 2p, S2p, and O 1s peaks were used to determinethe bond and valence state information. XPSreference binding energies for pyrrhotite provi-ded by Legrand et al. (2005) were used toassign and fit the spectra. The topography of selected surfaces was mea-sured by the confocal 3D microscopy systemµsurf custom (Nanofocus AG, Oberhausen). Itconsist of a monochromatic LED light source (l= 505 nm), a helically-shaped arrangement ofpinholes (Nipkow-Disk), a stage with piezomodule for z movement and a megapixel CCDcamera. The objective used (numerical apertu-re 0.9) provides 160 µm x 160 µm field of viewwith a resolution of 10 nm in z and about 500nm in x/y.

3.2. ResultsA. ferrooxidans develops a visible biofilm on

Figure 4: Overlays of the Fe 2p3/2, O 1s and S 2p narrow region scans of pyrrhotite surfaces abiotic control (left) andbiotic (right) experiments. Labels denote the duration of the experiments


190

the pyrrhotite cubes after approximately 5 daysof incubation. After 9 days the whitish biofilmcovered the entire pyrrhotite cubes and grewslowly until the experiment was finally stoppedat 40 days. After removing the biofilm by awashing procedure the color has changed todark blue/black and the reflectivity from shinyto matt. In contrast, the control pyrrhotitecubes (abiotic) were still shiny golden in allexperiments with only minor changes in reflec-tivity. XPS analyses (Fig. 4) reveal a significant-ly more oxidized surface in the biotic experi-ment with a clear shift from Fe(II)-S to Fe(III)-Obonding. Compared to the control experiment,the surface is more hydroxylated and showseven traces of sulfate. SEM images (Fig. 3) ofthe pyrrhotite surface exposed to A. ferrooxi-dans reveal long and deep trenches which startto develop already after 4 days of experiment.The position of the trenches is identical to thetroilite lamellae indicating that the reactivity ofthe lamellae is significantly higher than the sur-rounding NC-pyrrhotite. Trenches reach depthsof about 2 µm and a width of 3 µm as deter-mined by confocal microscopy measurementsafter 28 days (Fig. 5). Height profiles suggestthat the lamellae are inclined to the dissolvingsurfaces which lead to the nonsymmetricalshape of trenches. Although information ofthe original surface height is missing, the 3Dtopographic data can provide a differential dis-solution rate of the 2C-pyrrhotite compared tothe NC-pyrrhotite by using the matrix surfaceas a reference height. Five topographic imagestaken from surfaces after 21 and 28 days resul-

ted in a differential dissolution rate of8.8±1.2×10-9 and 1.2±0.3×10-8 mol m-2s-1,respectively.

3.3. DiscussionThe dissolution experiments clearly show thecatalyzing effect of microorganisms at condi-tions that are unfavorable for chemical dissolu-tion. The XPS data indicates a fast and extensi-ve oxidation of the biologically altered surfacewith dominating of Fe(III)-O bonds as well assome sulfate which points to the formation ofa iron hydroxide layer (possibly FeOOH). How -ever, this does not result in a passivation of thesurface as indicated by a continuous deepe-ning of the 2C lamellae which possible can beexplained by the presence of EPS and a rege-neration of ferric iron by the bacteria. For thefirst time, the clear difference in reactivity bet-ween two pyrrhotite superstructure (2C andNC) has been shown by the presented experi-ments. A direct and an indirect mechanism forbiologically-enhanced dissolution have beenproposed in the literature. For the directmechanism, bacteria are attached to the surfa-ce surrounded by EPS which provides comple-xed ferric iron (Rodriguez-Leiva & Tributsch,1988). The indirect mechanism interprets dis-solution as an inorganic process where non-attached bacteria only maintain a certain ferriciron concentration which accelerates the disso-lution (Rodriguez et al., 2003). Whether the dif -ference in dissolution rate is facilitated by apreferred attachment of bacteria due to thedifferent iron content or are caused by diffe-

Figure 5: Surface of the TYS sample after biologically-enhanced dissolution measured for 28 days measuredby confocal microscopy. The height of the ma trix NC-pyrrhotite (~0.5 µm) was used as reference height todetermine the differential dissolution rate from thedissolved volume of the 2C-lamellae. The squares illu-strate the principle of roughness parameter calcula-tion as function of sampling size performed withROUGHNECK.The upper profile clearly shows the non-symmetricalshape of the trenches which is explained by the incli-ned intersection of the lamellae with the surface


191

rences in reactivity of troilite and NC-pyrrhoti-te alone is not yet been conclusively evaluatedand will be subject of further activity. In anycase, the lamellar structure of the trenches cannot be interpreted as a primary microbial bio-signatures because it is originally caused by theexsolution process during equilibration of thepyrrhotite sample and acts like a templateduring dissolution. In any case, the differential dissolution rate isalready higher than all reported abiotic dissolu-tion rates and highlights the significance of thestructural variability and intergrowth texture.When we combine the differential dissolutionrate with the only available biotic dissolutionrate for pyrrhotite under acidic conditions(8×10-9 mol m-2s-1) reported in Belzile et al.(2004), the resulting net dissolution rate is a -bout one order of magnitude higher than re -ported rates in the absence of additional ferrousFe. Further accelerating factors towards theobserved rates include the combined effect ofbacterial consortia (combinations of iron- andsulfur-oxidizer) as well as the increase of reactivesurface area by the deepening of the trenches.

