Embed Size (px)
Citation preview
Technische Universität Dresden
Fakultät Umweltwissenschaften
Investigation of a polyether trisiloxane surfactant: Environmental fate and homologue specific trace analysis from surface water
Dissertation zur Erlangung des akademischen Grades
Doctor rerum naturalium (Dr. rer. nat.)
Vorgelegt von
M. Sc. Amandine Michel
Geboren am 11. November 1986 in Metz
Gutachter:
Herr Prof. Dr. Eckhard Worch, Technische Universität Dresden
Herr Prof. Dr. Heinz-Jürgen Brauch, DVGW-Technologiezentrum Wasser Karlsruhe
Herr Prof. Dr. Martin Jekel, Technische Universität Berlin
Tag der Verteidigung: 28.07.2015
i
Acknowledgments
I wish to express my sincere thanks to Prof. Brauch for allowing me to work in his department
at the DVGW–Technologiezentrum Wasser (TZW) and to Prof. Worch for supervising this
PhD and having allowed me to be a PhD student at the TU Dredsen – Institute of Water
Chemistry. Without their support this work would not have been possible. I am grateful to Dr.
Frank Thomas Lange for his support throughout this work. Thank you for keeping the door
open and for always finding some time for valuable open discussions.
I acknowledge funding from the German Association for Gas and Water (DVGW - Deutscher
Verein des Gas- und Wasserfaches e.V.), and the travelling grant from the GDCh –
Gesellschaft Deutscher Chemiker e.V..
Thanks to all the people who contributed to this work: Dr. Markus Godejohann and Li-Hong
Tseng from Brucker BioSpin, Prof. Michael Burkhardt, Conrad Dietschweiler and Martina
Böni from the UMTEC, people from CHT R. Beitlich for their cooperation and for providing
HANSA ADD 1055, and last but not least Dr. Oliver Happel for the great picture of
superspreading.
I originally did not have in mind to make a PhD and only during my master thesis I realized
that research can be exciting. A special thanks to Dr. Scheurer and Dr. Sacher.
Special thanks to all the people in TZW for their friendly welcome and for giving me the wish
to stay longer in Germany: Carola, the two Michaels, Andrea, Birgit, Brigitte, Thomas, Beat,
Martin, Ralf, Sybilla, Hans, Astrid… and all the others.
I have learnt a lot during this PhD and not only about silicones, HPLC-MS/MS, and
environmental fate. When I decided to move to Karlsruhe and start an internship in Germany,
it was because I was forced to for my diploma. In the end, the PhD and the life in Germany
have been very positive experiences and it made me realize that life is unexpected and full of
surprises. Special thanks to the people who made this PhD such a nice time:
Stephany, thank you for all the good moments we spent together. If Karlsruhe was a very
positive experience it’ s in a great part thanks to you.
Doro, the best flat mate! Thank you for your kind support, for all the nice moments together,
games, and discussions. Thank you for making flat sharing such a positive experience.
ii
Sarah, for your friendly help during loooong sampling trips on the Neckar, and for supporting
me even when I was in bad mood!
Florian for all the times I could just pop up your office and complain that nothing works! It
may have been boring for you but it has helped me a lot.
Brigitte for dinners, support during the time in TZW and for giving me the wish to travel
more and more!
To my family and friends,
My parents
Jouni
iii
ABSTRACT
Thanks to their adaptability and high efficiency compared to traditional carbon based
surfactants, silicone surfactants are a success in many different applications, from pesticides
to cosmetics, polyurethane foam, textile and car care products. In spite of those numerous
applications, no analytical method existed for their trace determination in environmental
samples and no data have been available regarding their environmental occurrence and fate.
An analytical method for the trace analysis of trisiloxane surfactants in the aqueous
environment was developed and validated. The method, based on liquid-liquid extraction and
HPLC-MS/MS, reaches limits of quantification in the ng L-1 range and allows an individual
quantification of every homologue of the targeted trisiloxane surfactant. The newly developed
analytical method was applied to analyze 40 river water samples. The targeted trisiloxane
surfactant was detected in 14 samples, between 1 ng L-1 and 100 ng L-1. The results showed
that the studied trisiloxane surfactant does not ubiquitously occur in the aquatic environment
in measurable concentrations, but can reach surface waters on a local scale.
In order to assess the persistence of the trisiloxane surfactant in surface waters, its hydrolysis
was studied in the lab, under various conditions (temperature, pH, and concentration). The
half-lifes at pH 7 and 2 mg L-1 were found to be between 29 days and 55 days at 25°C and
between 151 days and 289 days at 12°C. Taking only into account the hydrolysis, these
results indicate that the trisiloxane surfactant could persist several weeks in surface waters. A
degradation product of the trisiloxane surfactant was tentatively identified by high resolution
mass spectrometry.
When used as agricultural adjuvants, trisiloxane surfactants may reach the soil compartment
and might further leach to ground water. The behavior of the trisiloxane surfactant on soil was
therefore investigated to assess the possibility to reach ground water. With a sorption batch
equilibrium method, distribution coefficients between water and soil (Kd, Koc, and Kclay) were
estimated for two standard soils (loam and sandy loam) and for every homologue of the
trisiloxane surfactant. The obtained values for Kd were between 15 L kg-1 and 135 L kg-1,
indicating that the trisiloxane surfactant is only slightly mobile in soil. To further investigate
the possibility of leaching to ground water after application on agricultural fields, the leaching
in soil was simulated in the lab in a soil column. The experimental settings were designed to
simulate a worst case scenario where the application of the trisiloxane surfactant is done on
quartz sand and is immediately followed by a heavy rainfall. Even in these conditions, less
than 0.01 % of the initially applied trisiloxane surfactant leached through 20 cm of quartz
sand. Based on the Kd values and the results of the leaching in soil column, the studied
iv
trisiloxane surfactant is considered to be unlikely to leach to ground water after application as
an agricultural adjuvant.
v
TABLE OF CONTENTS
ABSTRACT .............................................................................................................................. iii
TABLE OF CONTENTS ........................................................................................................... v
LIST OF PUBLICATIONS .................................................................................................... viii
LIST OF FIGURES .................................................................................................................... x
LIST OF TABLES .................................................................................................................. xiii
ABBREVIATIONS ................................................................................................................. xiv
CHAPTER 1: General introduction ........................................................................................... 1
1. The challenge of water resources protection .......................................................................... 3
2. Assessing the environment risk of a chemical ....................................................................... 4
2.1. Environmental risk assessment ....................................................................................... 4
2.2. Ecotoxicity ...................................................................................................................... 4
2.3. Predicting the potential exposure to a chemical ............................................................. 4
3. Organosilicon compounds in the environment ....................................................................... 9
3.1. Definitions and applications ........................................................................................... 9
3.2. Organosilicon compounds in the environment ............................................................. 10
3.2.1. Environmental occurrence and fate of volatile methylsiloxanes .......................... 10
3.2.2. Environmental occurrence and fate of polydimethylsiloxanes ............................. 12
3.2.3. Environmental occurrence and fate of polyethermethylsiloxanes ........................ 13
4. Silicone surfactants .............................................................................................................. 14
4.1. Structure, properties, and applications ......................................................................... 14
4.2. The targeted trisiloxane surfactant ............................................................................... 17
4.2.1. Entry routes into the environment and expected environmental fate ................... 17
4.2.2. Occurrence in the environment ............................................................................. 19
4.2.3. Toxicty and ecotoxicity ........................................................................................ 20
5. Objectives ............................................................................................................................. 24
References ................................................................................................................................ 26
CHAPTER 2: Development of a liquid chromatography tandem mass spectrometry method
for trace analysis of trisiloxane surfactants in the aqueous environment: an alternative strategy
for quantification of ethoxylated surfactants ............................................................................ 34
Abstract .................................................................................................................................... 36
1. Introduction .......................................................................................................................... 37
2. Experimental ........................................................................................................................ 40
2.1. Reagents and material ................................................................................................... 40
2.2. Sample preparation ....................................................................................................... 40
2.3. Liquid chromatography-mass spectrometry ................................................................. 40
2.4. Method validation ......................................................................................................... 41
3. Results and discussion .......................................................................................................... 43
3.1. Characterization of the stock solution .......................................................................... 43
3.1.1. Theoretical approach: Poisson distribution ........................................................... 43
vi
3.1.2. Experimental data: single mass spectrometry ....................................................... 44
3.2. Liquid chromatography ................................................................................................ 48
3.3. Validation ..................................................................................................................... 51
Conclusion ................................................................................................................................ 54
Acknowledgment ..................................................................................................................... 55
References ................................................................................................................................ 55
Appendix A. Supplementary data ............................................................................................ 58
A.1 Theoretical approach: Poisson distribution .................................................................. 58
A.2 Experimental approach: LC-SPE-NMR/MS ................................................................ 59
A.2.1. Materials and method ........................................................................................... 59
A.2.2. Results and discussion ......................................................................................... 59
References ................................................................................................................................ 62
CHAPTER 3: Homologue specific analysis of a polyether trisiloxane surfactant in German
surface waters and study on its hydrolysis ............................................................................... 63
Abstract .................................................................................................................................... 65
1. Introduction .......................................................................................................................... 66
2. Experimental ........................................................................................................................ 68
2.1. Chemicals and standards .............................................................................................. 68
2.2. Sample preparation and analysis by liquid chromatography-mass spectrometry......... 68
2.3. Sampling method .......................................................................................................... 69
2.4. Quality control .............................................................................................................. 70
2.4.1 Calibration blanks .................................................................................................. 70
2.4.2 Laboratory blanks .................................................................................................. 70
2.4.3 Calibration ............................................................................................................. 70
2.4.4 Limits of quantification ......................................................................................... 70
2.4.5 Isobaric interference .............................................................................................. 70
2.4.6 Matrix effects ......................................................................................................... 71
2.5. Degradation by hydrolysis ............................................................................................ 71
3. Results and discussion .......................................................................................................... 72
3.1. Environmental samples................................................................................................. 72
3.2. Degradation by hydrolysis ............................................................................................ 79
3.2.1 Initial recoveries .................................................................................................... 79
3.2.2 Degradation kinetics .............................................................................................. 79
Conclusion ................................................................................................................................ 86
Acknowledgments .................................................................................................................... 87
References ................................................................................................................................ 87
Appendix A: Supplementary data ............................................................................................ 90
A.1 Discussion on the choice of container for hydrolysis study as a function of pH.......... 90
A.2 Description of the identification of degradation products by HPLC-ESI-TOF ............ 95
CHAPTER 4: Mobility of a polyether trisiloxane surfactant on soil: soil/water distribution
coefficients and leaching in a soil column ............................................................................... 97
Abstract .................................................................................................................................... 99
1. Introduction ........................................................................................................................ 100
2. Materials and methods ....................................................................................................... 102
2.1. Chemicals ................................................................................................................... 102
vii
2.2. HPLC-MS/MS analysis .............................................................................................. 102
2.3. Soil extraction ............................................................................................................. 103
2.4. Sorption experiments .................................................................................................. 103
2.4.1 Soils ..................................................................................................................... 103
2.4.2 Preliminary experiment ........................................................................................ 104
2.4.3 Sorption kinetics .................................................................................................. 104
2.4.4 Sorption isotherms ............................................................................................... 104
2.5. Leaching in soil column ............................................................................................. 104
3. Results and discussion ........................................................................................................ 106
3.1. Sorption using a batch equilibrium method ................................................................ 106
3.1.1 Preliminary experiment ........................................................................................ 106
3.1.2 Sorption kinetics .................................................................................................. 106
3.1.3 Sorption isotherms ............................................................................................... 109
3.2. Leaching in soil column ............................................................................................. 114
3.2.1 Acesulfame as a tracer ......................................................................................... 114
3.2.2 The trisiloxane surfactant HANSA ADD 1055 ................................................... 115
Conclusion .............................................................................................................................. 119
Acknowledgments .................................................................................................................. 119
References .............................................................................................................................. 120
Appendix A ............................................................................................................................ 122
References .............................................................................................................................. 132
SUMMARY AND CONCLUSION ....................................................................................... 133
viii
LIST OF PUBLICATIONS
PEER REVIEWED PUBLICATIONS
Michel, A.; Brauch, H.-J.; Worch, E.; Lange, F. T.; 2012. Development of a
liquid chromatography tandem mass spectrometry method for trace analysis
of trisiloxane surfactants in the aqueous environment: An alternative
strategy for quantification of ethoxylated surfactants. Journal of
Chromatography A 1245, 46-54.
Michel, A.; Brauch, H.-J.; Worch, E.; Lange, F. T.; 2014. Homologue
specific analysis of a polyether trisiloxane surfactant in German surface
waters and study on its hydrolysis. Environmental Pollution 186, 126-135.
CONFERENCE/WORKSHOP PROCEEDINGS
Michel, A.; Böni, M.; Dietschweiler, C.; Worch, E.; Brauch, H.-J.; Lange,
F. T.; 2013. Trisiloxane surfactants: trace analysis in the aquatic
environment and first insight into their environmental behavior. Vom
Wasser 111, 67-102.
Michel, A.; Böni, M.; Dietschweiler, C.; Worch, E.; Brauch, H.-J.; Lange,
F. T.; 2013. Trisiloxane surfactants: trace analysis in the aquatic
environment and first insight into their environmental behavior. Wasser
2013, Goslar, Abstract Book, 146-149.
Michel, A.; Worch, E.; Brauch, H.-J.; Lange, F. T.; 2013. First method for
the trace analysis of trisiloxane surfactants in surface waters and first
detection in river water. 8th Micropol & Ecohazard, Zurich, Programme and
Abstract Book, 64-65.
Michel, A.; Worch, E.; Brauch, H.-J.; Lange, F. T.; 2012. A method for
trace analysis of trisiloxane surfactants in aqueous environmental samples.
Wasser 2012, Neu-Ulm, Abstract Book, 175-179.
CONFERENCES/WORKSHOP CONTRIBUTIONS
Oral presentation at the Micropol and Ecohazard, Zürich, Switzerland, June
17-20, 2013.
Oral presentation at the Jahrestagung der Wasserchemischen Gesellschaft,
Goslar, Germany, May 06-08, 2013.
Oral presentation at the ANAKON, Essen, Germany, March 04-07, 2013.
Oral presentation at the Forum Wasseraufbereitung, Karlsruhe, Germany,
October 16, 2012.
Poster at the Society of Environmental Toxicology and Chemistry (SETAC)
World Congress, Berlin, Germany, May 20-24, 2012.
ix
Poster at the Jahrestagung der Wasserchemischen Gesellschaft, Ulm,
Germany, Mai 14-16, 2012.
Oral presentation at the 3rd Workshop on Organosilicones in the
Environment, Burlington, Ontario, Canada, May 8-9, 2012.
x
LIST OF FIGURES
CHAPTER 1
Figure 1: Schematic representation of the distribution of a chemical between the four
environmental compartments.
Figure 2: The steps of an environmental risk assessment.
Figure 3: Chemical structures of the three categories of organosilicon compounds having
noteworthy environmental loading.
Figure 4: Schematic representation of PDMS degradation by depolymerization, until the
dimethylsilanediol.
Figure 5: General chemical structure of the most common silicone surfactants.
Figure 6: (a) Wetting of a leaf by water with (lower part) and without (upper part)
trisiloxane surfactant. (b) Proposed situation at the spreading edge of an aqueous
droplet of trisiloxane surfactant on a hydrophobic surface.
Figure 7: Daphnia magna (a, image credit: Hajime Watanabe) and pimphales promelas (b,
image credit: Joseph Tomelleri)
CHAPTER 2
Figure 1: General chemical structure of silicone surfactants.
Figure 2: Schematic representation of the synthesis route of trisiloxane surfactants.
Figure 3: Mass spectra obtained by direct infusion of the trisiloxane surfactants at 0.5 mg L-
1 in MeOH/20 mM aqueous ammonium acetate solution 1/1 (v/v). a: MD (́-
(CH2)3-(EO)n-OH)M. b: MD (́-(CH2)3-(EO)n-O CH3)M.
Figure 4: Comparison of the individual concentrations of homologues in the stock solution
calculated with theoretical ( ) and experimental ( ) approaches. a: MD (́-(CH2)3-
(EO)n-OH)M. b: MD (́-(CH2)3-(EO)n-O CH3)M. The numbers on top of each
point give the absolute percentage difference between the results obtained by the
two approaches.
Figure 5: HPLC-MS/MS extracted ion chromatograms of a mixture MD (́-(CH2)3-(EO)n-
OH)M and MD (́-(CH2)3-(EO)n-OCH3)M at 0.1 mg L-1 (total surfactant
concentration) on two different Kinetex C18 columns. a: new Kinetex C18. b:
used Kinetex C18.
Figure 6: HPLC-MS/MS chromatograms of a mixture of the trisiloxane surfactants at
0.1 mg-1 (total surfactant concentration) with a PolymerX column. A: MD (́-
(CH2)3-(EO)n-OH) M. B: MD (́-(CH2)3-(EO)n-OCH3)M.
Figure S 1: 1H-NMR spectra between 0 ppm and 4 ppm for homologue n = 6, with peaks
attribution and integration.
Figure S 2: Comparison of the oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M
determined by theoretical Poisson distribution ( ), MS measurements ( ) and LC-
SPE-NMR/MS ( ).
CHAPTER 3
Figure 1: Chemical structures of the silicones considered to have noteworthy environmental
loading.
Figure 2: Sampling sites on the Neckar River and two of its tributaries, between Rottenburg
am Neckar and Besigheim.
xi
Figure 3: Oligomeric concentrations of MD (́-(CH2)3-(EO)n-OCH3)M in the Neckar River at
Esslingen (sampled on 16.07.12) determined with external calibration of the entire
method including sample preparation and with standard addition.
Figure 4: Oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M in the standard solution
at 75 ng L-1 ( ) and in the river water samples collected on the 16.07.12: ( )
Neckar at Deizisau, ( ) Neckar at Esslingen, ( ) Neckar at Poppenweiler, and ( )
Fils at Plochingen.
Figure 5: Graphical representation of ln(C/C0) as a function of time for the homologues
n = 6, n = 8, n = 10, and n = 12, for pH 7 and 9 and temperature tested in the
study: ( ) 12°C, ( ) 25°C, and ( ) 50°C.