4. From surface morphology to rates: Anautomated routine to evaluate convergedroughness parameters of heterogeneoussurfacesAn increase of surface roughness during biolo-gically-enhanced dissolution of pyrrhotitewhich, in turn, affects the dissolution rate wascaused by both the microbial activity and thesample heterogeneity. Two components contri-bute to the surface roughness evolution of theTYS sample, namely, the alteration of the NCmatrix and the deepening of trenches at the 2Clamellae. The high differential dissolution ratemay be at least partly explained by increasingsurface area. For the presented surfaces (Fig. 5)we are therefore interested in the evolution ofsurface roughness for the matrix, the lamellaeand the combined surface of the biologicallydissolved pyrrhotite surface, respectively. In geosciences, rates are usually determinedfrom dissolution studies (either flow throughor batch experiments) using a size fraction of

ground and sieved starting material by measu-ring the change in fluid chemistry as a functionof time. The surface area of the starting mate-rial is commonly determined by the BETmethod and used as a constant to normalizerates. Recently, topographic methods havebeen introduced to directly measure sulfidedissolution rates and surface roughness (Astaet al., 2008) and offer new insides into thereaction mechanisms and surface reactivity.Furthermore, it allows a systematic determina-tion of surface roughness parameters. As statistical quantities roughness parametersare strongly dependent on the field of view ofthe used technique or the sampling size. Theconcept of »converged roughness parameter«(Fischer & Lüttge, 2007) was shown to be veryuseful to characterize certain surface buildingblocks by employing recurrent (squared) bisec-tions of the measuring field with an edgelength of a. A surface parameter is defined ascon verged when a flat slope is found in theconvergence graph (roughness parameter ver-sus edge length a). Converged parametersintrinsically depend on the choice of »repre-sentative« areas for their analysis. Further more,artificial measuring points may lead to an ove-restimation of roughness parameters. Here, wepresent the newly developed program ROUGH -NECK, which calculates a number of surfaceroughness parameters from topographic data asfunction of sampling size allowing to analyzeand visualize roughness parameter distributions.

4.1. The program ROUGHNECKThe program ROUGHNECK utilizes the com-mon Surface Data Format (SDF) (Blunt & Stout,2001) as input and output file format and is astand-alone application that can be used onWindows and Linux systems. Our input filesproduced by confocal microscopy consist of984x984 points representing a 160x160 µmfield of view. For this study a quality criteria ofa minimum of 99% of measured points wasused for the topographic data. The residualnon-measured points were linearly interpola-ted. The following 3D roughness parametersare implemented so far: the roughness avera-ge Sa, the root-mean-square roughness Sq, the


192

peak-to-peak height Sz, the ten-point-heightS10z, the root-mean-square-gradient Sdq, thesurface-skewness Ssk, the surface-kurtosis Sku,and the surface area ratio F. The program usesfreely selectable edge lengths for the desiredsampling area, which is moved pixel by pixelover the entire field of view (cf. Fig. 5).

4.2. Evaluation of ROUGHNECK outputsThe surface area ratio, F=A3D/A2D, which des-cribes the surface area of a three dimensionalsurface normalized by the two dimensional(projected) surface, is used as an example para-meter to illustrate the statistical procedure. Aperfectly flat surface will result in F=1, while arough surface results in F>1. This parameter isa so-called hybrid surface parameter because it

does not mainly depend on the amplitude butis more sensitive to the high frequency modula-tions of the surface as it is measured by theslope between adjacent surface points. ROUGHNECK has been used to calculate theconverged F parameters of the biologically dis-solved pyrrhotite surface after 28 days (Fig. 5).The distribution of roughness parameters (herefrequency N vs. F) can be plotted as histogramby applying discrete bins (Fig. 6a). The maximaof the distributions show the roughness valuefor a significant portion of the measured surfa-ce. For heterogeneous surfaces, a number ofmaxima may indicate various components withdifferent surface roughness. In general, themaximum F value increases with increasingsampling area. At small scale (a=1 to 10 µm),

Figure 6: F value distributionsfor various squared samplingareas with a given edge length.With increasing sampling sizethe maximum of the curves isshifted to higher F values. How -ever, the transition from low va -lues around 1.05-1.1 to valuesaround 1.2 is clearly visible andused in the convergence graph(Fig. 7a). For some curves (e.g. a = 4 µm) up to 3 maxima canbe located indicating the hetero-geneity of surface components

Figure 7: a) Evolution of F as function of the edge length of the squared sampling area calculated for fresh and reactedpyrrhotite surfaces. F at larger sampling size (>10 µm) is controlled by the deepening of 2C lamellae. b) Spatial distribu-tion of F values for a squared sampling size with edge length a=4 µm. High F values (~1.6) can be found at the posi-tions of the dissolved 2C lamellae. Since these areas represent only a small portion of the entire field of view, they arenot detectable in the histogram. Very high F values originate from overlying particles (cf. Fig. 5), but they do not alterthe converged roughness parameter determined from the histogram


193

the first maxima represent only the roughnessof the NC pyrrhotite matrix which dissolves slo-wer than the 2C lamellae. At larger scale (a=15to100 µm) the F value includes contributionsfrom both because the spacing between thelamellae is smaller than the edge length of thesampling area. The maxima of each distribution can be trans-lated into a convergence graph plotting the Fvalue of the maxima versus the samplinglength a. Comparing predominant roughnessvalues as a function of time allows to trace theincrease in surface roughness of the NC matrixalone (a=1 to 10 µm) as well as the evolutionof roughness for the entire surface (a=15to100 µm). The output data for every samplingarea can also be visualized providing a spatialroughness parameter distribution (Fig. 6b).This is particularly useful if certain surface com-ponents, like the locations of the former 2Clamellae, represent only a small portion of theentire surface and are hence not well repres-ented in the histogram. From the roughnessdistribution image (a=4 µm) a roughness para-meter for the 2C lamellae of 1.6 can be deter-mined. In turn, this also validates that theroughness parameter distribution (Fig. 6a) isnot sensitive for artefacts like overlying parti-cles or inaccurately measured surface points.

5. The reaction pathway of secondaryiron sulfide formation under anoxic con-dition: An experimental study on thereaction of lepidocrocite with dissolvedsulfideAMD waters are typically characterized by highamounts of ferric iron and sulphate. The for-mation of secondary iron sulfides in the sub-surface is an important pathway to reduce thecontaminant concentration and to raise the pHof overlying waters (Bilek, 2006). In this res -pect, the interaction between dissolved sulfideand ferric oxides can be regarded as a key reac-tion leading ultimately to pyrite formation (Can -field et al., 1992, Poulton et al., 2004), butknow ledge on the pathways and on control-ling factors is still limited. Here, results frombatch experiments on the reaction betweenlepidocrocite and dissolved sulfide underanoxic conditions are presented with a specialemphasis on the characterization of nanocry-stalline products by TEM forming at differenttime steps.

5.1. Materials, experimental setup andmethodsKinetic batch experiments were conducted bymixing lepidocrocite (25 mmol/L) and differentconcentrations of dissolved sulfide at roomtemperature and a constant pH of 7 within ananoxic glove box. After 1-2 hours, 24 hours,

Figure 8: Evolution of elemental sulfur (left) and acid extractable ferrous iron (right) during two weeks of reaction forselected experimental runs. The initial sulfide concentration in run 10 and 14 was twice as the concentration in run 15.Mackinawite formation is rapid within the first hours of the experiments, whereas pyrite formation is relatively slow,starts after 2 days and continued to the end of the run


194

1 week and 2 weeks, respectively, dissolvedFe(II) and S(-II), Fe(II) extractable with 0.5 NHCl, S°, and the total iron content were deter-mined by wet chemistry extraction. At diffe-rent time steps aliquots of the reacting suspen-sion were analyzed by TEM, XRD and Möss -bauer spectroscopy. In order to limit oxidationin air during TEM sample preparation the sus -pension was first sampled in gas-tight vials. Adrop of solution was then taken with a syringeand put onto a Lacey carbon-coated coppergrid. The grid was immediately transferred tothe TEM holder and inserted into the highvacuum of the TEM. The short exposure of thesample to air can be limited to 1-2 minutes atmaximum with this procedure.

5.2. ResultsThe reaction can be divided into three phaseson the basis of wet chemistry (Fig. 8) and TEMobservations (Figs. 9 and 10). In the first minu-tes of the reaction (0-15min) dissolved sulfideis rapidly consumed. It is partly oxidized to S°species in solution and partly consumed by ini -tial mackinawite formation. In a second pha se(15-120 min) a continuous but slower forma-tion of S° is observed while acid extractableFe(II) remained constant. TEM measurementsre vealed the occurrence of a mackinawite rimcovering the lepidocrocite crystals that wasseparate from the lepidocrocite surface by aninterfacial magnetite layer (Fig. 9). The magne-

tite layer can be seen as an intermediate stagelinking two reactions, the formation of macki-nawite which reduces ferric iron at the lepido-crocite surface on one hand and the transportof electrons in the deeper regions of the lepi-docrocite bulk crystal which facilitates reducti-ve dissolution and mackinawite rim growthand the other hand. The third phase is charac -terized by a decrease of acid extractable Fe(II)and S° and results in pyrite formation togetherwith some magnetite (2-14 days). TEM measu-rements indicate that mackinawite completelydissolves from the surface of lepidocrocitewhile the precipitation of pyrite occurs separa-tely from the lepidocrocite surface. The pyriteformation is coupled to mackinawite dissolu-tion and probably accompanied by the forma-tion of FeS clusters that serve as precursors.The conversion of mackinawite to pyrite wastraced by TEM at 72 hours, 1 and 2 weeks. Themackinawite rims get gradually thinner andmore corrugated with time. After 48 hoursamorphous regions composed of iron and sul-fur in variable concentration can be found bet-ween lepidocrocite grains. Pyrite grains reachdiameters between 200 and 500 nm after 2weeks and commonly show striking cubic oroc tahedral outlines (Fig. 10). However, thegrains are composed of smaller cubic buildingblocks which point to a formation by orientedaggregation. In addition to pyrite, some mag-netite grains similar in size and shape could be