Figure 6: Comparison of the hydrolysis rates of MD (́-(CH2)3-(EO)8-OCH3)M in buffers at
pH 7 and Ph 9 and in sterilized river water (filtrated and non-filtrated)
Figure 7: Evolution of the oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M during
the hydrolysis test at pH 9, 12°C.
Figure 8: Proposed hydrolysis reaction of MD (́-(CH2)3-(EO)n-OCH3)M and relative
evolution of the concentrations of the trisiloxane surfactant homologue MD (́-
(CH2)3-(EO)4-OCH3)M ( ) and the identified degradation product ( ) over time
for the experiment at pH 9, 12°C.
Figure S 1: Initial recoveries plotted over the air/water contact surface.
Figure S 2: Graphical representation of ln(C/C0) as a function of time for the homologues
n = 5, n = 7, n = 9, and n = 11, for all conditions of pH and temperature tested in
the study: ( ) 12°C, ( ) 25°C, and ( ) 50°C.
Figure S 3: Identified degradation products of MD (́-(CH2)3-(EO)n-OCH3)M.
Figure S 4: A: HPLC-MS/MS chromatograms of direct injections of MD (́-(CH2)3-(EO)n-
OCH3)M at 20 µg L-1 (Figure S4 a) and Triton X-100 at 1 mg L-1 (Figure S4 b).
CHAPTER 4
Figure 1: (a) Chemical structure of the three main trisiloxane surfactants used in agriculture.
M stands for (CH3)3-Si-O1/2-, D’ for O1/2-Si(CH3)(R )́-O1/2-, and EO for
OCH2CH2. (b) Wetting of a leaf by water with (lower part) and without (upper
part) trisiloxane surfactant.
Figure 2: Distribution of every homologue of MD (́-(CH2)3-(EO)n-OCH3)M between the
aqueous phase and the soil together with the mass balance on soil 2.4 and 5M.
Figure 3: Sorption isotherm of MD (́-(CH2)3-(EO)7-OCH3)M on soil 2.4 and 5M.
Figure 4: Relative breakthrough curve (Figure 4a) and cumulated breakthrough curve
(Figure 4b) for acesulfame.
Figure 5: Relative breakthrough curve (Figure 5a) and cumulated breakthrough curve
(Figure 5b) for MD (́-(CH2)3-(EO)7-OCH3)M.
Figure 6: Distribution of MD (́-(CH2)3-(EO)7-OCH3)M in different segments of the soil
column. Duration of the experiment: 6 d, volumetric flow rate: 11.2 mL min-1,
total amount of water applied: 51 L.
Figure S 1: MD (́-(CH2)3-(EO)n-OCH3)M sorption kinetics on soils 2.1, 2.4, and 5M at three
soil/solution ratios (1/10, 1/25, and 1/50) as determined during the preliminary
experiments.
Figure S 2: Partition of MD (́-(CH2)3-(EO)n-OCH3)M between the aqueous phase and the soil
and mass balance for all homologues on soil 2.4.
Figure S 3: Partition of MD (́-(CH2)3-(EO)n-OCH3)M between the aqueous phase and the soil
and mass balance for all homologues on soil 5M.
Figure S 4: Sorption isotherms for all homologues of MD (́-(CH2)3-(EO)n-OCH3)M on soil
2.4.
xii
Figure S 5: Sorption isotherms for all homologues of MD (́-(CH2)3-(EO)n-OCH3)M on soil
5M.
Figure S 6: Acesulfame concentration and conductivity measured on side samples at 5 cm and
35 cm depth during a preliminary experiment with a 70 cm long quartz sand
column.
Figure S 7: Acesulfame concentration in the side samples taken at 5 cm, 10 cm, and 15 cm.
xiii
LIST OF TABLES
CHAPTER 1
Table 1: Classes of affinity of organic chemicals for the different environmental
compartments in relation to the physico-chemical characteristics of the substance.
Table 2: Comparison of the log Kow values for a few PEMS and PDMS.
Table 3: Toxicity of PEMS compounds to various aquatic organisms.
CHAPTER 2
Table 1: Description of the target trisiloxane surfactants: abbreviations, chemical structures
and CAS numbers.
Table 2: Precursor ions, product ions and corresponding MS parameters (DP declustering
potential, CE collision energy, CXP cell exit potential).
Table 3: Results of the validation: LOQs, correlation coefficients, sample preparation
recovery, RSDs and matrix effects for all analytes.
CHAPTER 3
Table 1: Concentrations of individual homologues of MD (́-(CH2)3-(EO)n-OCH3)M in
different environmental samples collected in 2012 and 2013.
Table 2: Hydrolysis rate constants, reaction half-life, correlation coefficients, and
activation energy for the hydrolysis of MD (́-(CH2)3-(EO)n-OCH3)M under the
different conditions tested in this study; na: not applicable.
Table S 1: Characteristics of the tested containers and initial recoveries (average over two
replicates ± standard deviation) for aqueous solutions of MD (́-(CH2)3-(EO)n-
OCH3)M at 2 mg L-1.
CHAPTER 4
Table 1: Sorption coefficient (Kd in L kg-1), organic carbon normalized sorption coefficient
(Koc in L kg-1), clay content normalized sorption coefficient (Kclay in L kg-1),
Freundlich sorption coefficient (KadsF µg1-1/n(cm³)1/ng-1), regression constants
(nads), and correlation coefficients (Rads²) for all homologues of MD (́-(CH2)3-
(EO)n-OCH3)M on both soils studied.
Table S 1: Limits of quantification obtained by direct injection on the HPLC-MS/MS system
and after liquid-liquid extraction.
Table S 2: Parameters of the chromatographic separations performed during analysis of
MD (́-(CH2)3-(EO)n-OCH3)M and acesulfame by HPLC-MS/MS.
Table S 3: Characteristics of the soils used in the sorption experiments.
Table S 4: Characteristics of the quartz sand used for the leaching in soil column
experiments.
xiv
ABBREVIATIONS
AAS Atomic Absorption Spectrometry
ACN Acetonitrile
AOF Adsorbable Organic Fluorine
AOX Adsorbable Organic Halogens
BCF Bioconcentration Factor
CE Collision Energy
CES European Silicone Center
CFTA Cosmetic, Toiletry and Fragrance Association
CMC Critical Micellar Concentration
cVMS Cyclic Volatile Methylsiloxanes
CXP Cell Exit Potential
D4 Octamethylcyclotetrasiloxane
D5 Decamethylcyclopentasiloxane
D6 Dodecamethylcyclohexasiloxane
DIN Deutsches Institut für Normung
DMSD Dimethylsilanediol
DP Declustering Potential
DVGW Deutscher Verein des Gas- und Wasserfaches e.V.
ECz Effective Concentration at z %
EO Ethylene Oxide
EPA Environment Protection Agency
EQC Equilibrium Criterion Model
ESI Electrospray Ionization
EU European Union
H Henry’ s constant
HPLC High Performance Liquid Chromatography
INCI International Nomenclature of Cosmetics Ingredients
LC Liquid Chromatography
LCx Lethal Concentration x
LDPE Low Density Polyethylene
LOD Limit of Detection
LOQ Limit of Quantification
LUFA Landwirtschaftliche Untersuchungs- und Forschungsanstalt
MeOH Methanol
MRM Multiple Reaction Mode
MS Mass spectrometry
MW Molecular weight
MWD Molecular Weight Distribution
NEC No Effect Concentration
NEL No Effect Level
NMR Nuclear Magnetic Resonance
NOEC No Observed Effect Concentration
NOESY Nuclear Overhauser Effect Spectroscopy
NPEO Nonylphenol Ethoxylates
OECD Organization for Economic Co-operation and Development
PDMS Polydimethylsiloxanes
PE Polyethylene
xv
PEC Predicted Environmental Concentration
PEG Polyethylene Glycol
PEMS Polyethermethylsiloxanes
PFA Perfluoroalkoxy Polymer
pH Potentia Hydrogenii
PNEC Predicted No Effect Concentration
PO Propylene oxide
PP Polypropylene
PPG Polypropylene glycol
PTFE Polytetrafluoroethylene
RSD Relative Standard Deviation
S Solubility
SPE Solid Phase Extraction
STP Sewage Treatment Plant
T Temperature
THF Tetrahydrofurane
TMB Trimethoxybenzene
TMS Trimethylsilanediol
TOC Total Organic Carbon
TOF Time Of Flight
UV Ultra Violet
VMS Volatile Methylsiloxanes
WWTP Wastewater Treatment Plant
wt% Si Weight percentage of Si in the polyethermethylsiloxane product
Chapter 1
1
Chapter 1
General introduction
Chapter 1
2
TABLE OF CONTENTS
1. The challenge of water resources protection ...................................................................... 3
2. Assessing the environmental risk of a chemical ................................................................ 4
2.1. Environmental risk assessment .................................................................................... 4
2.2. Ecotoxicity ................................................................................................................... 4
2.3. Predicting the potential exposure to a chemical .......................................................... 4
3. Organosilicon compounds in the environment................................................................... 9
3.1. Definitions and applications ........................................................................................ 9
3.2. Organosilicon compounds in the environment .......................................................... 10
3.2.1. Environmental occurrence and fate of volatile methylsiloxanes ....................... 10
3.2.2. Environmental occurrence and fate of polydimethylsiloxanes .......................... 12
3.2.3. Environmental occurrence and fate of polyethermethylsiloxanes ..................... 13
4. Silicone surfactants .......................................................................................................... 14
4.1. Structure, properties, and applications ...................................................................... 14
4.2. The targeted trisiloxane surfactant ............................................................................ 17
4.2.1. Entry routes into the environment and expected environmental fate ................. 17
4.2.2. Occurrence in the environment .......................................................................... 19
4.2.3. Toxicity and ecotoxicity ..................................................................................... 20
5. Objectives ......................................................................................................................... 24
References ................................................................................................................................ 26
Chapter 1
3
1. The challenge of water resources protection
Access to clean drinking water and sanitation is recognized as a human right by the United
Nations since 2010. Indeed, in Western Europe almost 100 % of the population has access to
drinking water that meets the World Health Organization quality standards (World Health
Organization and United Nations Children’s Fund, 2000). Reaching this high quality has only
been possible through continuous, substantial investment in water research and protection;
continuing this investment is essential to assure the availability of high-quality drinking water
for the community also in the future.
One main challenge for providing clean drinking water arises from the chemical pollution of
the raw waters used for drinking water production (Schwarzenbach et al., 2006; Jekel et al.,
2013). We use in our everyday life an enormous amount of products from the chemical and
pharmaceutical industries. In the EU, more than 100 000 chemicals are registered and
between 30 000 and 70 000 are in daily use (Schwarzenbach et al., 2006). Many of these
chemicals enter natural waters as a consequence of release during their production, storage,
transport, use, and/or disposal. For example, pharmaceutical and personal care products end
up in wastewaters and wastewater treatment plants (WWTPs) after consumer use (Daughton
and Ternes, 1999). However, the commonly applied treatments fail to remove all of them, and
consequently, some of them are released to the aquatic environment within WWTPs effluents.
Pesticides enter the environment by other patterns: after application on agricultural fields,
they may reach natural waters via leaching or runoff for example (Vasiljevic et al., 2013).
The entry paths of chemicals into the environment are diverse but the common result is the
ubiquitous contamination of natural waters by synthetic chemicals. The numerous substances
of anthropogenic origin present in water at low concentrations (µg L-1 or ng L-1) are referred
to as micropollutants (Kümmerer, 2011). Pharmaceuticals, personal care products, pesticides,
contrast media, surfactants, perfluorinated compounds, or engineered nanomaterials
(Kümmerer, 2013) are examples of micropollutants commonly found in natural waters.
Chapter 1
4
2. Assessing the environmental risk of a chemical
2.1. Environmental risk assessment
A compound’s risk for the environment is evaluated as the combination of its toxicity and
exposure. Toxicity is defined as the hazard of a substance which can cause poisoning and
exposure as the amount of chemical a given organism may be exposed to. Both factors are
required for a compound to represent an environmental risk. A highly toxic compound does
not represent risk if the exposure is negligible, nor does excessive exposure to a compound
with virtually no toxicity. On the contrary, a compound with low toxicity but excessive
exposure can represent a risk, as can a compound with high toxicity and infrequent exposure.
2.2. Ecotoxicity
Ecotoxicology is the science of contaminants in the biosphere and their effects on constituents
of the biosphere (Newman and Unger, 2002). In order to determine the effects of a given
chemical on the constituents of the biosphere, a variety of toxicity tests have been derived
(Jørgensen, 2010). Some of them involve experiments with subsets of natural systems like
microcosms for example, but the majority of them have been confined to studies in the
laboratory on a limited number of test species. The aim is to answer two key problematics: i)
Which hazards are associated with the application of the chemical? Answering this question
involves gathering data on neurological, mutagenic, endocrine disruption, and carcinogenic
effects. ii) What is the relation between dose and response? This implies knowledge of no
effect concentration (NEC), LCx values (the concentration that is lethal to x % of the
organisms considered), and ECz values (the concentration giving the indicated effect to z % of
the organisms considered). Based on those values, the most probable level of no effect, no
effect level (NEL) is assessed. The extrapolation of experimental laboratory data to real
situations implies a certain level of uncertainty, taken into account by applying safety factors
(typically from 50 to 100) to the NEL. This leads to the commonly used PNEC (predicted no
effect concentration), derived for the different environmental compartments (water, soil, air,
biota, and sediments).
2.3. Predicting the potential exposure to a chemical
The potential exposure to a chemical depends on its persistence and distribution patterns in
the environment. When a compound is released into the environment, it is subject to different
processes that can be classified into two categories: the processes that leave the structure of
Chapter 1
5
the chemical unchanged (transport and mixing within an environmental compartment, transfer
between the different compartments), and those that transform the chemical (chemical,
photochemical and biological transformations). Understanding the environmental fate of a
compound therefore requires characterizing the processes by which the chemical moves and is
transformed in the environment (United States Environmental Protection Agency, 2013;
Schwarzenbach et al., 2003).
Figure 1: Schematic representation of the distribution of a chemical between the four environmental
compartments. (Linde, 1994)
To understand and predict the distribution of a compound in the environment, a few key
parameters, representing partitioning of the compound between the four environmental
compartments (air, water, soil, and biota) in equilibrium situations are commonly used. In real
situations, partitioning equilibrium is often not reached but those parameters are very useful to
determine a compound’s tendency to accumulate in particular environmental compartments
(Hites and Raff, 2012; Schwarzenbach et al., 2003).
The Henry’s law constant (H) is the ratio of a compound’s partial pressure above water
to its water solubility. The higher the Henry’s law constant, the more likely the
compound volatilizes from water into air.
The solubility, S, is a measure of the maximum amount of chemical that can be
dissolved in water.
Sorption on soil or sediments is measured by the solid/water partition coefficient Kd,
which is the ratio of the compound’s concentration on the solid to its concentration in
water. For organic contaminants, the extent of sorption on soil and sediments often
strongly depends on the organic carbon content of the soil and therefore the organic
carbon normalized sorption coefficient (Koc) is commonly used.
Chemical
Soil/Sediments
Atmosphere
Water Biota
H
S
Kd /Koc
BCF
Chapter 1
6
The bioconcentration factor (BCF) is an indicator of how much of a chemical will
accumulate in living organisms. Fishes are good examples of living organisms in
equilibrium with their surroundings. The BCF can be defined as a partition coefficient
for the concentration of an organic compound in the fish relative to its concentration in
water.
The partitioning of a compound between the different environmental compartments is largely
dependent on the compound polarity (Hites and Raff, 2012; Schwarzenbach, 2003; Benjamin,
2002). The oxygen atom is for example more electronegative than the hydrogen atom. This
means that in a water molecule the oxygen attracts electrons more than the hydrogens do. As a
result, although water molecules are electrically neutral overall, they do have local regions of
finite charge. Because of this separation of charges, the molecule is referred to as being polar.
When a polar molecule A is dissolved into a polar solvent, like water, attractions between
electron poor parts of A and electrons rich sites of water take place. In contrast, non-polar
molecules interact only weakly with water molecules. The electrostatic interactions between
A and water stabilize A molecules in water and make their dissociation favorable. Molecules
whose dissolution is favorable are called hydrophilic (water-loving), and those whose
dissolution is unfavorable are called hydrophobic (water-hating). Polar molecules are
generally hydrophilic, while non-polar compounds are often hydrophobic. In the environment,
polar molecules are more likely to partition to polar media, like water, while non-polar
molecules tend to partition to non-polar media such as lipids in biota, soil, or sediments.
A model system containing water and n-octanol is commonly used as a measure of the
hydrophobicity of molecules. The n-octanol/water partition coefficient (Kow1) describes the
partition of a compound between n-octanol, a non-polar solvent, and water, a polar solvent. A
large Kow value means that the chemical has a higher affinity for non-polar environments than
for polar environments.
Many of the partition coefficients described previously are related to the Kow. For example the
soil/water partition coefficient Koc is related to Kow by a log-log relationship of the type:
log Koc = a·log Kow + b (a and b being constants). Similar relationships have been described
for the bioconcentration factor or the solubility (Schwarzenbach, 2003).
1 Kow = solute n-octanol
[solute]water; [solute]n-octanol being the concentration of the solute in n-octanol and
[solute]water being the concentration of the solute in water.
Chapter 1
7
Based on the physico-chemical properties of a chemical, a rough prediction of the
environmental fate can be made according to the following classification (Vighi and
Calamari, 1993).
Table 1: Classes of affinity of organic chemicals for the different environmental compartments in relation
to the physico-chemical characteristics of the substance.
Affinity Air
H in Pa m3 mol-1 Water
S in g L-1 Soil
Koc in L kg-1 Biota
Kow
Low < 10-3 < 10-3 <1 <10-3
Medium 10-3 - 1 10-3 - 1 1 – 103 103 - 105
High >1 >1 >103 >105
In each environmental compartment, the degradation of the chemical is governed by chemical
and/or biological processes. Chemical processes generally occur in water and in the
atmosphere and involve oxidation, reduction, hydrolysis, and photolysis. Biological
mechanisms in soil and living organisms utilize oxidation, reduction, hydrolysis, and
conjugation to degrade chemicals.