Figure 9: Bright field (a) and high resolution (b,c) TEM images of lepidocrocite crystals with sulfur-rich rims after 2 hoursof reaction. The spotted contrast on lepidocrocite grains in (a) is due to nanocrystalline mackinawite. In (b) characteri-stic (001) and (111) lattice fringes of mackinawite (0.52 and 0.23nm, respectively) are visible in the outer rim. A conti-nuous intermediate layer (arrow) shows fringes matching d220 of magnetite/maghemite (arrow). This layer can also benoticed in (a). A nanocrystal of mackinawite in [010] zone axis orientation is shown in (c) together with its calculatedFFT and a simulated diffraction pattern as inset


195

identified by STEM-EDX and electron diffrac-tion as well.

5.3. DiscussionThe temporal evolution of the chemical speciesconcentrations and solid phases indicate thatthe reaction progress is highly dynamic. TEMdemonstrates that several phases form duringthe reaction towards pyrite. On the basis of theobserved concentrations a number of balancedchemical equations can be formulated invol-ving sorbed iron species as well as polysulfites.However, it is obvious that the system is kineti-cally controlled towards chemical equilibration(pyrite formation) due to the slow reductivedissolution process and the redox disproportio-nation. This study has provided novel insightsinto the redox processes occurring at the sur-face and in the vicinity of the lepidocrocite cry-stals and reveals the occurrence of magnetiteand mackinawite in the initial phase of thereaction covering the lepidocrocite surface. Itfurther demonstrates that dissolution of mak-kinawite occurs prior to formation of pyritepresumably being followed by further reactionwith polysulfides in the absence of dissolvedsulfide. It has also demonstrated that the kine-tics of precursor formation from reductive dis-solution of ferric (hydr)oxides by hydrogen sul-fide and of subsequent pyrite formation arekinetically decoupled. This observation impliesthat in iron-rich aqueous systems of periodicalsulfide formation (e.g. movement of the capil-lary fringe in ground waters) ferric oxides may

undergo »charging« with reduced substancesthat may lead to pyrite (and magnetite) forma-tion even in the absence of dissolved sulfide.This reactions sequence has also implicationsfor the retention of toxic metal by adsorptionon nanocrystalline mackinawite which may getmobilized by further reaction to pyrite.

6. ConclusionThe most important scientific insights obtainedwithin the MIMOS project are:(i) A new structural model (TIM model) has

be en successfully applied to non-sto i chi o -me tric NC-pyrrhotite. It well explains im -portant structural details that are observedin electron diffraction patterns and dark-field images and can lead to a better under-standing of surface and bulk physical pro-perties of pyrrhotite relevant to geomagne-tics, petrology, environmental mineralogy,and technical mineral processing.

(ii) Differences in reactivity caused by small stru -ctural and chemical variations have be endirectly proven for biologically-enhanced dis -solution of pyrrhotite. The intergrowth tex-ture has therefore a direct influence on thedissolution rate by a continuous increase ofreactive surface area. This result is relevantfor the assessment of the weathering ratesof pyrrhotite tailings and bioleaching appli-cations.

(iii) The evolution of surface area was quanti-fied by combining topographic data and a

Figure 10: High resolution images, electron diffraction pattern and EDX spectra of pyrite. Note the aggre-gative nature of the grain consisting of cubic building blocks. Slight misorientations are also reflected inthe diffraction pattern


196

statistical evaluation of surface roughnessusing the newly developed programROUGHNECK for heterogeneous surfaces.This approach provides new prospects formineral dissolution studies as well as formore general topics dealing with surfacealteration like corrosion.

(iv) Analytical TEM was successfully applied totra cing the reaction pathway of iron sulfi-de for mation. The new results are relevantfor our understanding of processes andelement distributions at geochemicalboundaries which can help to deciphersuccessful remediation strategies for pol-luted waters and soils.

ReferencesAmelinckx, S. & Van Dyck, D. (1992) Electrondiffraction effects due to modulated structu-res. Electron diffraction techniques, IUCr/Ox -ford Science Publications, 2, 309-373.

Asta, M.P., Cama, J., Soler, J.M., Arvidson, R.S.& Lüttge, A. (2008) Interferometric study ofpyrite surface reactivity in acidic conditions.Am. Mineral., 93, 508-519.

Becker, M., de Villiers, J. & Bradshaw, D. (2010)The flotation of magnetic and non-magneticpyrrhotite from selected nickel ore deposits.Miner. Eng., in press, doi:10.1016/j.mineng.2010.07.002

Belzile, N., Chen Y., Cai, M. & Li, Y. (2004) Are view on pyrrhotite oxidation. Journal of Geo -chemical Exploration, 84, 65-76.

Blunt, L. & Stout, K.J. (2001) EC Project FinalReport »Development of a Basis for 3D SurfaceRoughness Standards« SMT4-CT98-2209.

Bilek, F. (2006) Column tests to enhance suphi-de precipitation with liquid organic electron do -nators to remediate AMD-influenced ground -water. Environ. Geol., 49, 674-683.

Canfield D. E., Raiswell R. & Bottrell S. (1992)The reactivity of sedimentary iron minerals to -

ward sulphide. Am. J. Sci., 292, 659-683.

Church, C.D., Wilkin, R.T., Alpers, C.A., Rye, R.O. & McCleskey, R.B. (2007) Microbial sulfatereduction and metal attenuation in pH 4 acidmine water. Geochem. Transac., doi:10.1186/1467-4866-8-10.

Ekmekçi, Z., Becker, M., Bagci Tekes, E. & Brag -shaw, D. (2010) An impedance study of theadsorption of CuSO4 and SIBX on pyrrhotitesamples of different provenances. Miner. Eng.,in press, doi:10.1016/j.mineng.2010.02.007

Fischer, C. & Lüttge, A. (2007) Converged sur-face roughness parameters – A new tool toquantify rock surface morphology and reactivi-ty alteration. Am. J. Sci. 2007, 307, 955-973.Gehrke, T., Telegdi, J., Thierry, D. & Sand, W.(1998) Importance of extracellular polymericsubstances from Thiobacillus ferrooxidans forbioleaching. Appl. Environ. Microbiol., 64,2743 – 2747.

Izaola, Z., Gonzalez, S., Elcoro, L., Perez-Mato,J.M., Madariaga, G. & Garcia, A. (2007) Re vi -sion of pyrrhotite structures within a commonsuperspace model. Acta Cryst. B, 63, 693-702.

Janzen, M.P., Nicholson, R.V. & Scharer, J.M.(2000) Pyrrhotite reaction kinetics: reaction ra -tes for oxidation by oxygen, ferric iron, and fornonoxidative dissolution. Geochim. Cos mo chim.Acta 64, 1511 –1522.

Jiang, L., Zhou, H.Y. & Peng, X.T. (2007) Bio-oxidation of pyrite, chalcopyrite and pyrrhotiteby Acidithiobacillus ferrooxidans. Chin. Sci.Bull., 52, 2702-2714.

Ke, J. & Li, H. (2006) Bacterial leaching of nik-kel-bearing pyrrhotite. Hydrometallurgy, 82,172-175.

Legrand, D.L., Bancroft, G.M. & Nesbitt, H.W.(1995) Oxidation/alteration of pentlandite andpyrrhotite surfaces at pH 9.3: Part 1. Assign -ment of XPS spectra and chemical trends. Am.Min., 90, 1042-1054.


197

Morimoto, N., Gyobu, A., Tsukuma, K. &Koto, K. (1975) Superstructure and nonstoi-chiometry of intermediate pyrrhotite. Am.Min., 60, 240-248.

Nakazawa, H. & Morimoto, N. (1970) Pyrrho -tite phase relations below 320°C. Proc. JapanAcad., 46, 678-683.

Nicholson, R.V. & Scharer, J.M. (1994): La bor -atory studies of pyrrhotite oxidation kinetics.In: Environmental geochemistry of sulfide oxi-dation. Alpers C.N., Blowes D.W. (eds.), .ACSSymp. Series 550, 14-30.

Ohfuji, H. & Rickard, D. (2006) High resolutiontransmission electron microscopic study of syn-thetic nanocrystalline mackinawite. Earth andPlanetary Science Letters, 241, 227-233.

Pierce, L. & Buseck, P.R. (1974) Electron ima-ging of pyrrhotite superstructures. Science,186, 1209-1212.

Pollok, K., Langenhorst, F., Hopf, J., Kothe, E.,Geisler, T., Putnis, C.V. & Putnis, A. (2008) Micro -structural Controls on Monosulfide Weatheringand Heavy Metal Release (MIMOS), GEOTECH-NOLOGIEN Science Report, 12, 79-88.

Poulton S.W., Krom, D.M. & Raiswell R. (2004)A revised scheme for the reactivity of iron (oxy-hydr)oxide minerals towards dissolved sulfide.Geochim. Cosmochim. Acta, 68, 3703-3715.

Rawlings, D.E. (2002) Heavy metal mining u -sing microbes. Annual Reviews in Micro bio -logy, 56, 65–91.

Rodriguez, Y., Ballester, A., Blazquez, M. A. &Munoz, J. A. (2003) Study of bacterial attach-ment during the bioleaching of pyrite, chalco-pyrite and sphalerite. Geomicrobiol. J. 20,131–141.