For a more precise analysis and for the estimation of the predicted environmental
concentrations (PECs), computer models are commonly applied. For example the “fugacity
model” uses for prediction a standard “unit of world” of 1 km² divided into six compartments
(air, water, soil, sediments, suspended solids, aquatic biomass). In this model, the fugacity is
used to describe mathematically the rates at which chemicals diffuse or are transported
between phases (McKay, 2001). With physico-chemical parameters as input parameters, the
fugacity model predicts the concentration of a chemical in every environmental compartment.
Different complexity levels can be applied. The level I fugacity model considers a one-time
input of chemical, no reaction, system at equilibrium, fugacity equal in all phases. The level
III fugacity model considers a constant emission of the chemical, system at steady state, and
includes first-order reactions (hydrolysis, photolysis, redox, and biolysis). Generally speaking
and independently from the level of complexity applied, the computer models aim at
predicting the environmental concentrations of a chemical and are therefore a useful tool to
assess the potential exposure to a chemical.
Chapter 1
8
As summarized in Figure 2, the environmental risk assessment of a chemical requires
information both on the toxicological properties of the considered substance and on its
environmental fate.
Figure 2: The steps of an environmental risk assessment
Knowledge on the environmental fate of the compound is required at an early stage of the
environmental risk assessment. The toxicity tests can then be better targeted towards those
environmental compartments for which the exposure is expected to be the highest.
Basic data on the environmental fate
Effect assessment
PNEL/PNEC
Identification of hazard
PEC
Environmental Risk Assessment
Toxicity
Exposure
Chapter 1
9
3. Organosilicon compounds in the environment
Current research on the environmental risk assessment of micropollutants focuses mainly on
pesticides, pharmaceuticals, personal care products, or more recently engineered nano-
particles. However, other chemicals, intensively used and potentially reaching the
environment have been very little under concern. The starting point of this work was the
observation that organosilicon compounds are used in many products of daily life, from
personal care products to façade paint or pesticides but in spite of this widespread use, their
environmental impact has been very little investigated so far. This section 3 gives an overview
of the organosilicon compounds and point out the knowledge gap on their environmental
occurrence and fate.
3.1. Definitions and applications
Organosilicon compounds (silicon based substances in which at least one organic group is
directly bonded to silicon via a Si-C bond (Rösch et al., 2000)) are man-made compounds and
do not exist naturally. Organosilicon compounds are divided into two groups: monomeric
compounds and polymers. Among polymers, silicones are commercially the most important.
The name silicone refers to all chemical substances in which silicon atoms are linked via
oxygen atoms, each silicon bearing one or several organic groups (Moretto et al., 2000). In
industrially important silicones, the organic groups are methyl. The name “silicone” was first
mentioned by Frederick S. Kipping in analogy with the carbon based ketones (Klosowski and
Wolf, 2009). It was kept afterwards for all siloxanes although it was soon understood, by
Kipping himself, that ketones and silicones have very different chemical structures and very
different properties (Robinson and Kipping, 1908).
The silicone family includes a large variety of products. An infinite number of substances can
be created by modifying the nature of the organic substituent attached to the silicon or the
polymerization degree of the siloxane chain. According to the European Silicone Center
(CES), it exists around 2000 forms of silicone (European Silicones Centre, 2004) and the
active research in the field indicates that new silicones are likely to appear in the near future.
Their application fields have been reviewed in details elsewhere (Andriot et al., 2007).
Briefly, one can mention: food, textile, construction, electronics, medical, coatings,
transportation (cars and airplanes), pulp and paper, and adhesives. Silicones are present in
every segment of our everyday life, from contact lenses to façade paint, from dry cleaning to
food or from cosmetics to baking forms.
Chapter 1
10
3.2. Organosilicon compounds in the environment
Although silicones are increasingly used in many consumers and industrial products, the
knowledge on their environmental occurrence and fate is scarce. Because of the diversity of
silicones, the evaluation of their environmental occurrence and fate is a substantial task and
has to be limited to the most relevant compounds. Based on exposure ranking, three
categories of organosilicon compounds are considered to have noteworthy environmental
loading (Figure 3, Allen et al., 1997):
volatile methylsiloxanes (VMS), including cyclic volatile methylsiloxanes (cVMS) and
linear volatile methylsiloxanes
polydimethylsiloxanes (PDMS),
and polyethermethylsiloxanes (PEMS).
PDMS
R = H, CH3, C(O)CH3
PEMS
VMS
Figure 3: Chemical structures of the three categories of organosilicon compounds having noteworthy
environmental loading.
3.2.1. Environmental occurrence and fate of volatile methylsiloxanes
cVMS find their main applications as carriers and emollients in personal care products and as
starting agents for the production of PDMS (Graiver, Farminer, and Narayan 2003; Wang et
al. 2013; Buser, 2015). cVMS meet the criteria defined by the United Nation Environment
Programme for persistent, bioaccumulative and toxic substances (Environment Canada and
Health Canada, 2008a, 2008b, 2008c; United Nations, 2009; Stockholm Convention). cVMS
are potentially released into the environment during their manufacture, use and disposal.
Based on their physico-chemical properties, essentially on their high volatility (Henry´s law
xx 5
x y
n m
OR
xx < 5
x
x < 7
Chapter 1
11
constant estimated between 3.4 and 32 ) and hydrophobicity (water solubility between 2.1
µg L-1 and 2000 µg L-1), and on the results of computer modelling (Environment Canada and
Health Canada, 2008a, 2008b, 2008c; Kaj et al., 2005a; Hughes et al., 2012), cyclic siloxanes
are expected to end up mainly in the atmosphere and in sediments.
cVMS from D3 to D6 have been detected in the atmosphere, at various locations in the world,
in the ng m-3 range (Genualdi et al., 2011; McLachlan et al., 2010; Kierkegaard and
McLachlan, 2010; Kaj et al., 2005b; Brooke et al., 2009; Environment Canada and Health
Canada, 2008a, 2008b, 2008c, Kaj et al., 2005a; Buser et al., 2013; Krogseth et al., 2012;
Yucuis et al., 2013). Higher concentrations were found around highly populated areas (Wang
et al., 2001), and especially around WWTPs and landfills (Cheng et al., 2011). The detection
of D5 in the Arctic confirmed that it is subject to long-range atmospheric transport (Genualdi
et al. 2011; Krogseth et al., 2012; Warner et al., 2010). The dominant decay mechanism of
cVMS in the atmosphere is the reaction with OH radicals (Atkinson, 1991; Sommerlade et al.,
1993; Graiver et al., 2003). This reaction leads to the formation of silanols that are removed
mainly by wet deposition (Whelan et al. 2004). Taking only into account the reaction with OH
radicals at 7.7·105 molecule cm-3, the atmospheric life time of D5 is estimated at ~10 days and
~30 days for D3 (Atkinson, 1991).
cVMS may reach the aquatic environment partly because of an incomplete removal in
WWTPs. The measured removal rates of cVMS in WWTPs were found to be between 80 %
and 100 % (Wang et al., 2013; Buser, 2015; Bletsou et al., 2013; Xu et al., 2013). In a recent
paper (Bletsou et al., 2013), D3 and D4 were found to be not removed, while D5 was removed
at 35 % and D6 and D7 at 97 % (applied treatments: primary sedimentation, activated sludge
process, and secondary sedimentation). The elimination during wastewater treatment is
mainly due to volatilization and sorption to sludge. cVMS have been mentioned as emerging
contaminants in water (Richardson and Ternes, 2011) but only little data is available
regarding their occurrence in the aquatic compartment. In river waters, the measured
concentrations range from below the LOQ to the low ng L-1 range. (Different LOQs from one
study or another: from 20 ng L-1 to 120 ng L-1). (Environment Canada and Health Canada,
2008a, 2008b, 2008c, Kaj et al., 2005a; Brooke et al., 2009; Kaj et al., 2005b; Schlabach et
al., 2007; Sparham et al., 2008; Sparham et al., 2011; Companioni-Damas et al., 2012). In
water, D5 is not biodegradable (Whelan et al., 2010) and its hydrolysis at pH 7 is limited
(half-life 449 days at pH 7 and at 9°C) (Brooke et al., 2009). The elimination mainly comes
Chapter 1
12
from volatilization and adsorption to dissolved and particulate organic carbon (Sparham et al.,
2011; Whelan et al., 2009; Whelan et al., 2010). The concentrations of cVMS in sediments
are generally lower than 100 ng g-1 dry weight and are higher downstream from important
urban areas. (Kaj et al., 2005b; Brooke et al., 2009, Environment Canada and Health Canada,
2008a, 2008b, 2008c, Kaj et al., 2005a; Schlabach et al., 2007; Sparham et al., 2011; Zhang et
al., 2011). Currently, no data exists regarding the occurrence of cVMS in drinking water.
Cyclic siloxanes may reach the soil compartment through the amendment of biosolids on
fields. D5 and D6 were detected in several Spanish soils (agricultural, industrial and sludge
amended soils), with concentrations between 5 ng g-1 and 500 ng g-1 dry weight (Sanchez-
Brunete in 2010). Measurements of two soil samples from the Faroe Islands did not reveal the
presence of cVMS (LOQs between 2 ng L-1 and 10 ng L-1) (Kaj et al., 2005b). The elimination
of cVMS from soil may come from a combination of volatilization and degradation (Xu and
Chandra, 1999; Xu, 1999). The relative importance of each phenomenon varies with the
conditions, especially with the soil moisture. The degradation mechanism on soil starts with a
surface catalyzed ring-opening hydrolysis to form linear oligomeric diols until
dimethylsilandiol.
The environmental behavior of linear polydimethylsiloxanes depends on its polymerization
degree i.e. its molecular mass. The environmental fate of low molecular weight PDMS
(0 < n < 4) is very similar to the one of cVMS and that is why cVMS and linear VMS are
often studied together. Linear VMS (0 < n < 3) have been measured in air, soil, sediment, soil,
surface water, and sewage water and were only rarely detected (Kaj et al., 2005a; 2005b).
Generally speaking their concentrations were lower than the one of cVMS. As expected based
on the water solubility of PDMS, the higher is the molecular weight of PDMS, the higher is
the concentrations in sludge and sediments and the lower the concentrations in atmosphere
and water (Kaj et al., 2005b; Bletsou et al., 2013; Cheng et al., 2011; Genualdi et al. 2011;
Kaj et al., 2005a; Schlabach et al., 2007).
3.2.2. Environmental occurrence and fate of polydimethylsiloxanes
When high molecular weight PDMS reaches WWTPs, it partitions rather to sludge because of
its hydrophobicity. The removal efficiency of PDMS is between 70 % and 95 % (Fendinger et
al., 1997; Bletsou et al., 2013). In a monitoring program of eight WWTPs in North America
(Fendinger et al., 1997), the concentration of PDMS in effluents was below 5 µg L-1 and the
sludge concentrations were between 290 mg kg-1 and 5155 mg kg-1. The fate of PDMS then
Chapter 1
13
depends on the fate of the sludge. It could eventually enter the environment if the sludge is
amended on fields. This expected fate explains why most of the studies on PDMS in the
environment focus on its soil degradation (Carpenter et al., 1995; Griessbach and Lehmann,
1999; Lehmann et al., 1994a; Lehmann et al., 1994b; Lehmann et al., 1995; Lehmann and
Miller, 1996; Lehmann et al., 1998a; Lehmann et al., 1998b; Lehmann et al., 2000; Powell et
al., 1999; Sabourin et al., 1996; Stevens, 1998; Xu et al., 1998). On soils, PDMS undergoes
hydrolysis catalyzed by clay minerals and dimethylsilanediol is formed.
Figure 4: Schematic representation of the degradation of PDMS by depolymerization, until the
dimethylsilanediol. Modified from Griessbach and Lehmann, 1999.
In some WWTPs, the sludge is used to produce biogas by anaerobic digestion. The obtained
biogas contains mainly methane and carbon dioxide and is combusted to generate heat and
electricity. In a context of fossil fuel depletion, energy import dependence, and climate
change, energy recovery from WWTPs is an emerging topic and biogas, as source of
renewable energy, is one possible answer (Stutton et al. 2011). Present as impurities,
siloxanes hamper the production of biogas (Dewil et al., 2006; Appels et al., 2011, Rasi et al.,
2011). During anaerobic digestion, the siloxanes contained in the WWTP sludge volatilize
into the biogas and precipitate as micro crystalline silica during the combustion process. This
phenomenon can cause serious damage to the moving parts of the engine and hampers the use
of biogas. Finding a solution to this issue would increase the efficiency and the profitability of
biogas production. Several technical solutions are currently investigated but so far no fully
satisfying answer has been found and more research would be needed.
3.2.3. Environmental occurrence and fate of polyethermethylsiloxanes
On the contrary to VMS and PDMS, PEMS have not been much investigated yet and data on
their environmental occurrence and fate is scarce. The following section summarizes the
existing knowledge on PEMS.
OHSi
OHnSi
OH2+
Chapter 1
14
4. Silicone surfactants
4.1. Structure, properties, and applications
All surfactants have in common an asymmetrical chemical structure with a combination of a
hydrophobic and a hydrophilic moiety (Knepper and Berna, 2003). This particular chemical
structure is the origin of their surface active properties. Surfactants decrease the surface
tension of aqueous solutions and tend to partition at the interfaces, for example at the interface
between water and oil. In the case of silicone surfactants, the hydrophobic moiety consists of
a permethylated siloxane chain. Several types of hydrophilic moieties can be attached to the
siloxane chain: non-ionic groups like polyoxyethylene, polyoxyethylene/polyoxypropylene, or
saccharides, anionic moieties like sulfate for instance, cationic groups like quaternary
ammonium groups, or zwitterionic moieties with betaines. Polyethermethylsiloxanes (PEMS)
are the most common silicone surfactants. They consist of a siloxane backbone (the
hydrophobic part), a spacer (CH2CH2CH2), and a combination of ethylene oxide groups
(OCH2CH2), propylene oxide groups (OCH2CHCH3), constituting the hydrophilic part. The
general structure of PEMS is shown in Figure 5.
Figure 5: General chemical structure of the most common silicone surfactants.
The properties of PEMS can be adapted to a specific application by changing the
polymerization degree of the siloxane chain (x and y), the length and the composition of the
polyether chain (n and m), and the nature of the end capping group (R). Increasing the number
x y
n m
OR
Hydrophobic part
Hydrophilic part
Chapter 1
15
of ethylene oxide groups increases for instance the water solubility. The possible tailorization,
together with the higher efficiency of silicone surfactants to reduce surface tension of aqueous
solutions compared to traditional carbon based surfactants, made the PEMS successful in
various applications: from polyurethane foam production to personal care products or
pesticides.
For the production of polyurethane foam: The silicone surfactants used during the
production of polyurethane foam perform various tasks. They stabilize the foam bubbles in
the flexible foam or they emulsify the ingredients in the rigid foam. The possibility to adapt
their physico-chemical properties to a specific need is a key point for their use in polyurethane
foam production. The application of silicone surfactants for the production of polyurethane
foam have been reviewed by Snow and Stevens (1999).
In personal care products: The high adaptability of the silicone surfactant structure is one
reason for their success in personal care products. Many organic substituents can be attached
to the siloxane chain, depending on the desired properties: anionic groups like phosphates,
sulfates, carboxylates, sulfosuccinates, sulfonates or thiosulfates; amphoteric groups like
betaines (good antistatic agent); cationic groups like quaternary ammonium groups: non-ionic
groups like alcohol ethoxylates, esters or alkanolamides. The obtained properties for each
type of substituent have been detailed by Floyd (1999). According to the International
Nomenclature of Cosmetics Ingredients (INCI) from the Cosmetic, Toiletry and Fragrance
Association (CFTA), when alcohol ethoxylate is the organic moiety, the silicone surfactant is
named with the following acronyms:
polyethermethylsiloxanes: Dimethicone copolyol
PEG x dimethicone when it contains only EO (m = 0)
PEG/PPG-x/y dimethicone when it contains EO and PO (n and m ≠ 0).
Dimethicone copolyol is used for its ability to decrease the surface tension, as wetting agent,
conditioners, glosser, emulsifier and foam builder. The example of application of PEMS to a
bath foam formulation is quite interesting and illustrates how the formulator uses the physico-
chemical properties to meet the needs or desires of the consumer. PEMS possess a so-called
cloud-point i.e. a temperature above which the substance is not soluble in water anymore and
causes the aqueous solution to become cloudy or milky. If the right silicone surfactant is used,
it is possible to create bath oil which is clear and transparent at ambient temperature and
Chapter 1
16
which causes a bathtub of hot water to become milky (Vick, 1984). The exact cloud-point can
be adapted by choosing the correct EO/PO ratio.
As agricultural adjuvants: Application in agriculture was mentioned for the first time by
Jansen (Jansen, 1973). Trisiloxane surfactants are nowadays increasingly used as agricultural
adjuvants in combination with herbicides, fungicides, insecticides, nutrients, or growth
regulators for their ability to improve the activity and application characteristics of the active
substance. They reduce the value of surface tension of water from 72 mN m-1 to 22 mN m-1
(at 24°C and 0.1 %) and promote a rapid and important spreading on hydrophobic surfaces
like leaves (Knoche et al., 1991; Penner et al., 1999). This effect gave them the name of
superspreaders or superwetters. The mechanism of superspreading is not fully understood yet
but seems to be related to the ability of the surfactant to form bilayer aggregates. An
illustration of spreading on a leaf, with and without silicone surfactant is shown in Figure 6. A
schematic representation of the situation at the spreading edge of an aqueous droplet
containing trisiloxane surfactant is also represented, as suggested in the review of Venzmer
(2011).
Figure 6: (a) Wetting of a leaf by water with (lower part) and without (upper part) trisiloxane surfactant.
(b) Proposed situation at the spreading edge of an aqueous droplet of trisiloxane surfactant on a
hydrophobic surface.
Among all silicone surfactants, trisiloxane surfactants containing an average of 7.5 EO groups
(x = 1, y = 0, n = 7.5, m = 0) have been found to be the optimum structure for agricultural
Chapter 1
17
applications. The methyl capped trisiloxane surfactant (R = CH3, Figure 5) is the subject of
the great majority of the research on silicone surfactants as agricultural adjuvants. It is often
referred to with one of its commercial names: Silwet® L77 (x = 1, y = 0, n = 7.5, m = 0,
R = CH3). Trisiloxane surfactants are used as tank mix at concentrations around 0.01 %.