Rodriguez-Leiva, M. & Tributsch, H. (1988) Mor -phology of bacterial leaching patterns by Thi o -bacillus ferrooxidans on synthetic pyrite. Arch.Microbiol. 149, 401–405.

Sand W., Gehrke, T., Hallman, R. & Schippers,A. (1995) Sulfur chemistry, biofilm, and the(in) direct attack mechanism – a critical evalua-tion of bacterial leaching. Appl. Microbiol.Biotechnol., 43, 961-966.

Schippers, A. (2004) Biogeochemistry of metalsulfide oxidation in mining environments, sedi-ments and soils. In:Amend, J.P., Edwards, K.J.,Lyons, T.W. (Eds.), Sulfur Biogeochemistry—Past and Present. Special Paper, vol. 379. Geo -logical Society of America, Boulder, Colorado,pp. 49-62.

Schippers, A., Kock, D., Schwartz, M., Bött cher,M.E., Vogel, H. & Hagger, M. (2007) Geomic -robiological and geochemical investigation of apyrrhotite-containing mine waste tailings damnear Selebi-Phikwe in Botswana. J. Geochem.Expl., 92, 151-158.

Yamamoto, A. & Nakazawa, H. (1982)Modulated structure of the NC-type (N = 5.5)pyrrhotite, Fe1-xS. Acta Cryst. A, 38(1), 79-86.


198


199

AAgné T. . . . . . . . . . . . . . . . . . . . . . 126Altermann W. . . . . . . . . . . . . . . . . . 65Arbeck D. . . . . . . . . . . . . . . . . . . . . 76Azzam R. . . . . . . . . . . . . . . . . . . . . 111

BBahr C. . . . . . . . . . . . . . . . . . . . . . 170Bayard E. . . . . . . . . . . . . . . . . . . . . 126Below M. . . . . . . . . . . . . . . . . . . . . 140Böhnke S. . . . . . . . . . . . . . . . . . . . . 34Bollschweiler C. . . . . . . . . . . . . . . . . . 3Bosbach D. . . . . . . . . . . . . . . . . . . . 76Burghardt D. . . . . . . . . . . . . . . . . . 154

CChust T. . . . . . . . . . . . . . . . . . . . . . 182

DDamian,C. . . . . . . . . . . . . . . . . . . . 154Darbha G.K.. . . . . . . . . . . . . . . . . . . 47Daus B. . . . . . . . . . . . . . . . . . . . . . 170Diedel R. . . . . . . . . . . . . . . 34, 96, 126Dombrowski I. . . . . . . . . . . . . . . . . 140Dörr H. . . . . . . . . . . . . . . . . . . . . . . 96Driehaus W. . . . . . . . . . . . . . . . . . . 170Drobek T. . . . . . . . . . . . . . . . . . . . . . 65Dultz S. . . . . . . . . . . . . . . . . . . . . . 140

EEhinger S. . . . . . . . . . . . . . . . . . . . 154Emmerich H. . . . . . . . . . . . . . . . . . 126Emmerich K. . . . . . . . . . . . . . . . . . 126Engels M.. . . . . . . . . . . . . . . . . . . . 126Ernst R. . . . . . . . . . . . . . . . . . . . . . 111

Etzel K. . . . . . . . . . . . . . . . . . . . . . 182Eulenkamp C. . . . . . . . . . . . . . . . . . 34

FFeinendegen M.. . . . . . . . . . . . . . . 111Fernández-Steeger T. M. . . . . . . . . 111Fischer R. . . . . . . . . . . . . . . . . . . . . . 3Fischer C. . . . . . . . . . . . . . . . . . . . . 47Fischer H. . . . . . . . . . . . . . . . . . . . . 19Fischer U. . . . . . . . . . . . . . . . . . . . . 76Frei M. . . . . . . . . . . . . . . . . . . . . . . 65

GGeiß P.L. . . . . . . . . . . . . . . . . . . . . . 96Gemming S. . . . . . . . . . . . . . . . . . . 19Glowacky J. . . . . . . . . . . . . . . . . . . 76Götz A. . . . . . . . . . . . . . . . . . . . . . 154Grefhorst C. . . . . . . . . . . . . . . . . . . 34Gutt B. . . . . . . . . . . . . . . . . . . . . . . . 3

HHaas S. . . . . . . . . . . . . . . . . . . . . . 126Habäck M. . . . . . . . . . . . . . . . . . . . 34Haist M. . . . . . . . . . . . . . . . . . . . . . 76Harries D. . . . . . . . . . . . . . . . . . . . 182Heberling F. . . . . . . . . . . . . . . . . . . . 76Heckl W.M. . . . . . . . . . . . . . . . . . . . 65Hellige K. . . . . . . . . . . . . . . . . . . . 182Hochella M.F., Jr. . . . . . . . . . . . . . . 182Hopf J. . . . . . . . . . . . . . . . . . . . . . 182Hsieh K. . . . . . . . . . . . . . . . . . . . . 154Huber J. . . . . . . . . . . . . . . . . . . . . . 76