4.2. The targeted trisiloxane surfactant
Among all PEMS, the present work focuses on trisiloxane surfactants (x = 1, y = 0, n = 7.5,
m = 0)2. Existing data on the environmental occurrence, fate and toxicity of trisiloxane
surfactants is scarce. Based on its applications, especially in agriculture, and its physico-
chemical properties (relatively high polarity compared to others PEMS), trisiloxane
surfactants are expected to be released to the environment and may reach natural waters.
4.2.1. Entry routes into the environment and expected environmental fate
Entry routes into the environment: Knowing the uses of trisiloxane surfactants, several
potential entry routes to the environment can be hypothesized. On the one hand, trisiloxane
surfactants may reach the soil compartment and eventually reach natural waters when they are
used as agricultural adjuvants. On the other hand, their use in personal care products imply
that trisiloxane surfactants may reach WWTPs and eventually surface waters if the applied
treatments in WWTPs fail to eliminate 100 % of the incoming trisiloxane surfactant. Other
entry routes can also be hypothesized: environmental release related to the production,
accidental contamination, or improper disposal.
Expected environmental fate: The hydrophobicity of a chemical is very important to predict
its environmental fate (Toshio, 2001). Log Kow values have been estimated for a few silicone
surfactants and for PDMS by using the prediction method from Molinspiration available
online 3.
2 The analytical method described in chapter 2 includes two trisiloxane surfactants: a methyl capped trisiloxane surfactant (x = 1, y = 0, n = 7.5, m = 0, R = CH3) and a hydrogen capped trisiloxane surfactant (x = 1, y = 0, n = 7.5, m = 0, R = H). However, NMR confirmation of the EO oligomeric distribution have been performed only for the methyl capped trisiloxane surfactant. The composition of the analytical standard is therefore only known with certainty for one trisiloxane surfactant. Moreover, the initial screening of surface water samples have not revealed any detectable hydrogen capped trisiloxane surfactant. As a consequence, it was decided to focus further investigations on the environmental occurrence and fate of the methyl capped trisiloxane surfactant, even if the analytical method originally included two trisiloxane surfactants. 3 http://www.molinspiration.com/services/logp.html
Chapter 1
18
Table 2: Comparison of the log Kow values for a few PEMS and PDMS (chemical structures, see Figure 5).
Methyl capped polyether silicone surfactants PDMS
x 1 1
0
y 0 9
1 2 3 4
m 0 0
0
n 4 5 6 7 8 9 10 11 12 7
0
log Kow 3.9 3.7 3.5 3.3 3.1 2.9 2.7 2.5 2.3 6.0
5.2 5.6 5.9 6.3
The predicted log Kow values for a few PEMS and PDMS (Table 2) show that the
hydrophilicity increases when the number of ethylene oxide groups increases and when the
siloxane chain length decreases. With a short siloxane chain (three silicon atoms) and a
relatively long hydrophilic moiety (on average 7.5 ethylene oxide groups), the targeted
trisiloxane surfactant is one of the most polar PEMS. The trisiloxane surfactant is also clearly
more polar than PDMS having no hydrophilic moiety (x = 0). Finally on the contrary to VMS,
the targeted trisiloxane surfactant is not volatile. Consequently, among organosilicon
compounds with noteworthy environmental loading (PDMS, VMS, and PEMS), the targeted
trisiloxane surfactant is expected to be the more likely to be mobile in the environment and to
reach the aquatic environment.
In the aquatic environment, hydrolysis is thought to be an important elimination mechanism
of trisiloxane surfactants. This hypothesis is based on several studies showing that trisiloxane
surfactants are sensitive to hydrolysis. However, most of those studies have been focused on
the stability in pesticide formulations, under acidic conditions, and at high concentrations,
from 0.1 g L-1 to 1 g L-1 (Knoche et al., 1991; Radulovic et al., 2009; Radulovic et al., 2010).
Moreover, the hydrolysis has been investigated by measuring the evolution of surface tension
over time, but since nothing is known on the surface activity of the by-products and
degradation products, those measurements are not specific and could have led to a wrong
interpretation.
Sorption on soil or sediments is a major factor influencing the transport and degradation of
chemicals in the environment. A study already exists on the soil/water distribution of one
trisiloxane surfactant (R = C(O)CH3) but the lack of a specific analytical method led to
ambiguous results (Griessbach et al. 1997; Griessbach et al. 1998). The test substance was a
Chapter 1
19
trisiloxane surfactant, 14C-radiolabelled at the terminal methyl group of the hydrophilic
moiety, and the detection was performed by liquid scintillation counting. The
sorption/desorption experiments concluded that the trisiloxane surfactant is immobile on soil
(KF between 8 µg1-1/n(cm³)1/ng-1 and 75 µg1-1/n(cm³)1/ng-1) (Griessbach et al. 1997). However
soil column leaching experiments revealed the presence of 14C-radiolabelled substance in the
leachate. The lack of a specific analytical method prevented discrimination between the
trisiloxane surfactant and its degradation products (Griessbach et al. 1998) and no definitive
conclusion could be draw from the results.
Based on their usage and their physico-chemical properties trisiloxane surfactants are
expected to be released to the environment and to partition between water, soil, and
sediments. However existing data are scarce and do not allow to get a full picture of the
environmental fate of trisiloxane surfactants. To increase our knowledge on the environmental
fate of this category of compounds and based on the expected environmental fate described
above, the present work was focused on two environmental compartments: surface waters and
soil.
4.2.2. Occurrence in the environment
Very little data exist on the occurrence of trisiloxane surfactants in the environment. The latter
have never been analyzed in the aquatic environment, neither in soil, air or biota.
Recently, three trisiloxane surfactants (R = H, CH3, and C(O)CH3) have been analyzed in
honey, pollen, and beeswax (Chen and Mullin, 2013). None of the trisiloxane surfactants was
detected in honey (method detection limits: 0.53 ng g-1, 0.60 ng g-1, 0.56 ng g-1 for the methyl,
acetyl, and hydroxyl end capped trisiloxane surfactants, respectively). The average total
trisiloxane surfactant concentration was 116 ng g-1 in beeswax and 18 ng g-1 in pollen. The
methyl capped trisiloxane surfactant has been detected in 6 pollen samples out of 10, with an
average concentration of 15 ng g-1 (8 ng g-1 – 22 ng g-1) and in 5 beeswax samples out of 10,
with an average concentration of 75 ng g-1 (11 ng g-1 – 153 ng g-1).
One reason for the lack of information on the environmental occurrence of trisiloxane
surfactants comes from the fact that until recently no analytical method for their trace analysis
in environmental matrices was available.
Chapter 1
20
4.2.3. Toxicity and ecotoxicity
Pests: Most of the existing studies on the toxicity of trisiloxane surfactants focus on the
toxicity to pests, especially of the methyl capped trisiloxane surfactant. The latter is toxic to
arthropod pests including two-spotted spider mites, aphids, citrus leafminers, tephritid fruit
flies and armyworms (Shapiro et al., 1998; Cocco and Hoy, 2008; Tipping et al., 2003;
Cowles et al., 2000; Imai et al., 1995; Chandler, 1995; Purcell and Schroeder, 1996).
According to Cowles (2000), its mode of action is related to its surfactant properties: it acts as
an extremely active soap and thereby permits water to infiltrate spider mites’s respiratory
system. In addition, when the trisiloxane surfactant enters the body, it can interact with nerve
and cell membranes and disrupt their function.
The aphicidal effects of the methyl capped trisiloxane surfactant depend on several factors:
i) The type of plant on which it is applied. According to Cowles (2000), higher
concentrations of the trisiloxane surfactant were required to kill mites on strawberry
leaves than on bean leaves. This may come from the differences in the ability of the
surfactant to wet the leaf surface, and therefore to surround and get in contact with the
mites.
ii) The rapidity of the evaporation: When evaporation is inhibited, (by a high relative
humidity or the presence of a compound which inhibit evaporation like calcium
chloride, glycerin or sodium carboxymethyl cellulose), a continuous watery film can
spread over the aphid and stay long enough to induce suffocation (Imai, 1995).
iii) The size of the insect: As suggested by Imai (1995), the insecticidal effect of the
trisiloxane surfactant is probably limited to small insects due to the difficulty to
sustain a watery film on the entire body surface of bigger insects.
Non-target species: Unpublished data from the industry has been compiled by Powell and
Carpenter (1997). Information on the toxicity of one trisiloxane surfactant (x =1, y = 0,
R = CH3) to several aquatic organisms is also available from several Material Safety Data
Sheets (MSDSs). Available data on the toxicity of silicone surfactants to aquatic organisms is
summarized in Table 3.
Chapter 1
21
Table 3: Toxicity of PEMS compounds to various aquatic organisms
Compound Species Test
Effect
Concentration
(mg L-1)
Reference
PEMS 1
Pow
ell
and
Car
pent
er, 1
997
MW = 31 400 g mol-1 Sewage microorganisms IC50 (glucose metabolism) 297 m
Wt % Si = 6.7 Bacteria (Photobacterium phosphorem) EC50 (Microtox®) >455 m
EO and PO Green algae (Selenastrum capricornutum) 96-h EC50 (growth inhibition) 623 m
PE substitution = 8.8 M % Duck weed (Lenna gibba) 7-dEC50 (frond number) >977 m
R = C(O)CH3 Water flea (Daphnia magna) 48-h LC50 (static) >960 m
Rainbow trout (Onchorhynchus mykiss) 96-h LC50 (static) >884 m
Zebra fish (Brachidanio rerio) 96-h LC50 (static) >500 m
Zebra fish (Brachidanio rerio) 105-d LC50(15-d renewal) >10 m
PEMS 2
Pow
ell
and
Car
pent
er, 1
997
MW = 3100 g mol-1 Bacteria (Photobacterium phosphorem) EC50 (Microtox®) >455 m
Wt % Si = 11.4 Blue-green algae (Anabaena flos-aquae) 96-h EC50 (growth inhibition) 755 n
EO only Green algae (Selenastrum capricornutum) 96-h EC50 (growth inhibition) 746 m
PE substitution = 25 M % Duck weed (Lenna gibba) 7-dEC50 (frond number) >1020 m
R = H Brine shrimp (Artemia salina) 24-h LC50 (static) >500 n
Water flea (Daphnia magna) 48-h LC50 (static) >930 m
Water flea (Daphnia magna) 48-h LC50 (static) 816 n
Water flea (Daphnia magna) 48-h LC50 (static) 311 n
Water flea (Daphnia magna) 48-h LC50 (flow-through) 486 n
Water flea (Daphnia magna) 21-d LC50(7-d renewal) >10 n
Bluegill sunfish (Lepomis macrochirus) 96-h LC50 (static) >1000 n
Rainbow trout (Onchorhynchus mykiss) 96-h LC50 (static) 860 m
Rainbow trout (Onchorhynchus mykiss) 96-h LC50 (static) 250 n
Chapter 1
22
Compound Species Test
Effect
Concentration
(mg L-1)
Reference
PEMS 3
Pow
ell
and
Car
pent
er, 1
997 MW = 3600 g mol-1 Bacteria (Photobacterium phosphorem) EC50 (Microtox®) >455 m
Wt % Si = 20.7 Green algae (Selenastrum capricornutum) 96-h EC50 (growth inhibition) 741 m
EO only Duck weed (Lenna gibba) 7-dEC50 (frond number) >1010 m
PE substitution = 8.1 M % Water flea (Daphnia magna) 48-h LC50 (static) >960 m
R = C(O)CH3 Rainbow trout (Onchorhynchus mykiss) 96-h LC50 (static) 115 m
PEMS 4
Pow
ell
and
Car
pent
er, 1
997
MW = 672 g mol-1 Water flea (Daphnia magna) 48-h LC50 (static) 41 n
Wt % Si = 9.8 Fathead minnow (Pimphales promelas) 96-h LC50 (static) 4 n
EO only
PE substitution = 34 M %
R = C(O)CH3
PEMS 5
EO only
x = 1
y = 0
R = CH3
Rainbow trout (Onchorhynchus mykiss) 96-h LC50 4.5
MS
DS
(M
omen
tive
, GE
,
Nuf
arm
, DeS
ango
sse)
96-h NOEC 3.2
Water flea (Daphnia magna) 48-h LC50 24
48-h EC50 37
48-h NOEC 25
Water flea (Daphnia similis) 48-h EC50 22.61
48-h NOEC 10
Zebra fish (Brachidanio rerio) 96-h NOEC 0.56
Bluegill (Lepomis macrochirus) 96-h LC50 6
Chapter 1
23
Compound Species Test Effect
Concentration
(mg L-1)
Reference
PEMS 5
MS
DS
(M
omen
tive
, G
E,
Nuf
arm
, D
eSan
goss
e)
Spirilum volutans 120 min-MEC (uncoordinated mobility in 90 % of the population)
> 0.201
Green algae (Selenastrum capricornutum) 96-h EC50 5.5 96-h NOEC 1
m = measured concentration, n = initial concentration, MW = molecular weight, wt % Si = weight % of Si in the PEMS product, PE substitution = degree of polyether substitution on the copolymer in mole %, R = polyether endcap.
Figure 7: Daphnia magna (a, image credit: Hajime Watanabe) and pimphales promelas (b, image credit: Joseph Tomelleri)
a b
Chapter 1
24
Most of the data presented in Table 3 relate to high molecular weight silicone surfactants
(PEMS 1 to PEMS 3 have MW > 3000). They indicate that the trisiloxane surfactant (PEMS
5) is more toxic to aquatic organisms than high MW silicone surfactants. However, the mode
of action of silicone surfactants on aquatic invertebrates is unknown (Stark and Walthall,
2003).
5. Objectives
The overall objective of this work is to fill the lack of knowledge on the environmental
occurrence and fate of silicone surfactants and to evaluate their possibility to contaminate
natural waters. The present work is more specifically focused on one trisiloxane surfactant
(y = 0, x = 1, m = 0, average n = 7.5, R = CH3).
The first task was to develop an analytical method allowing the determination of the
trisiloxane surfactant at low concentrations (ng L-1 range) in environmental samples. Two
main challenges had to be overcome: the lack of analytical standards and the oligomeric
nature of the trisiloxane surfactant. The only standard available was a mixture of homologues,
each having a different and unknown concentration. The first step of the method development
was the characterization of the oligomeric distribution, i.e. the determination of the exact
composition of the standard, based on mathematical description of the Poisson distribution
and on experimental measurements by MS and NMR.
The next aim was to analyze surface waters to get information on the occurrence of trisiloxane
surfactants in surface waters. Internal standards are commonly used in analytical chemistry to
compensate for losses of analyte during sample preparation or any variation of intensity
during detection by mass spectrometry. It is common to use the isotopically labeled analyte as
an internal standard. Such compounds are currently not commercially available for trisiloxane
surfactants and this was the main hardship faced during the application of the analytical
method to environmental samples. The lack of internal standard made the method sensitive to
variations of sensitivity of the mass spectrometer. To overcome this problem, a control
sample of known concentration was analyzed at regular intervals (every ten samples) during
the analysis sequence. With this control sample, the intensity of the signal could be controlled
and samples with inadequate intensity could be identified and discarded. This resulted in a
decreased number of environmental samples and unfortunately limited the amount of results
that could be produced on trisiloxane surfactants in the aquatic environment.
Chapter 1
25
For a better understanding of the fate in the aquatic compartment, the degradation by
hydrolysis was studied under environmental relevant conditions of pH and temperature
according to the OECD guideline 111. The aim of this study was to confirm that hydrolysis is
a relevant elimination mechanism of trisiloxane surfactants from the environment. The rates
of degradation by hydrolysis were compared in various conditions of pH, temperature,
concentrations, and in different matrices (pure water, river water, and filtrated river water).
The third part of the work was dedicated to the study of the mobility of the trisiloxane
surfactant on soil. In order to predict the partition between water and soil and to evaluate the
possibility of leaching or runoff, the sorption behavior was studied based on the OECD
guideline 106. Furthermore, to clarify the possibility of leaching through soil after application
as agricultural adjuvant, leaching in soil column experiments were carried out.
Chapter 1
26
References
Allen R.B., Kochs P., and Chandra G.,1997. Industrial organosilicon materials, their environmental entry and predicted fate, pp. 1, in: Organosilicon materials, Chandra G. (Ed.), Springer-Verlag, Berlin, Heidelberg.
Andriot M., Chao S.H., Colas A., Cray S., de Buyl F., DeGroot J.V., Dupont A., Easton T., Garaud J.L., Gerlach E., Gubbels F., Jungk M., Leadley S., Lecomte J.P., Lenoble B., Meeks R., Mountney A., Shearer G., Stassen S., Stevens C., Thomas X., and Wolf A.T., 2007. Silicones in industrial applications, pp. 61, in: Inorganic polymers, De Jaeger, R. and Gleria, M. (Eds.), Nova Publishers, Hauppauge.
Appels, L., Lauwers, J., Degrève, J., Helsen, L., Lievens, B., Willems, K., Van Impe, J., and Dewil, R., 2011. Anaerobic digestion in global bio-energy production: Potential and reaserch challenges. Renew. Sust. Energ. Rev., 15, 4295.
Atkinson, R., 1991. Kinetics of the gas-phase reactions of a series of organosilicon compounds with hydroxyl and nitrate (NO3) radicals and ozone at 297 +/- 2 K. Environ. Sci. Technol., 25, 863.
Benjamin, M.M., 2002. Concepts in aquatic chemistry, pp. 1, in: Water chemistry. Waveland Press, Long Grove, Illinois.
Bletsou, A.A., Asimakopoulos, A.G., Stasinakis, A.S., Thomaidis, N.S., and Kannan, K., 2013. Mass loading and fate of linear and cyclic siloxanes in a wastewater treatment plant in Greece. Environ. Sci. Technol., 47, 1824.
Brooke D.N., Crookes M.J., Gray D., and Robertson S., 2009. Environmental risk assessment report: Octamethylcyclotetrasiloxane, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/290565/scho0309bpqz-e-e.pdf (Access June 2014)
Buser, A.M., Kierkegaard, A., Bogdal, C., MacLeod, M., Scheringer, M., and Hungerbühler, K., 2013. Concentrations in ambient air and emissions of cyclic volatile methylsiloxanes in Zurich, Switzerland. Environ. Sci. Technol., 47, 7045.
Buser, A.M., 2015. Corrigendum to "Review of recent advances in research on the toxicity, detection, occurrence and fate of cyclic volatile methyl siloxanes in the environment” [Chemosphere 93 (5) (2013) 711-725]. Chemosphere, 119, 1275.
Carpenter, J.C., Cella, J.A., and Dorn, S.B., 1995. Study of the degradation of polydimethylsiloxanes on soil. Environ. Sci. Technol., 29, 864.
Chandler, L., 1995. Effect of surfactants on beet armyworm and fall armyworm larvae. Arthropod Manage. Tests, 20, 353.
Chen, J. and Mullin, C.A., 2013. Quantitative determination of trisiloxane surfactants in beehive environments based on liquid chromatography coupled to mass spectrometry. Environ. Sci. Technol., 47, 9317.
Chapter 1
27
Cheng, Y., Shoeib, M., Ahrens, L., Harner, T., and Ma, J., 2011. Wastewater treatment plants and landfills emit volatile methyl siloxanes (VMSs) to the atmosphere: investigations using a new passive air sampler. Environ. Pollut., 159, 2380.
Cocco, A. and Hoy, M.A., 2008. Toxicity of organosilicone adjuvants and selected pesticides to the asian citrus psyllid (Hemiptera: Psyllidae) and its parasitoid Tamarixia Radiata (Hymenoptera: Eulophidae). Fla. Entomol., 91, 610.
Companioni-Damas, E.Y., Santos, F.J., and Galceran, M.T., 2012. Analysis of linear and cyclic methylsiloxanes in water by headspace-solid phase microextraction and gas chromatography-mass spectrometry. Talanta, 89, 63.
Cowles, R.S., Cowles, E.A., McDermott, A.M., and Ramoutar, D., 2000. "Inert" formulation ingredients with activity: toxicity of trisiloxane surfactant solutions to twospotted spider mites (Acari : Tetranychidae). J. Econ. Entomol., 93, 180.
Daughton, C.G. and Ternes, T.A., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Perspect., 107, 907.
Dewil, R., Appels, L., and Baeyens, J., 2006. Energy use of biogas hampered by the presence of siloxanes. Energ. Convers. Manage., 47, 1711.
Environment Canada and Health Canada, 2008a. Screening assessment for the challenge octamethylcyclotetrasiloxane (D4). http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=2481B508-1 (Access June 2014)
Environment Canada and Health Canada, 2008b. Screening assessment for the challenge decamethylcyclopentasiloxane (D5). http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=13CC261E-1 (Access June 2014)
Environment Canada and Health Canada, 2008c. Screening assessment for the challenge dodecamethylcyclohexasiloxane (D6), 2008. http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=FC0D11E7-1 (Access June 2014)
European Silicones Centre, 2004. Silicones: overview II. http://www.silicones.eu/uploads/Modules/Library/silicones_overview2.pdf (Access November 2013)
Fendinger, N.J., Mcavoy, D.C., Eckhoff, W.S., and Price, B.B., 1997. Environmental occurrence of polydimethylsiloxane. Environ. Sci. Technol., 31, 1555.
Floyd, D.T., 1999. Silicone surfactants: applications in the personal care industry, pp. 181, in: Silicone surfactants, Surfactant science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker, New York.
Genualdi, S., Harner, T., Cheng, Y., MacLeod, M., Hansen, K.M., van Egmond, R., Shoeib, M., and Lee, S.C., 2011. Global distribution of linear and cyclic volatile methyl siloxanes in air. Environ. Sci. Technol., 45, 3349.
Graiver, D., Farminer, K.W., and Narayan, R., 2003. A review of the fate and effects of silicones in the environment. J. Polym. Environ. 11, 129.
Chapter 1
28
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze, P., 1997. Mobility of a siliconepolyether studied by adsorption and desorption isotherms and soil thin-layer chromatography. Sci. Total Environ., 201, 89.
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze P., 1998. Mobility of a siliconepolyether studied by leaching experiments through disturbed and undisturbed soil columns. Sci. Total Environ., 221, 159.
Griessbach, E.F.C. and Lehmann, R.G., 1999. Degradation of polydimethylsiloxane fluids in the environment - A review. Chemosphere, 38, 1461.
Hites, R.A., Raff, J.D., 2012. Fates of organic compounds, pp. 223, in: Elements of environmental chemistry, John Wiley & Sons, Hoboken, New Jersey.
Hughes, L., Mackay, D., Powell, D.E., and Kim, J., 2012. An updated state of the science EQC model for evaluating chemical fate in the environment: application to D5 (decamethylcyclopentasiloxane). Chemosphere , 87, 118.
Imai, T., Tsuchiya, S., and Fujimori, T., 1995. Aphicidal effects of Silwet L-77, organosilicone nonionic surfactant. Appl. Entomol. Zool., 30, 380.
Jansen, L.L., 1973. Enhancement of herbicides by silicone surfactants, Weed Sci., 21, 130.
Jekel, M., Ruhl, A.K., Meinel, F., Zietzschmann, F., Pflugmacher Lima, S., Baur, N., Wenzel, M., Gnirß, R., Sperlich, A., Dünnbier, U., Böckelmann, U., Hummelt, D., van Baar, P., Wode, F., Petersohn, D., Grummt, T., Eckhardt, A., Schulz, W., Heermann, A., Reemtsma, T., Seiwert, B., Schlittenbauer, L., Lesjean, B., Miehe, U., Remy, C., Stapf, M., and Mutz, D., 2013. Anthropogenic organic micro-pollutants and pathogens in the urban water cycle: assessment, barriers and risk communication (ASKURIS). Environ. Sci. Eur., 25, 20.
Jørgensen, S.E., 2010. Introduction, pp. 3, in: Ecotoxicology, Jørgensen, S.E. (Ed.), Elsevier, Amsterdam.
Kaj L., Andersson J., Palm Cousins A., Remberger M., and Brorström-Lundén E., 2005a. Results from the Swedish national screening programme 2004, http://www.imm.ki.se/Datavard/PDF/B1643_siloxaner.pdf (Access June 2014)
Kaj, L., Schlabach, M., Andersson, J., Cousins, A.P., Schmidbauer, N., and Brorström-Lundén, E., 2005b. Siloxanes in the nordic environment, http://www.norden.org/da/publikationer/publikationer/2005-593/at_download/publicationfile (Access June 2014)
Kierkegaard, A. and McLachlan, M.S., 2010. Determination of decamethylcyclopentasiloxane in air using commercial solid phase extraction cartridges. J. Chromatogr. A, 1217, 3557.
Klosowski, J.M., and Wolf, A.T., 2009. The history of sealants. in: Handbook of sealant technology, Mittal, K.L., Pizzi, A. (Eds.), CRC Press, Boca Raton, United States.
Knepper, P.T., and Berna, J.L., 2003. Surfactants: properties, production, and environmental aspects, pp. 1, in: Analysis and fate of surfactants in the aquatic environment, Knepper, T.P., Barcelò, D., DeVoogt, P. (Eds.), Elsevier, Amsterdam.
Chapter 1
29
Knoche, M., Tamura, H., and Bukovac, M.J., 1991. Performance and stability of the organosilicone surfactant L-77 - Effect of pH, concentration, and temperature. J. Agric. Food Chem., 39, 202.
Krogseth, I.S., Kierkegaard, A., McLachlan, M.S., Breivik, K., Hansen, K.M., and Schlabach, M., 2012. Occurrence and seasonality of cyclic volatile methyl siloxanes in Arctic air. Environ. Sci. Technol., 47, 502.
Kümmerer K., 2011. Emerging contaminants versus micropollutants. CLEAN, 39, 889.
Kümmerer K., 2013. The issue of micropollutants in urban water management, pp. 71, in: Source separation and decentralization for wastewater management, Larsen A.T., Udert K.M., and Lienert J. (Eds.), IWA Publishing, London.
Lehmann, R.G., Varaprath, S., and Frye, C.L., 1994a. Degradation of silicone polymers in soil. Environ. Toxicol. Chem., 13, 1061.
Lehmann, R.G., Varaprath, S., and Frye, C.L., 1994b. Fate of silicone degradation products (Silanols) in soil. Environ. Toxicol. Chem., 13, 1753.
Lehmann, R.G., Varaprath, S., Annelin, R.B., and Arndt, J.L., 1995. Degradation of silicone polymer in a variety of soils. Environ. Toxicol. Chem., 14, 1299.
Lehmann, R.G. and Miller, J.R., 1996. Volatilization and sorption of dimethylsilanediol in soil. Environ. Toxicol. Chem., 15, 1455.
Lehmann, R.G., Miller, J.R., Xu, S., Singh, U.B., and Reece, C.F., 1998a. Degradation of silicone polymer at different soil moistures. Environ. Sci. Technol., 32, 1260.
Lehmann, R.G., Miller, J.R., and Collins, H.P., 1998b. Microbial degradation of dimethylsilanediol in soil. Water Air Soil Poll., 106, 111.
Lehmann, R.G., Miller, J.R., and Kozerski, G.E., 2000. Degradation of silicone polymer in a field soil under natural conditions. Chemosphere, 41, 743.
Linde, C.D., 1994. Physico-chemical properties and environmental fate of pesticides. http://www.cdpr.ca.gov/docs/emon/pubs/ehapreps/eh9403.pdf (Access June 2014)
McKay, D., 2001. Multimedia environmental models. The fugacity approach, CRC Press, Boca Raton, United States.
McLachlan, M.S., Kierkegaard, A., Hansen, K.M., van Egmond, R., Christensen, J.H., and Skjoth, C.A., 2010. Concentrations and fate of decamethylcyclopentasiloxane (D(5)) in the Atmosphere. Environ. Sci. Technol., 44, 5365.
Moretto, H.H., Schulze, M., and Wagner, G., 2000. Silicones. In: Ullmann's encyclopedia of industrial chemistry. Wiley-VCH, Weinheim.
Newman, M.C., and Unger, M.A., 2002. Introduction, pp. 3, in: Fundamentals of ecotoxicology, Newman, M.C., Unger, M.A. (Eds.), CRC Press, Boca Raton, United States.
Chapter 1
30
Penner, D., Burow, R., and Roggenbuck, F.C., 1999. Use of organosilicone surfactants as agrichemical adjuvants, pp. 241, in: Silicone surfactants, Surfactant science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker, New York.
Powell, D.E., Annelin, R.B., and Gallavan, R.H., 1999. Silicone in the environment: a worst-case assessment of poly(dimethylsiloxane) (PDMS) in sediments. Environ. Sci. Technol., 33, 3706.
Powell, D.E., and Carpenter, J.C., 1997. Polyethermethylsiloxanes, pp. 225, in: Organosilicon materials, Chandra G. (Ed.), Springer-Verlag, Berlin, Heidelberg.
Purcell, M.F. and Schröder, W.J., 1996. Effect of Silwet L-77 and diazinon on three tephritid fruit flies (Diptera: Tephritidae) and associated endoparasitoids. J. Econom. Entomol., 89, 1566.
Radulovic, J., Sefiane, K., and Shanahan, M.E.R., 2010. Ageing of trisiloxane solutions. Chem. Eng. Sci., 65, 5251.
Radulovic, J., Sefiane, K., and Shanahan, M.E.R., 2009. Spreading and wetting behaviour of trisiloxanes. J. Bionic Eng., 6, 341.
Rasi, S., Läntelä, J., and Rintala, J., 2011. Trace compounds affecting biogas utilisation – A review, Energ. Conves. Manage., 52, 3369.
Richardson, S.D. and Ternes, T.A., 2011. Water analysis: emerging contaminants and current issues. Anal. Chem., 83, 4614.
Robinson, R., and Kipping, F.S., 1908. XLIII.-Organic derivatives of silicon. Part V. Benzylethylsilicone, dibenzylsilicone, and other benzyl and benzylethyl derivatives of silicone. J. Chem. Soc., Trans., 93, 439.
Rösch, L., John, P., and Reitmeier, R., 2000. Silicon compounds, organic. in: Ullmann's encyclopedia of industrial chemistry. Wiley-VCH, Weinheim.
Sabourin, C.L., Carpenter, J.C., Leib, T.K., and Spivack, J.L., 1996. Biodegradation of dimethylsilanediol in soils. Appl. Environ. Microb., 62, 4352.
Sanchez-Brunete, C., Miguel, E., Albero, B., and Tadeo, J.L., 2010. Determination of cyclic and linear siloxanes in soil samples by ultrasonic-assisted extraction and gas chromatography-mass spectrometry. J. Chromatogr. A, 1217, 7024.
Schlabach, M., Strand Andersen, M., Green, N., Schoyen, M., and Kaj, L., 2007. Siloxanes in the environment of the inner Oslofjord, Norwegian Institute for Air Research (Eds.).
Schwarzenbach, R.P., Escher, B.I., Fenner, K., Hofstetter, T.B., Johnson, C.A., Urs, v.G., and Wehrli, B., 2006. The challenge of micropollutants in aquatic systems. Sci., 313, 1072.
Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M., 2003. Environmental organic chemistry. John Wiley & Sons, Hoboken, New Jersey.
Chapter 1
31
Shapiro, J.P., Schroeder, W.J., and Stansly, P.A., 1998. Bioassay and efficacy of Bacillus
thuringiensis and an organosilicone surfactant against the citrus leafminer (Lepidoptera : Phyllocnistidae). Fla. Entomol., 81, 201.
Snow, S.A., and Stevens, R.E., 1999. The science of silicone surfactant application in the formulation of polyurethane foam, pp. 137, in: Silicone surfactants, Surfactant science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker, New York.
Sommerlade, R., Parlar, H., Wrobel, D., and Kochs, P., 1993. Product analysis and kinetics of the gas-phase reactions of selected organosilicon compounds with OH radicals using a smog chamber-mass spectrometer system. Environ. Sci. Technol., 27, 2435.
Sparham, C., van Egmond, R., OConnor, S., Hastie, C., Whelan, M., Kanda, R., and Franklin, O., 2008. Determination of decamethylcyclopentasiloxane in river water and final effluent by headspace gas chromatography/mass spectrometry, J. Chromatogr. A, 1212, 124.
Sparham, C., van Egmond, R., Hastie, C., OConnor, S., Gore, D., and Chowdhury, N., 2011. Determination of decamethylcyclopentasiloxane in river and estuarine sediments in the UK. J. Chromatogr. A, 1218, 817.
Stark, J.D., and Walthall, W.K., 2003. Agricultural adjuvants: acute mortality and effects on population growth rate of Daphnia pulex after chronic exposure. Environ. Toxicol Chem., 22, 3056.
Stevens, C., 1998. Environmental degradation pathways for the breakdown of polydimethylsiloxanes. J. Inorg. Biochem., 69, 203.
Stutton, P.M., Rittmann, B.E., Schraa, O.J., Banaszak, J.E., and Togna, A.P., 2011. Wastewater as a resource: a unique approach to achieving energy sustainability. Water Sci. Technol., 63.
Tipping, C., Bikoba, V., Chander, G.J., and Mitcham, E.J., 2003. Efficacy of Silwet L-77 against several arthropod pests of table grape. J. Econom. Entomol., 96, 246.
Toshio, F., Keiichiro, N., Chiyozo, T., Masanori, Y., and Matazaemon, U., 2001. Hydrophobicity as a key physicochemical parameter of environmental toxicology of pesticides, Handbook of pesticide toxicology, Krieger, R.I., Krieger, W.C. (Eds.), Elsevier, Amsterdam.
United Nations, United Nations Environment Programme, 2009. Stockholm convention on persistent organic pollutants (POPs), http://chm.pops.int/TheConvention/Overview/TextoftheConvention/tabid/2232/Default.aspx (Access June 2014)
United States Environmental Protection Agency, 2013. Environmental fate summary, http://www.epa.gov/dfe/pubs/tools/ctsa/ch5/mod5-3.pdf (Access June 2014)
Vasiljevic T., Grujić S., Radišić M., Dujaković N., and Laušević M., 2013. Pesticide residues in surface water and groundwater, pp. 259, in: Pesticides evaluation of environmental pollution, Rathore H.S. and Nollet L.M.L. (Eds.),. CRC Press, Boca Raton, United States.
Chapter 1
32
Venzmer, J., 2011. Superspreading - 20 years of physicochemical research. Curr. Opin. Colloid In., 16, 335.
Vick, S.C., 1984. Structure property relationships for silicone polyalkyleneoxide copolymers and their effects on performance in cosmetics. Soap Cosmet. Chem. Spe., 60, 36.
Vighi, M. and Calamari, D., 1993. Prediction of the environmental fate of chemicals. Ann. Ist. Super. Sanita, 29, 209.
Wang, D.G., Norwood, W., Alaee, M., Byer, J.D., and Brimble, S., 2013. Review of recent advances in research on the toxicity, detection, occurrence and fate of cyclic volatile methyl siloxanes in the environment. Chemosphere, 93, 711.
Wang, X.M., Lee, S.C., Sheng, G.Y., Chan, L.Y., Fu, J.M., Li, X.D., Min, Y.S., and Chan, C.Y., 2001. Cyclic organosilicon compounds in ambient air in Guangzhou, Macau and Nanhai, Pearl River Delta. Appl. Geochem., 16, 1447.
Warner, N.A., Evenset, A., Christensen, G., Gabrielsen, G.W., Borga, K., and Leknes, H., 2010. Volatile siloxanes in the European Arctic: assessment of sources and spatial distribution. Environ. Sci. Technol., 44, 7705.
Whelan, M.J., Estrada, E., and van Egmond, R., 2004. A modelling assessment of the atmospheric fate of volatile methyl siloxanes and their reaction products. Chemosphere, 57, 1427.
Whelan, M.J., Sanders, D., and van Egmond, R., 2009. Effect of Aldrich humic acid on water-atmosphere transfer of decamethylcyclopentasiloxane. Chemosphere, 74, 1111.