JJanneck E. . . . . . . . . . . . . . . . . . . . 154

Author’s Index


200

Jennissen H.P. . . . . . . . . . . . . . . . . . 19Jordan G. . . . . . . . . . . . . . . . . . . . . 34

KKantioler M. . . . . . . . . . . . . . . . . . . 65Kappler A. . . . . . . . . . . . . . . . . . . 170Karabacheva S. . . . . . . . . . . . . . . . 170Kersten M. . . . . . . . . . . . . . . . . . . 170Kipry J. . . . . . . . . . . . . . . . . . . . . . 154Kirsten A. . . . . . . . . . . . . . . . . . . . . 19Koch T. . . . . . . . . . . . . . . . . . . . . . 154Koczur K. . . . . . . . . . . . . . . . . . . . . 19Kolbe F. . . . . . . . . . . . . . . . . . . . . . 170Kramar U. . . . . . . . . . . . . . . . . . . . . 76

LLangenhorst F. . . . . . . . . . . . . . . . 182Latief O. . . . . . . . . . . . . . . . . . 34, 126Lindner M. . . . . . . . . . . . . . . . . . . . 19Lüttge A. . . . . . . . . . . . . . . . . . . . . 47

MMartin M. . . . . . . . . . . . . . . . . . . . 154Meißner M. . . . . . . . . . . . . . . . . . . . 19Meister D. . . . . . . . . . . . . . . . . . . . . 34Meyer J. . . . . . . . . . . . . . . . . . . . . 154Michler A. . . . . . . . . . . . . . . . . . . . . 47Müller H.S. . . . . . . . . . . . . . . . . . . . 76Müller-Mai C. . . . . . . . . . . . . . . . . . 19

NNeher H.P. . . . . . . . . . . . . . . . . . . . 111Neumann T. . . . . . . . . . . . . . . . . . . 76

OObst U. . . . . . . . . . . . . . . . . . . . . . . . 3Oliveira A. . . . . . . . . . . . . . . . . . . . . 19Otte K. . . . . . . . . . . . . . . . . . . . . . 154

PPaikaray S. . . . . . . . . . . . . . . . . . . 154Peiffer S. . . . . . . . . . . . . . . . . 154, 182Pentcheva R. . . . . . . . . . . . . . . . . . 154Peuker M. . . . . . . . . . . . . . . . . . . . 126Phuong K.L. . . . . . . . . . . . . . . . . . . 65Pollok K. . . . . . . . . . . . . . . . . . . . . 182Posth N. . . . . . . . . . . . . . . . . . . . . 170Presser M. . . . . . . . . . . . . . . . . . . . . 96Pust C. . . . . . . . . . . . . . . . . . . . . . . 76

RReich T.Y. . . . . . . . . . . . . . . . . . . . 170Rieder A. . . . . . . . . . . . . . . . . . . . . . . 3Roth E. . . . . . . . . . . . . . . . . . . . . . . 96

SSchäfer T. . . . . . . . . . . . . . . . . . . . . 47Schäfer T. . . . . . . . . . . . . . . . . . . . . 76Schellhorn M. . . . . . . . . . . . . . . . . 140Schenk M.K. . . . . . . . . . . . . . . . . . 140Schlömann M. . . . . . . . . . . . . . . . 154Schmahl W. . . . . . . . . . . . . . . 34, 154Schmidt E. . . . . . . . . . . . . . . . 34, 140Schmilewski G. . . . . . . . . . . . . . . . 140Schöne G. . . . . . . . . . . . . . . . . . . . 154Schurk K. . . . . . . . . . . . . . . . . . . . 170Schwartz T. . . . . . . . . . . . . . . . . . . . . 3Seifert G. . . . . . . . . . . . . . . . . . . . . 19Spagnoli G. . . . . . . . . . . . . . . . . . . 111


201

Stanjek H. . . . . . . . . . . . . 34, 111, 170Stark R.W. . . . . . . . . . . . . . . . . . . . . 65Steinkemper U. . . . . . . . . . . . . . . . . 34Stelling J. . . . . . . . . . . . . . . . . . . . . 76Steudel A. . . . . . . . . . . . . . . . . . . . 126Strobel C. . . . . . . . . . . . . . . . . . . . . 65

VVinograd V. . . . . . . . . . . . . . . . . . . . 76Vucak M. . . . . . . . . . . . . . . . . . . . . 76Vuin A. . . . . . . . . . . . . . . . . . . . . . 126

WWalsch J. . . . . . . . . . . . . . . . . . . . 140Wang Z. . . . . . . . . . . . . . . . . . . . . 154Weh M. . . . . . . . . . . . . . . . . . . . . 111Wennrich R. . . . . . . . . . . . . . . . . . 170Wiacek C. . . . . . . . . . . . . . . . . . . . 154Winkler B. . . . . . . . . . . . . . . . . . . . . 76Wittwer W. . . . . . . . . . . . . . . . . . . . 96Wolff H. . . . . . . . . . . . . . . . . . . . . . 34

YYang H. . . . . . . . . . . . . . . . . . . . . 126

ZZiegler A. . . . . . . . . . . . . . . . . . . . 154Ziegler M. . . . . . . . . . . . . . . . . . . . 111Zoller J. . . . . . . . . . . . . . . . . . . . . . . . 3Zurlinden K. . . . . . . . . . . . . . . . . . . 19


202


203

GEOTECHNOLOGIEN Science Reports –Already published/Editions

No. 1 Gas Hydrates in the Geosystem – Sta -tus Seminar, GEOMAR Research CentreKiel, 6–7 May 2002, Programme &Abstracts, 151 pages.