Whelan, M.J., van Egmond, R., Gore, D., and Sanders, D., 2010. Dynamic multi-phase partitioning of decamethylcyclopentasiloxane (D5) in river water, Water Res., 44, 3679.
World Health Organization and United Nations Children's Fund, 2000. Global water supply and sanitation assessment 2000 report, http://www.who.int/water_sanitation_health/monitoring/jmp2000.pdf (Access June 2014)
Xu, L., Shi, Y., and Cai, Y., 2013. Occurrence and fate of volatile siloxanes in a municipal wastewater treatment plant of Beijing, China. Water Res., 47, 715.
Xu, S.H., 1999. Fate of cyclic methylsiloxanes in soils. 1. The degradation pathway. Environ. Sci. Technol., 33, 603.
Xu, S.H. and Chandra, G., 1999. Fate of cyclic methylsiloxanes in soils. 2. Rates of degradation and volatilization. Environ. Sci. Technol., 33, 4034.
Xu, S.H., Lehmann, R.G., Miller, J.R., and Chandra, G., 1998. Degradation of polydimethylsiloxanes (silicones) as influenced by clay minerals. Environ. Sci. Technol., 32, 1199.
Yucuis, R.A., Stanier, C.O., and Hornbuckle, K.C., 2013. Cyclic siloxanes in air, including identification of high levels in Chicago and distinct diurnal variation. Chemosphere, 92, 905.
Chapter 1
33
Zhang, Z.F., Qi, H., Ren, N.Q., Li, Y.F., Gao, D.W., and Kannan, K., 2011. Survey of cyclic and linear siloxanes in sediment from the Songhua River and in sewage sludge from wastewater treatment plants, Northeastern China. Arch. Environ. Cont. Tox., 60, 204.
Chapter 2
34
Chapter 2
Development of a liquid chromatography tandem
mass spectrometry method for trace analysis of
trisiloxane surfactants in the aqueous
environment: an alternative strategy for
quantification of ethoxylated surfactants
Published in Journal of Chromatography A, 1245 (2012) 46-54
Chapter 2
35
TABLE OF CONTENTS
Abstract .................................................................................................................................... 36
1. Introduction ...................................................................................................................... 37
2. Experimental ........................................................................................................................ 40
2.1 Reagents and materials ............................................................................................... 40
2.2 Sample preparation ..................................................................................................... 40
2.3 Liquid chromatography-mass spectrometry ............................................................... 40
2.4 Method validation ....................................................................................................... 41
3. Results and discussion ...................................................................................................... 43
3.1 Characterization of the stock solution ......................................................................... 43
3.1.1 Theoretical approach: Poisson distribution ........................................................ 43
3.1.2 Experimental data: single mass spectrometry .................................................... 44
3.2 Liquid chromatography ............................................................................................... 48
3.3 Validation .................................................................................................................... 51
Conclusion ................................................................................................................................ 54
Acknowledgment ..................................................................................................................... 55
References ................................................................................................................................ 55
Appendix A: Supplementary data ............................................................................................ 58
A.1 Theoretical approach: Poisson distribution ................................................................... 58
A.2 Experimental approach: LC-SPE-NMR/MS ................................................................. 59
A.2.1 Materials and method ............................................................................................. 59
A.2.2 Results and discussion ............................................................................................ 59
References ................................................................................................................................ 62
Chapter 2
36
Abstract
Trisiloxane surfactants, often referred to as superspreaders or superwetters, are added to
pesticides to enhance the activity and the rainfastness of the active substance by promoting
rapid spreading over hydrophobic surfaces. To fill the lack of data on the environmental
occurrence of these compounds, we have developed and validated a method for their trace
analysis in the aqueous environment. The method is based on liquid-liquid extraction
followed by liquid chromatography and tandem mass spectrometry. The oligomeric
distribution of trisiloxane surfactant in a reference solution was determined by a theoretical
calculation and by experimental measurements. Based on these results, the quantification was
performed by comparison with a calibration made with a single homologue instead of a
mixture of homologues. This approach avoids a time-consuming synthesis of pure
homologues and reduces the risk of wrong estimation of the concentration because of
different response factors of the sample and the standard. Such an approach could be applied
to the quantification of other ethoxylated surfactants following a similar distribution. The
validation was performed from 2 ng L-1 to 250 ng L-1 (total surfactant concentration) in
deionized water, tap water, and river water (Rhine water). Knowing the oligomeric
distribution of the polymer in the reference solution, the corresponding calibration ranges
were estimated for individual homologues. Limits of quantification were found to be between
0.37 ng L-1 and 15 ng L-1. The total recovery of sample preparation was between 77 % and
116 %. Matrix effects were lower than 10 % with river water and the relative standard
deviation evaluated over ten identical samples of spiked river water was below 12 %.
Chapter 2
37
1. Introduction
Because of their outstanding surface properties, silicone surfactants or
polyethermethylsiloxanes (PEMS) are today added to a broad range of products, e.g.
pesticides, cosmetics, polyurethane foams, textiles and fibers, paints and coatings, and car
care products. They were first commercialized in the 1950s for the production of polyurethane
foam and this application is still their largest commercial market. In addition, their high
efficiency to reduce surface tension and their adaptability led to a diversification of their
applications (Hill, 1999; Chandra, 1997). Like all other surfactants, PEMS are composed of a
hydrophilic and a hydrophobic part. The hydrophobic part is made of a flexible polysiloxane
backbone, and the hydrophilic part of neutral PEMS is composed of a spacer, ethylene oxide
units (EO), propylene oxide units (PO) and an end group (R). The general structure of silicone
surfactants is shown in Figure 1 (Hill, 1999). The molecule can be tailored toward a specific
application by changing the polymerization degree of the siloxane chain (x and y), the length
of the polyether chain (n and m), and the end group. Increasing the number of EO groups
increases, for example, the water solubility of the molecule (Vick, 1984).
R = H, CH3, C(O)CH3
Figure 1: General chemical structure of silicone surfactants.
Trisiloxane surfactants (x = 1, y = 0, and m = 0) are used as agricultural adjuvants to increase
the activity and the rainfastness of pesticides. They are often referred to as superspreaders or
superwetters because of their ability to promote rapid spreading over the hydrophobic surface
of leaves (Ananthapadmanabhan, 1990; Hill, 1998; Penner et. al., 1999; Radulovic et. al.,
2009; Venzmer, 2011; Zhu, 1994). Even if the mechanism of superspreading is not fully
understood yet, studies have proven that the best wetting agents contain between 2 and 5
silicon atoms (Kanner et. al., 1967). Trisiloxane surfactants are water soluble and may reach
surface waters after their application in agriculture by runoff and emissions through
x y
n m
OR
Chapter 2
38
wastewaters from cleaning of the spray equipment. The growing interest for silicone based
agricultural adjuvants raises the question of their occurrence in the aqueous environment.
Only a few scientific articles are related to trisiloxane surfactants in the environment: A
HPLC-MS/MS method has been developed to characterize one silicone based agricultural
adjuvant and has been applied to investigate its foliar uptake and its degradation (Bonnington,
2000; Bonnington et. al., 2004). The behavior of silicone polyether in different soils has been
simulated (Griessbach et. al., 1997; Griessbach et. al., 1998), but to the best of our knowledge
no measurements of environmental concentrations of trisiloxane surfactants have been
performed and no method for trace analysis of trisiloxane surfactants in environmental
samples has been described. Most of the literature on the environmental occurrence and fate
of organosilicones is dedicated to polydimethylsiloxanes (PDMS) (Chandra, 1997) and cyclic
siloxanes. Since a few years, cyclic siloxanes have been appearing as new subject of concern
because of their toxic properties and their persistence in the environment (Nordic Council of
Ministers, 2005; Environment Canada, 2008; Environment Agency, 2009; Kierkegaard et. al.,
2010; McLachlan et. al., 2010). By hydrolysis of the siloxane chain of trisiloxane surfactants,
silanols with Si-OH function(s) can be formed and can condensate to form various linear or
cyclic siloxanes. Cyclic siloxanes have been described as degradation products of trisiloxane
surfactants (Bonnington, 2000) and it is likely that the degradation process of PEMS leads to
the formation of cyclic siloxanes, like octamethylcyclotetrasiloxane (D4) and
decamethylcyclopentasiloxane (D5). The first step for further research on the environmental
occurrence and the fate of trisiloxane surfactants is to be able to analyze them at low
concentrations (ng L-1) in complex matrices. This study describes the development and
validation of a method for trace analysis of trisiloxane surfactants in the aqueous
environment. The target molecules of the novel method are described in Table 1. According
to the silicone nomenclature (Hill, 1999), they are denoted MD´(-(CH2)3-(EO)n-OCH3)M and
D´(-(CH2)3-(EO)n-OH)M where M = (CH3)3-Si-O1/2-, D´ = O1/2-Si(CH3)(R´)-O1/2- with
R´ = -(CH2)3-(EO)n-OR and EO = OCH2CH2. MD´(-(CH2)3-(EO)n-OCH3)M is well-known
under its commercial name, Silwet® L77 (originally Union Carbide, now Momentive).
Chapter 2
39
Table 1: Description of the target trisiloxane surfactants: abbreviations, chemical structures and CAS
numbers. Abbreviation with silicone nomenclature
MD´(-(CH2)3-(EO)n-OH)M MD´(-(CH2)3-(EO)n-OCH3)M
CAS N° 67674-67-3 27306-78-1
Chemical structure
Trisiloxane surfactants used in agriculture are commercialized as mixtures of homologues,
with an average number of 7.5 ethylene oxide groups (Venzmer, 2011). The comparison of
the oligomeric distribution of the surfactant in commercial products and in environmental
samples may be important to identify the origin of trisiloxane surfactants in the environment
and to understand their degradation (Ventura and deVoogt, 2003). We have therefore chosen
to develop a method which quantifies every homologue separately and allows obtaining
information on the oligomeric distribution of the polymer. The development of such a method
faces the lack of standards of individual homologues. The lack of standards is a recurring
problem for the analysis of surfactants (Petrovic et. al., 2001) and two main strategies are
usually applied (deVoogt and Knepper, 2003). A first possibility is to use a commercial
mixture as standard. In this case, the accuracy of the results depends on the similarity of the
molecular weight distribution (MWD) in the standard and the sample. If the oligomeric
composition in the sample differs from the one in the standard, the response factors may be
different and this approach can lead to a wrong estimation of the concentration in the sample.
The second option is to synthesize pure oligomers of the target surfactant, but this synthesis is
time-consuming and requires characterizing the purity of the obtained product. The method
described in this work proposes an alternative strategy. The individual concentrations in a
reference solution were estimated by two approaches: i) a theoretical calculation based on the
Poisson distribution of ethylene oxide polymers and ii) an experimental determination by
single mass spectrometry. By combining the two approaches, a reference solution where all
concentrations are known was generated and was used as a standard for quantification. This
approach constitutes a reliable alternative to overcome the lack of analytical standards.
n n
Chapter 2
40
2. Experimental
2.1 Reagents and materials
Reference substances of the two trisiloxane surfactants were provided by ABCR (Karlsruhe,
Germany). Each trisiloxane surfactant is a mixture of homologues with different numbers of
ethylene oxide groups. The average number of ethylene oxide units is 7.5 (Venzmer, 2011)
but various chain lengths are present in the product and homologues constitute a so-called
oligomeric distribution. Stock solutions were prepared at 1000 mg L-1 and 1 mg L-1 (total
surfactant concentration) in methanol. They were then stored at -18 °C. All organic solvents
used for the analysis were HPLC-grade. Their purity was at least 99.8 %. Methanol and
dichloromethane were provided by Carl Roth GmbH (Karlsruhe, Germany), tetrahydrofurane
and isopropanol were purchased from Merck (Darmstadt, Germany) and acetonitrile was
obtained from Scharlau S.L. (Sentmenat, Spain). Ultrapure water was produced by an Arium
611 UV laboratory water purification system from Sartorius (Göttingen, Germany).
Ammonium acetate was purchased from Fluka (Steinheim, Germany, purity higher than
98.0 %). All glassware was treated in a pyrolysis oven from Carbolite (Hope Valley, England)
at 550 °C for 60 min before use.
2.2 Sample preparation
200 mL of water sample were placed in a 250 mL glass bottle and 20 mL of dichloromethane
were added. The sample was stirred on a magnetic stirrer for 30 min before pouring it into a
separating funnel of 250 mL. The glass bottle was rinsed with 5 mL dichloromethane which
was added to the separating funnel. After decantation, the organic phase was collected in a
glass vial, blown down to dryness under a gentle nitrogen stream and then reconstituted with
250 µL methanol and 250 µL ultrapure water. In the case of spiked water samples, the spiked
volume of stock solution was kept below 0.01 % to avoid modifying the extraction recoveries
by adding methanol. The spiked water was homogenized by 5 min stirring before liquid-liquid
extraction.
2.3 Liquid chromatography-mass spectrometry
The analysis was performed by using a model 1200 HPLC system (Agilent Technologies,
Waldbronn, Germany) coupled with a triple-quadrupole mass spectrometer. The HPLC unit
was equipped with a solvent cabinet, a micro vacuum degasser, a binary pump, a high-
performance autosampler with 54 vial plates and a temperature controlled column
Chapter 2
41
compartment. The separation was carried out on a PolymerX RP-1 column
(250 mm × 4.6 mm i.d., 5 µm, 100 Å) from Phenomenex. The injection volume was set at
40 µL, the flow rate was 0.8 mL min-1 and the column compartment was regulated at 20 °C.
The solvents were: (A) 48 % acetonitrile, 40 % water 10 mM ammonium acetate, 6 %
methanol, 6 % tetrahydrofurane, (B) tetrahydrofurane. The elution gradient started with
100 % of solvent A and decreased to 40 % of A within 24 min. This composition was held for
3 min before going back to initial conditions. The column was equilibrated at 100 % of A for
8 min between each run. The detection was achieved with a triple quadrupole mass
spectrometer API 4000 Q-Trap (Applied Biosystems/MDS Sciex Instruments, Concord,
Canada) equipped with an ESI source and used in multiple reaction monitoring (MRM). The
mass spectrometer operated in positive ionization mode and silicone surfactants were detected
as ammonium adducts. Two product ions were used for every homologue: One for
quantification (m/z 103 for MD´(-(CH2)3-(EO)n-OCH3)M and m/z 147 for MD´(-(CH2)3-
(EO)n-OH)M) and one for confirmation. The MS parameters were determined successively
for the two target trisiloxane surfactants by direct infusion of each surfactant homologue
mixture at a total concentration of 0.5 mg L-1 in MeOH/20 mM aqueous ammonium acetate
solution and are summarized in Table 2. The settings of the ESI source were: ion spray
voltage: 5.5 kV, heater temperature: 500°C, collision gas: medium, ion source gas 1/2:
70/90 psi and curtain gas: 30 psi. Data acquisition was performed with Analyst software
(version 1.5.1). An extracted ion chromatogram was generated for every homologue and the
area obtained by integration of the peak was used for quantification. The attribution of a peak
in a sample was confirmed by the retention time, the presence of the two products ions, and
the ratio between the two products ions.
2.4 Method validation
A ten-point calibration between 2 ng L-1 and 250 ng L-1 (total surfactant concentration) was
established by applying the whole sample preparation procedure to ten samples of deionized
water spiked with trisiloxane surfactants standards. The limit of quantification (LOQ) was
calculated with the SQS software (version 2.01) according to DIN 32645 (Deutsches Institut
für Normung, 1994). The absolute recovery of the whole sample preparation was calculated at
20 % and 80 % of the calibration range by comparing the peak area of the calibration points
with the peak area of a direct injection. Relative standard deviation (RSD) was evaluated on
ten replicates at two different concentrations. Similar procedures (calibration, recovery and
LOQ determinations) were performed in river water (river Rhine) and tap water (Karlsruhe) to
Chapter 2
42
determine matrix impact on linearity, recovery and LOQs. To estimate the matrix effect on
ionization in the interface of the mass spectrometer, a blank sample of river Rhine water was
extracted and trisiloxane surfactants were spiked to the extract, after its evaporation. The
sample was compared to a direct injection at the same concentration. The eventual loss of
compound during evaporation of the extract was also estimated by comparing a sample spiked
before and after evaporation.
Table 2: Precursor ions, product ions and corresponding MS parameters (DP declustering potential, CE
collision energy, CXP cell exit potential).
Precursor ion Product ion DP (V) CE (eV) CXP (V)
MD´(-(CH2)3-(EO)n-OCH3)M 488.4
103.1 381.2
86 35 23
18 10
532.5 102.9 147.1
91 35 29
18 8
576.5 103.0 147.0
91 41 33
26 28
620.5 103.0 147.0
101 45 37
18 8
664.5 102.9 221.1
131 51 51
20 44
708.4 103.0 221.1
91 43 63
18 22
752.5 103.2 221.1
106 49 63
8 14
796.4 103.0 147.1
146 51 41
18 8
840.8 103.0 221.1
76 51 73
18 16
MD´(-(CH2)3-(EO)n-OH)M 474.5
147.0 189.1
51 39 33
8 12
518.6 147.1 321.1
71 39 23
14 8
562.4 147.1 365.1
86 37 23
10 10
606.7 147.1 409.2
91 39 25
8 12
650.6 147.0 453.2
96 41 27
8 14
694.5 147.1 77.0
106 43 77
8 14
738.6 147.0 191.1
96 43 39
8 12
782.7 147.0 191.1
116 45 43
8 12
826.6 147.1 191.1
81 47 43
8 12
Chapter 2
43
3. Results and discussion
3.1 Characterization of the stock solution
Quantitative analytical methods are based on the comparison of a sample with a calibration
made by successive dilutions of a standard of known concentration. No analytical standards of
individual homologues were available for trisiloxane surfactants but the individual
concentrations in a reference solution were estimated by two approaches: A theoretical
approach and experimental measurements.
3.1.1 Theoretical approach: Poisson distribution
The theoretical molecular weight distribution of the homologues of trisiloxane surfactants
can be predicted by knowing the synthesis route. The different steps and the
different reagents are simplified in Figure 2.
Figure 2: Schematic representation of the synthesis route of trisiloxane surfactants.
Since trisiloxane surfactants are synthesized by reaction of 1,1,1,3,5,5,5-
heptamethyltrisiloxane with an allyloxypolyetheroxide containing an average of 7.5 ethylene
oxide groups (reaction (3)) (Kanner et. al., 1967; Union Carbide Corporation, 1967), the
oligomeric distribution is expected to be centered on 7.5. In addition, the synthesis route
indicates that the oligomer composition of trisiloxane surfactant directly depends on the
oligomer composition of R-O-[EO]n-H. As demonstrated by Flory (Flory, 1940), ethylene
oxide polymers obtained by reaction (1) follow a Poisson distribution. The expected MWD of
the trisiloxane surfactants is therefore a Poisson distribution centered on 7.5. A mathematical
treatment of the problem leads to equations (1) and (2) and allows calculating the theoretical
concentrations of all homologues in the stock solution (1 mg L-1):
)!1n(M
M
1
n1n
1n
(1)
Si O Si O Si
CH3CH3
CH3
CH3
CH3
CH3
CH3
H
CH2 CH CH2 EO O R Si O Si O Si
CH3CH3
CH3
CH3
CH3
CH3
CH3
CH2
CH2
CH2
EO
O
R
R O EO H
CH2 CH CH2 XO
R = H, CH3
X = undisclosed halogen
n n
n-1
R O EO H
n
(1) (2) (3)
Chapter 2
44
L/mg11n
ntot
(2)
βn is the mass concentration of the homologue containing n ethylene oxide groups (in mg L-1),
Mn is the molecular weight of homologue n (in g mol-1) and ν is the number of propagating
units reacted per initiator. The detailed calculation is presented in the supplementary material.
Measurements by single mass spectrometry (Figure 3) were carried out to compare
experimental data to the distribution calculated by mathematical treatment. The
concentrations estimated by theoretical and experimental approaches are compared in
Figure 4.
3.1.2 Experimental data: single mass spectrometry
Adducts of MD´(-(CH2)3-(EO)n-OCH3)M with NH4+, K+, Na+ and H+ are reported
(Bonnington, 2000). In this work, ammonium acetate was added to the HPLC solvents to
simplify the spectra by promoting the formation of only one type of adduct. This salt was
chosen for its compatibility with the ESI source of the mass spectrometer and because
polyethylene oxide is known to have a strong complexation ability toward K+, which has a
similar radius as NH4+ (Yanagida et. al., 1977; Yanagida et. al., 1978). Figure 3 shows the
addition of ten multi channel analysis obtained by direct injection of the target compounds at
0.5 mg L-1 in MeOH/20 mM aqueous ammonium acetate solution 1/1 (v/v). Like every
polymer containing EO units, a characteristic pattern was obtained with equidistant signals at
Δm/z 44. The homologues below n = 3 for MD´(-(CH2)3-(EO)n-OCH3)M and n = 4 for MD´(-
(CH2)3-(EO)n-OH)M) could not be distinguished from the background noise.
Chapter 2
45
Figure 3: Mass spectra obtained by direct infusion of the trisiloxane surfactants at 0.5 mg L-1 in
MeOH/20 mM aqueous ammonium acetate solution 1/1 (v/v). a: MD (́-(CH2)3-(EO)n-OH)M.
b: MD (́-(CH2)3-(EO)n-OCH3)M.
The mass spectra could be used to estimate the MWD and the individual concentrations of all
homologues but this calculation required two assumptions. The first approximation is that the
response of the detector is the same for every homologue, which means that the intensity
measured by single MS for homologue n is directly proportional to its molar concentration:
In = B × cn (3)
m/z in Da
200 400 600 800 1000
Inte
nsit
y in
cps
0
5e+6
1e+7
2e+7
2e+7
n = 13870.6
n = 8650.6
n = 9694.5
n = 10738.6
n = 7606.7
n = 6562.4
n = 5518.6
n = 4474.5
n = 11782.7
n = 12826.6
n = 14914.7
n = 15958.6
m/z in Da
200 400 600 800 1000
Inte
nsit
y in
cps
0
5e+6
1e+7
2e+7
2e+7
n = 13884.7
n = 8664.6
n = 9708.6
n = 10752.8
n = 7620.6
n = 6576.7
n = 5532.4
n = 4488.5
n = 11796.7
n = 12840.7
n = 14928.8
n = 3442.5 n = 15
972.7
a
b
Chapter 2
46
Where In is the intensity measured on the mass spectrum (in cps) (Figure 3), cn is the molar
concentration of homologue n (in mol L-1) and B is a common constant to all homologues.
With nMnβ
nc , equation (3) could be written:
In = B × n
n
M
(4)
Another expression of βn could be deduced from equation (4) by introducing the ratio In/In-1:
1n1n
nn1nn
MI
MI
(5)
The second approximation consisted of limiting the calculation to the homologues which
could be detected by MS (3 ≤ n ≤ 15 for MD´(-(CH2)3-(EO)n-OCH3)M and 4 ≤ n ≤ 15 for
MD´(-(CH2)3-(EO)n-OH)M) and assuming that the concentrations of all other homologues
could be neglected. Equation (5) can thus be written:
4or 3 4or 3 4or 3
nnn β
M I
MIβ (6)
To calculate all βn, a boundary condition was required: Since the amount of trisiloxane
surfactant used to prepare the stock solution was known, the overall mass concentration (βtot
in mg L-1) was known:
βtot =
15
4or3nnβ = 1 mg L-1 (7)
The set of equations formed by (6) and (7) allows estimating the concentrations in the stock
solution by using the intensities measured by mass spectrometry, weighted by the molecular
masses of individual homologues. Calculated concentrations are represented by squares in
Figure 4. Figure 4 compares the concentrations obtained by theoretical and experimental
determinations. The percentage difference was calculated as the standard deviation between
the concentrations obtained with the two approaches divided by the mean value. Further
method development was limited to 4 ≤ n ≤ 12. Outside this range, individual
concentrations of the homologues were lower than 5 % of the total concentration of
surfactant. In addition, high deviations were observed. For MD´(-(CH2)3-(EO)n-OCH3)M,
within the 4 ≤ n ≤ 12 range, the deviation between the concentrations calculated with
experimental and theoretical approaches were lower than 10 % (except for n = 4). This
homologue was however kept in the method to keep certain symmetry of the distribution.
Chapter 2
47
Figure 4: Comparison of the individual concentrations of homologues in the stock solution calculated with
theoretical ( ) and experimental ( ) approaches. a: MD (́-(CH2)3-(EO)n-OH)M;
b: MD (́-(CH2)3-(EO)n-OCH3)M. The numbers on top of each point give the absolute percentage difference
between the results obtained by the two approaches.
The very good correlation between the theoretical and experimental approaches confirmed
the validity of the two assumptions: The contribution of expelled homologues was
negligible and the response of the mass spectrometer did not strongly differ from one
homologue to the other. Because of the good agreement between the concentrations
obtained with different approaches, the calculated composition of the stock solution could
Number of ethylene oxide groups
5 10 15 20
Con
cent
rati
on i
n µ
g L
-1
0
50
100
150
200
106%
30%
2%
5%1% 14%
27%
15%
24%
8%
3%16%
Number of ethylene oxide groups
5 10 15 20
Con
cent
rati
on i
n µ
g L
-1
0
50
100
150
200
19%
2%
3%
3% 8%
6%
9%
3%
1%
11%
55% 77%
12%
a
b
Chapter 2
48
be considered as reliable and the Poisson distribution appeared as a good model of the
MWD of trisiloxane surfactants. Regarding MD´(-(CH2)3-(EO)n-OH)M, the deviations
between the theoretical and experimental concentrations within the 4 ≤ n ≤ 12 range were
higher than for MD´(-(CH2)3-(EO)n-OCH3)M. The difference was however lower than
30 % (except for the homologue n = 4), which indicates a reasonable correlation between
the two approaches. Since it was impossible to assert which distribution is closer to reality,
both were kept for further method development. Each quantification was made by referring
to both reference concentrations and the difference was taken as an estimation of the
uncertainty on the concentration. Knowing the oligomer composition of the stock solution,
one calibration could be generated for each homologue, based on the individual
concentration in the reference solution and on the area integrated from the extracted ion
chromatogram. Without analytical standards for individual oligomers of trisiloxane
surfactants and internal standards, the method should be described as semi-quantitative.
Since our purpose was to provide a first screening of the occurrence of trisiloxane
surfactants in environmental waters, this approach was considered as suitable and reliable
enough.
Additional LC-SPE-NMR/MS measurements were performed and support the oligomeric
distribution of MD´(-(CH2)3-(EO)n-OCH3)M measured by MS and calculated by Poisson
distribution. The results are given in the supplementary material.
3.2 Liquid chromatography
The chromatographic separation was first developed on a Kinetex C18 column
(100 mm × 2.1 mm i.d., 2.6 µm 100 Å) from Phenomenex with isocratic conditions at 60 % of
a ternary mixture (80 % ACN, 10 % MeOH, 10 % THF) and 40 % of water with 20 mM
ammonium acetate. This column was chosen among other tested columns because it allowed a
baseline separation of homologues (Figure 5a). This ability to separate molecules with very
similar structures (the difference between two homologues only consists in one ethylene oxide
group) may be explained by the narrow particles size distribution and by the nature of the not
fully porous silica particles of the Kinetex column. However, a quick loss of the
chromatographic performance of the Kinetex C18 appeared when analyzing samples of
silicone surfactants. 60 consecutive injections of a sample at 0.1 mg L-1 in MeOH/water 1/1
(v/v) (total surfactant concentration) led to a decrease of the resolution from 1.7 to 0.5 for
MD´(-(CH2)3-(EO)n-OH)M homologues and from 3.0 to 2.0 for MD´(-(CH2)3-(EO)n-OCH3)M
homologues. Two successive measurements of one sample (trisiloxane surfactants at
Chapter 2
49
0.1 mg L-1 in MeOH/water 1/1 (v/v)) on the same HPLC-MS/MS system but on two different
Kinetex C18 columns (new and used) proved that the HPLC column aging was responsible
for this loss of performance (Figure 5a and 5b). An important difference of intensity was
noticed between the two chromatograms. The loss of chromatographic performance could
have led to co-elutions of homologues and thus to ion suppression.
Figure 5: HPLC-MS/MS extracted ion chromatograms of a mixture MD (́-(CH2)3-(EO)n-OH)M and
MD (́-(CH2)3-(EO)n-OCH3)M at 0.1 mg L-1 (total surfactant concentration) on two different Kinetex C18
columns. a: new Kinetex C18. b: used Kinetex C18.
Time in min
10 20 30 40
Inte
nsit
y in
cps
0
1e+5
2e+5
3e+5
4e+5
5e+5
n=4
n=5
n=7
n=8
n=9
n=10
n=11
n=12
n=3n=4
n=6
n=7n=9
n=10
n=11
n=6
n=8 n=5
MD´(-(CH2)3-(EO)n-OCH3)-M
MD´(-(CH2)3-(EO)n-OH)M
Time in min
10 20 30 40
Inte
nsit
y in
cps
0
1e+5
2e+5
3e+5
4e+5
5e+5
MD´(-(CH2)3-(EO)n-OCH3)M
MD´(-(CH2)3-(EO)n-OH)M
a
b
Chapter 2
50
A degradation of performance of an HPLC column during the analysis of silanols has already
been mentioned in literature (Dorn, 1994) and was believed to be due to a strong interaction
between silanols and the silica-based stationary phase of the column. Based on these findings,
a polymeric column made of polystyrene divinylbenzene was chosen for the separation of
trisiloxane surfactants and showed a good stability of the chromatographic performance. A
chromatogram obtained with the polymeric column for MD´(-(CH2)3-(EO)n-OCH3)M and
MD´(-(CH2)3-(EO)n-OH)M is represented in Figure 6.
Figure 6: HPLC-MS/MS chromatograms of a mixture of the trisiloxane surfactants at 0.1 mg-1 (total
surfactant concentration) with a PolymerX column. a: MD (́-(CH2)3-(EO)n-OH)M.
b: MD (́-(CH2)3-(EO)n-OCH3)M.
Time in min
12 14 16 18 20
Inte
nsit
y in
cps
0.0
2.0e+4
4.0e+4
6.0e+4
8.0e+4
1.0e+5
1.2e+5
1.4e+5
n = 4
n = 5
n = 6
n = 7
n = 8
n = 9
n = 10
n = 11
n = 12
a
Time in min
16 18 20 22
Inte
nsit
y in
cps
0
10000
20000
30000
40000
n = 4
n = 5
n = 6
n = 7
n = 8
n = 9
n = 10
n = 11
n = 12
b
Chapter 2
51
The separation obtained with the polymeric column is not as good as the one obtained with
the Kinetex column but this may not influence the quantification since the latter is performed
on extracted ion chromatograms. Possible competition between homologues in the source of
the mass spectrometer may be limited since homologues are still separated from each other.
3.3 Validation
Contamination of samples is common during the analysis of silicones (Smith and Parker,
1991). Even if the combination of liquid-liquid extraction and HPLC-MS/MS makes the
described method selective, it is important to keep this problem in mind when analyzing
trisiloxane surfactants. Silicone surfactants are used in many personal care products under the
denomination dimethicone copolyol (Floyd, 1999). A sample of a commercial hair styling
product containing dimethicone copolyol was analyzed to check the presence of the target
trisiloxane surfactants and the possibility of contamination from the operator. The sample was
prepared by diluting 1 g of product in 1 mL of MeOH/water 1/1 (v/v) and was analyzed with
HPLC-(ESI)-MS/MS and HPLC-(ESI)-TOF. The HPLC-MS/MS measurements of the sample
and the comparison with a sample of the same cosmetic spiked with the reference substances
showed the presence of MD´(-(CH2)3-(EO)n-OCH3)M (confirmation with retention time and
ratio between the quantification and confirmation product ions). The measurement with
HPLC-(ESI)-TOF confirmed with the exact mass and the isotopic pattern that
MD´(-(CH2)3-(EO)n-OCH3)M was present. These experiments proved that contaminations
from personal care products are possible. In order to prevent contamination, all personal care
products with the mentioned dimethicone copolyol were avoided. To make sure that no
problems of contamination by an external source of trisiloxane surfactants occurred, six blank
samples of deionized water and tap water were analyzed by following the whole analytical
procedure. No peaks with signal to noise ratios higher than ten were observed. Results of the
validation are summarized in Table 3. Calibrations were performed from 2 ng L-1 to
250 ng L-1 (overall surfactant concentration). Knowing the oligomeric distribution of the
reference solution (cf. Section 3.1.2), the corresponding concentrations of individual
homologues were estimated. For example for MD´(-(CH2)3-(EO)7-OCH3)M, the calibration
was from 0.29 ng L-1 to 37 ng L-1. Calibrations lines were obtained by plotting the area of the
peak determined from the extracted ion chromatogram of the considered homologue versus its
individual concentration and were fitted with linear least square regression. The correlation
coefficients were found to be higher than 0.990, except for homologues with n = 10, 11, 12 of
both trisiloxanes in deionized water, for which they were between 0.966 and 0.988. The
Chapter 2
52
linearity of the calibration curves was considered as satisfactory in the considered
concentrations range.
LOQ values were calculated from the residual standard deviation of the calibration lines
obtained with the two approaches used to characterize the reference solution (Poisson
distribution and single MS measurements). The given LOQ corresponds to a mean value and
the standard deviation was taken as the error on the determination. The LOQ were found to be
between 0.36 ng L-1 (MD´(-(CH2)3-(EO)4-OCH3)M in tap water) and 15.2 ng L-1
(MD´(- (CH2)3-(EO)9-OCH3)M in river water). Another method for the estimation of the LOQ
is to choose the lowest concentration that can be measured with a signal to noise ratio of at
least ten. For example for MD´(-(CH2)3-(EO)7-OCH3)M, the LOQ determined in tap water
with the signal to noise ratio would be 0.3 ng L-1 while the LOQ calculated with the DIN
method was 1.6 ng L-1. The higher LOQ obtained according to DIN 32645 is more
conservative and was therefore chosen for the estimation of the LOQ.
The total recovery of the sample preparation procedure was found to be between 77 %
(MD´(-(CH2)3-(EO)12-OH)M in river water, at 20 % of the calibration range) and 116 %
(MD´(-(CH2)3-(EO)12-OCH3)M in tap water, at 20 % of the calibration range). No significant
variation of the total recovery was observed between 20 % and 80 % of the calibration range
and between the three tested matrices. The total recovery does not discriminate between the
losses during sample preparation and ion suppression/enhancement in the interface of the
mass spectrometer. Since no internal standards were available for trisiloxane surfactants, it
was important to estimate matrix effects. For this purpose, a blank sample of river Rhine
water was extracted according to the liquid-liquid extraction procedure described in
Section 2.2. The extract was blown down to dryness and was spiked with trisiloxane
surfactants standards. The obtained sample was compared to an external standard and the
percentage difference was calculated as the ratio of the standard deviation over the mean
value. The results were always lower than 10 % indicating that no significant matrix effect
occurred in the tested matrix. The losses of compound during evaporation of the extract were
also estimated by comparing samples spiked before and after evaporation. No losses higher
that 10 % were observed. Relative standard deviation was estimated at 20 % and 80 % of the
calibration range by applying the whole analytical procedure to ten identical samples of river
Rhine water spiked with trisiloxane surfactants. RSD was calculated as the ratio of the
standard deviation to the mean value and was found to be between 3 % and 12 %.