No. 2 Information Systems in Earth Mana -gement – Kick-Off-Meeting, Universityof Hannover, 19 February 2003, Pro -jects, 65 pages.

No. 3 Observation of the System Earth fromSpace – Status Seminar, BLVA Munich,12–13 June 2003, Programme & Ab -stracts, 199 pages.

No. 4 Information Systems in Earth Mana -gement – Status Seminar, RWTHAachen University, 23–24 March 2004,Programme & Abstracts, 100 pages.

No. 5 Continental Margins – Earth’s FocalPoints of Usage and Hazard Potential– Status Seminar, Geo For schungs Zen -trum (GFZ) Potsdam, 9–10 June 2005,Programme & Abstracts, 112 pages.

No. 6 Investigation, Utilization and Protec -tion of the Underground – CO2-Stor -age in Geological Formations, Tech -nologies for an Underground SurveyAreas – Kick-Off-Meeting, Bundes an -stalt für Geowissenschaften und Roh -stoffe (BGR) Hannover, 22–23 Sep tem -ber 2005, Programme & Ab stracts,144 pages.

No. 7 Gas Hydrates in the Geosystem – TheGerman National Research Pro gram -me on Gas Hydrates, Results from theFirst Funding Period (2001–2004), 219 pages.

No. 8 Information Systems in Earth Mana -gement – From Science to Application,Results from the First Funding Period(2002–2005), 103 pages.

No. 9 1. French-German Symposium on Geo -logical Storage of CO2, Juni 21./22.2007, GeoForschungsZentrum Pots -dam, Abstracts, 202 pages.

No. 10 Early Warning Systems in Earth Mana -gement – Kick-Off-Meeting, TechnicalUniversity Karlsruhe, 10 October 2007,Programme & Abstracts, 136 pages.

No. 11 Observation of the System Earth fromSpace – Status Seminar, 22–23 Novem -ber 2007, Bavarian Academy of Scien -ces and Humanities, Munich, Pro -gramme & Abstracts, 194 pages.

No. 12 Mineral Surfaces – From Atomic Pro -cesses to Industrial Application – Kick-Off-Meeting, 13–14 October 2008,Ludwig-Maximilians Universtität, Mu -nich, Programme & Abstracts, 133pages.


No. 13 Early Warning Systems in Earth Mana -gement – Status Seminar, 12–13October 2009 Technische UniversitätMünchen, Programme & Abstracts,165 pages.

No. 14 Die dauerhafte geologische Spei che -rung von CO2 in Deutschland – Ak tu -elle Forschungsergebnisse und Pers -pek tiven, Herausgegeben von: LudwigStroink, J. Peter Gerling, MichaelKühn, Frank R. Schilling, 140 Seiten.

No. 15 Early Warning Systems for Trans -portation Infrastructures, Workshop9-10 February 2009, Fraunhofer IITBKarlsruhe, Karlsruhe Institute of Tech -nology (KIT), 160 p.

204



Status Seminar26-27 October 2010Johannes-Gutenberg-UniversitätMainz



No. 16


ISSN: 1619-7399

In the frame of the R&D-programme GEOTECHNOLOGIEN thirteen joint interdiscipli-nary projects have been launched in April 2008. The object of research is the betterunderstanding of the multiple physical and chemical reactions at mineral surfaces.The abstract volume summarizes the scientific results presented during the statusseminar at the Johannes Gutenberg University in Mainz, Germany in October 2010.

Research activities focus on different applications of geomaterials:

– reactions on external and internal interfaces (e. g., immobilization of toxic ele-ments/compounds and radioactive irradiated phases) and bio-mineralization suchas biomimetic functional materials as well as bone implants.

– development of high technology and its applications for the exploration of theEarth's interior on the one hand and the synthesis of new materials (e. g. defor-mation hardening, ultra hard phases) on the other.

The GEOTECHNOLOGIEN programme is funded by the Federal Ministry for

Education and Research (BMBF) and the German Research Council (DFG)

No.

16

Min

eral

Sur

face

s –

From

Ato

mic

Proc

esse

s to

Indu

stria

l App

licat

ion

GEO

TECH

NO

LOG

IEN

Scie

nce

Repo

rt

Umschlag_SR16.qxd:Umschlag_Infosyst.II.qxd 30.09.10 13:44 Seite 1

Documents

GEOTECHNOLOGIEN Science Report - gfz-potsdam.degfzpublic.gfz-potsdam.de/pubman/item/escidoc:42034:3/component/... · GEOTECHNOLOGIEN Science Report No. 16 Mineral Surfaces – From