Chapter 2
53
Table 3: Results of the validation: LOQs, correlation coefficients, sample preparation recovery, RSDs and matrix effects for all analytes. LOQ (ng L-1) Correlation coefficient Recovery (%) RSD (%)
Matrix effect (%)
Deionized water Tap water River water Deionized
water Tap
water River water
Deionized water Tap water River water River water River water
n MS Poisson LOQ
Std dev
MS Poisson LOQ Std dev
MS Poisson LOQ Std dev
50
ng L-1 200
ng L-1 50
ng L-1 200
ng L-1 50
ng L-1 200
ng L-1 50
ng L-1 200
ng L-1 50
ng L-1 200
ng L-1
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
4 1.58 2.06 1.8 0.3 0.31 0.40 0.36 0.06 0.47 0.61 0.5 0.1 0.9983 0.9999 0.9999 113 87 111 107 104 97 4 6 6 4
5 3.56 3.66 3.6 0.07 1.01 1.04 1.03 0.02 1.02 1.05 1.04 0.02 0.9983 0.9999 0.9999 108 87 113 108 96 99 3 7 3 3
6 5.41 5.19 5.3 0.2 0.74 0.71 0.73 0.02 1.48 1.42 1.45 0.04 0.9983 1.0000 0.9999 114 86 109 105 98 99 3 8 5 4
7 6.22 6.45 6.3 0.2 1.54 1.59 1.57 0.04 1.66 1.72 1.69 0.04 0.9980 0.9999 0.9999 108 85 109 105 97 99 3 10 3 3
8 6.66 7.42 7.0 0.5 2.51 2.79 2.7 0.2 1.37 1.53 1.5 0.1 0.9974 0.9996 0.9999 100 84 115 105 100 97 5 9 3 3
9 8.70 9.46 9.1 0.5 4.79 5.21 5.0 0.3 14.6 15.8 15.2 0.9 0.9943 0.9983 0.9845 102 85 111 107 97 101 3 8 3 4
10 11.99 10.59 11 1 7.50 6.62 7.1 0.6 1.20 1.06 1.1 0.1 0.9881 0.9953 0.9999 104 86 114 109 91 98 3 7 4 2
11 9.37 9.02 9.2 0.2 5.78 5.68 5.73 0.07 0.81 0.78 0.80 0.02 0.9823 0.9930 0.9999 104 83 113 107 91 104 4 9 2 2
12 6.20 6.29 6.3 0.06 3.36 3.40 3.38 0.03 0.43 0.44 0.44 0.01 0.9785 0.9933 0.9999 107 82 116 106 80 99 7 12 5 6
MD
´(-(
CH
2)3-
(EO
) n-O
H)M
4 0.29 0.60 0.4 0.2 0.46 1.38 0.9 0.7 0.26 0.54 0.4 0.2 0.9998 0.9996 0.9999 106 83 101 103 91 98 4 6 1 3
5 0.81 1.06 0.9 0.2 0.83 1.57 1.2 0.5 0.86 1.12 1.0 0.2 0.9999 0.9999 0.9999 101 83 101 104 89 99 5 6 2 4
6 1.61 1.64 1.6 0.02 0.99 1.44 1.2 0.3 1.30 1.32 1.31 0.01 0.9998 0.9999 0.9999 96 83 108 105 94 100 4 8 1 3
7 3.59 3.58 3.7 0.1 1.01 1.54 1.3 0.4 1.50 1.58 1.54 0.06 0.9993 0.9999 0.9999 94 84 110 105 91 101 3 9 3 3
8 5.18 5.14 5.2 0.03 1.55 2.23 1.9 0.5 1.59 1.57 1.58 0.01 0.9988 0.9999 0.9999 93 83 110 106 83 100 4 9 2 4
9 11.04 9.45 10 1 1.50 1.86 1.7 0.3 1.81 1.55 1.7 0.2 0.9944 0.9999 0.9999 91 83 109 106 79 101 4 8 0 2
10 16.87 12.35 15 3 5.30 5.59 5.4 0.2 1.64 1.20 1.4 0.3 0.9842 0.9984 0.9999 90 82 109 106 81 103 4 9 2 2
11 14.68 11.12 13 3 3.53 2.86 3.7 0.2 1.07 0.81 0.9 0.2 0.9748 0.9984 0.9999 88 82 110 101 79 105 5 9 0.5 2
12 9.86 8.38 6 6 4.12 5.12 4.6 0.7 0.58 0.49 0.54 0.06 0.9662 0.9930 0.9999 89 83 105 105 77 103 4 11 1 3
Chapter 2
54
Conclusion
A method for the trace analysis of two trisiloxane surfactants in the aqueous environment has
been developed and validated. The combination of a clean-up and concentration step, HPLC
separation, and detection by tandem mass spectrometry ensures specificity and sensitivity to
the method, which makes it suitable for the analysis of complex environmental samples. The
quantification by comparison with a calibration based on a single homologue instead of a
mixture of homologues avoids a wrong estimation of the concentrations in the sample because
of different response factors of sample and standard. This individual quantification of
homologues was possible due to the characterization of the oligomer composition of a
reference solution with the theoretical Poisson distribution and with MS measurements. This
approach could be extended to the quantification of other ethoxylated surfactants, supposed to
follow a similar Poisson distribution.
No concentrations of trisiloxane surfactants higher than the LOQ were found in the Karlsruhe
tap water and the river Rhine water used for method development but no conclusion on the
environmental occurrence of these compounds could be drawn only on these first results. The
aim of the work was to make available a method for the trace analysis of trisiloxane
surfactants in environmental waters and further screening of different environmental waters is
ongoing. Positive detections are expected when going toward the sources of emission of these
compounds. In addition, as shown by a preliminary lab experiment (data not shown) the
degradation process of trisiloxane surfactants by hydrolysis may lead to the formation of more
polar degradation products which may be found in environmental waters.
Chapter 2
55
AcknowledgmentThis work was financially supported by the German Association of Gas and Waterworks
(Deutscher Verein des Gas- und Wasserfaches e.V. DVGW), project W2/01/10. We thank
Markus Godejohann and Li-Hong Tseng from Brucker BioSpin, Rheinstetten, for performing
the LC-SPE-NMR/MS measurements.
References
Ananthapadmanabhan, K.P., Goddard, E.D., and Chandar, P., 1990. A study of the solution, interfacial and wetting properties of silicone surfactants. Colloids Surf. 44, 281.
Bonnington, L.S., 2000. Analysis of organosilicone surfactants and their degradation products, PhD Thesis, University of Waikato, Hamilton, New Zealand.
Bonnington, L.S., Henderson, W., and Zabkiewicz, J.A., 2004. Characterization of synthetic and commercial trisiloxane surfactants materials. Appl. Organomet. Chem. 18, 28.
Chandra, G., 1997. Organosilicon Materials, Springer-Verlag, Berlin Heidelberg New York.
Deutsches Institut für Normung e.V., 1994. DIN 32645 Chemical analysis – Decision limit, detection limit and determination limit; estimation in case of repeatability; terms, methods, evaluation, Beuth, Berlin.
Dorn, S.B., and Frame, E.M.S., 1994. Development of a high-performance liquid chromatographic-inductively coupled plasma method for speciation and quantification of silicones: from silanols to polysiloxanes. Analyst 119, 1687.
de Voogt, P., and Knepper, T.P., 2003. Quantification and quality assurance in surfactant analysis, pp. 443, in: Comprehensive Analytical Chemistry, Analysis and Fate of Surfactants and the Aquatic Environment, Knepper, T.P. (Ed.), Elsevier, Amsterdam.
Environment Agency, 2009. Environmental Risk Assessment Report: Octamethylcyclotetrasiloxane, http://publications.environment-agency.gov.uk/PDF/SCHO0309BPQZ-E-E.pdf (accessed Nov. 2011).
Environment Canada 2008. Screening Assessment for the Challenge Octamethylcyclotetrasiloxane (D4), http://www.ec.gc.ca/substances/ese/eng/challenge/batch2/batch2_556-67-2.cfm (accessed Nov. 2011).
Flory, P.J., 1940. Molecular size distribution in ethylene oxide polymers. J. Am. Chem. Soc. 62, 1561.
Floyd, D.T., 1999. Silicone Surfactants: applications in the personal care industry, pp. 181, in: Silicone surfactants, Surfactants science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker Inc, New York.
Chapter 2
56
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze, P., 1997. Mobility of a siliconepolyether studied by adsorption and desorption isotherms and soil thin-layer chromatography. Sci. Total Environ. 201, 89.
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze, P., 1998. Mobility of a siliconepolyether studies by leaching experiments through disturbed and undisturbed soil columns. Sci. Total Environ. 221, 159.
Hill, R.M., 1998. Superspreading. Curr. Opin. Colloid Interface Sci. 3, 247.
Hill, R.M., 1999. Silicone surfactants, Surfactants science series, Vol. 86, Marcel Dekker Inc, New York.
Kanner, B., Reid, W.G., and Petersen, I.H., 1967. Synthesis and properties of siloxane-polyether copolymer surfactants. Ind. Eng. Chem. Prod. R.D. 6, 88.
Kierkegaard, A., and McLachlan, M.S., 2010.Determination of decamethylcyclopentasiloxane in air using commercial solid phase extraction cartridges. J. Chromatogr., A. 1217, 3557.
McLachlan, M.S., Kierkegaard, A., Hansen, K.M., van Egmond, R., Christensen, J.H., and Skjoth, C.A., 2010. Concentration and fate of decamethylcyclopentasiloxane (D5) in the atmosphere. Environ. Sci. Technol. 44, 5365.
Nordic Council of Ministers 2005. Siloxanes in the Nordic Environment, http://www.norden.org/en/publications/publikationer/2005-593 (accessed Nov. 2011).
Penner, D., Burow, R., and Roggenbuck, F.C., 1999. Use of organosilicone surfactants as agricultural adjuvants, pp. 241, in: Silicone surfactants, Surfactants science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker Inc, New York.
Petrovic, M., and Barcelo, D., 2001. Analysis of ethoxylated nonionic surfactants and their metabolites by liquid chromatography/atmospheric pressure ionization mass spectrometry. J. Mass Spectrom. 36, 1173.
Radulovic, J., Sefiane, K., and Shanahan, M.E.R., 2009. Superspreading and wetting behavior of trisiloxanes. J. Bionic Eng. 6, 341.
Smith, A.L., and Parker, R.D., Trace analysis involving silicones, pp. 71, in: The analytical chemistry of silicones, Chemical analysis: A Series of Monographs on Analytical Chemistry and Its Applications, Vol 112, Smith, A.L. (Ed.), John Wiley & Sons Inc, 1991.
Union Carbide Corportation (Editor), 1967. Patent n°3 299 112, D.L. Bailey, in a.c.o.N.Y..
Ventura, F., and de Voogt, P., Separation and detection, pp. 51, in: Comprehensive Analytical Chemistry, Analysis and Fate of Surfactants and the Aquatic Environment, Knepper, T.P. (Ed.), Elsevier, Amsterdam, 2003.
Venzmer, J., 2011. Superspreading - 20 years of physicochemical research. Curr. Opin. Colloid Interface Sci. 16, 335.
Vick, S.C., 1984. Structure property relationships for silicone polyalkyleneoxide copolymers and their effects on performance in cosmetics. Soap Cosmet. Chem. Spe. 60, 36.
Chapter 2
57
Yanagida, S., Takahashi, K., and Okahara, M., 1977. Metal-ion complexation of noncyclic poly(oxyethylene) derivatives, I. Solvent extraction of alkali and alkaline earth metal thiocyanates. Bull. Chem. Soc. Jpn. 50, 1386.
Yanagida, S., Takahashi, K., and Okahara, M., 1978. Metal-ion complexation of noncyclic poly(oxyethylene) derivatives. II. PMR studies of the complexation with alkali and alkaline-earth metal cations. Bull. Chem. Soc. Jpn. 51, 1294.
Zhu, S., Miller, W.G., Scriven, L.E., and Davis, H.T., 1994. Superspreading of water – silicone surfactant on hydrophobic surfaces. Colloids Surf., A. 90, 63.
Chapter 2
58
Appendix A: Supplementary data
A.1 Theoretical approach: Poisson distribution
Flory described in 1940 the MWD of ethylene oxide polymers obtained from the reaction of
ethylene oxide with ethylene glycol and by successive additions of ethylene oxide (Flory,
1940):
OH CH2CH2 OH
O
OH CH2CH2 O H
O O
OH CH2CH2 O H
2 3
The numbers of molecules of the different sizes must be represented by a Poisson distribution
so that:
ν1n
n e1)!(n
νN
with n [1; ]
Nn is the number of molecules possessing n ethylene oxide groups and ν is the number of
propagating units reacted per initiator. According to Flory, ν is the average number of
ethylene oxide groups minus one. From this relation, the ratio of Nn-1 to Nn can be expressed
as:
ν1n
N
N
n
1n
(S1)
On the other hand, Nn is proportional to the mass concentration, βn so that the ratio n
1n
N
N
can also be written as a function of the mass concentrations:
n1n
1nn
n
1n
βM
βM
N
N
(S2)
βn is the mass concentration of molecules with n ethylene oxide groups and Mn is the
molecular weight of homologues with n ethylene oxide groups.
Combining equations (S1) and (S2) allows expressing βn as:
1)!(nM
Mνββ
1
n1n
1n
(S3)
Since the total amount of trisiloxane surfactant used to prepare the stock solution is known,
the total mass concentration is known:
mg/L1ββ1n
ntot
(S4)
etc.
Chapter 2
59
The combination of equations (S3) and (S4) allows calculating the theoretical concentrations
of all homologues in the stock solution.
A.2 Experimental approach: LC-SPE-NMR/MS
A.2.1 Materials and method
The chromatographic separation was carried out on a PolymerX RP-1 (250 mm × 4.6 mm i.d.,
5 µm, 100 Å) from Phenomenex connected to a model 1100 HPLC system (Agilent
Technologies, Waldbronn, Germany). 50 µL of MD´(-(CH2)3-(EO)n-OCH3)M at 49.3 mg/mL
in water/MeOH 1/1 were injected on the column. The mobile phase consisted of solvent A:
20 mM aqueous ammonium acetate and solvent B: ACN/MeOH/THF 80/10/10 with 10 mM
ammonium acetate. The flow rate was 0.9 mL/min and the column was kept at ambient
temperature. The elution was made under isocratic conditions (60 % solvent B during 31 min)
followed by an increase of solvent B to 70 % at 43 min. These conditions were kept for 6 min
before increasing B to 90 % from 50 to 55 min. The return to initial conditions was made
within 1 min. Compounds eluted from the column were diluted with water supplied by a
makeup pump (flow rate 1.5 mL/min) and were trapped on Hypersphere GP SPE cartridges,
10 × 2 mm (Spark, Emmen, Holland) which were then dried with nitrogen and eluted into a
3 mm NMR tube with around 180 µL of CD3CN containing 0.27 mg/mL of
trimethoxybenzene (TMB). NMR data were acquired on a Bruker Avance III 600 MHz
spectrometer equipped with a 5 mm QNP cryo probe. For each measurement, 128 scans were
accumulated into 32 k data points with a sweep width of 12 kHz using a 1D version of the
NOESY pulse sequence for double solvent suppression.
A.2.2 Results and discussion
To confirm the MWD calculated with a Poisson distribution and measured by MS, LC-SPE-
NMR/MS was applied to MD´(-(CH2)3-(EO)n-OCH3)M. 1H-NMR spectra were acquired for
each detectable homologue (2 < n < 10) after separation by liquid chromatography and
concentration by solid phase extraction. An example of 1H-NMR spectrum for homologue
n = 6 with peaks attribution is presented in Figure S 1.
Chapter 2
60
Figure S 1: 1H-NMR spectra between 0 ppm and 4 ppm for homologue n=6, with peaks attribution and
integration.
Trimethylsilanol usually taken as a reference was not suitable here because of the overlap
with the signals of the target analytes, it was thus replaced by TMB. Si(CH3)3 and Si-CH3
were chosen for quantification because of the possibility to assign them with certainty and
because of the high intensity of these signals: they represent 18 and 3 protons, respectively.
After integration of the peaks, the quantification was performed by comparison with TMB.
NMR is an absolute technique since the intensity of a signal does not depend on the analyte
but is only proportional to the number of protons of the molecule and its molar
concentration. The area of a peak (A) can therefore be expressed as a function of the mass
concentration (β), the molecular weight (M) and the number of protons (n´):
n´M
βCA (S5)
with C a common constant for all analytes. equation S5 can be applied to the standard
(denoted with a subscript s) and to homologue n (denoted with a subscript n) to calculate
the mass concentration of every homologue:
nn´nMnA
sMsAsn´sβ
nβ (S6)
The mass concentrations of individual homologues in the sample were calculated with
equation (S6) and are represented in Figure S 2 together with the MWD determined by MS
OCH3O
CH3
OCH3
CH3
Si
O
Si
O
Si
CH3
CH2CH2CH2EO
CH3CH3
CH3CH3
CH3OCH36
OCH3O
CH3
OCH3
CH3
Si
O
Si
O
Si
CH3
CH2CH2CH2EO
CH3CH3
CH3CH3
CH3OCH36
Chapter 2
61
measurements and calculated with the Poisson distribution. A different scale was used for
the NMR data because they do not correspond to the same sample as the two other
distributions. Even if the concentrations do not refer to the same sample, one can notice the
good correlation between the shapes of the three distributions, which supports the validity
of the two assumptions set in 3.1.2. However, the concentrations determined by NMR were
much lower than the concentrations expected from the volume injected on the HPLC
column and the concentration of the sample used. One possible explanation would be the
low recovery of MD´(-(CH2)3-(EO)n-OCH3)M by SPE. Because of this limitation, the
distribution determined by NMR was just given here as a confirmation but was not used for
quantification.
Figure S 2: Comparison of the oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M determined by
theoretical Poisson distribution ( ), MS measurements ( ) and LC-SPE-NMR/MS ( ).
Number of ethylene oxide groups
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Con
cent
rati
on i
n th
e st
ock
solu
tion
in
µg/
L
0
20
40
60
80
100
120
140
160
Con
cent
rati
on c
alcu
late
d fo
r N
MR
exp
erim
ents
in
mg/
L
0
100
200
300
Chapter 2
62
References
Flory, P.J., 1940. J. Am. Chem. Soc. 62, 1561.
Chapter 3
63
Chapter 3
Homologue specific analysis of a polyether
trisiloxane surfactant in German surface waters
and study on its hydrolysis
Published in Environmental Pollution, 186 (2014) 126-135
Chapter 3
64
TABLE OF CONTENTS
Abstract .................................................................................................................................... 65
1. Introduction ...................................................................................................................... 66
2. Experimental .................................................................................................................... 68
2.1 Chemicals and standards .......................................................................................... 68
2.2 Sample preparation and analysis by liquid chromatography-mass spectrometry .... 68
2.3 Sampling method ...................................................................................................... 69
2.4 Quality control .......................................................................................................... 70
2.4.1 Calibration blanks .............................................................................................. 70
2.4.2 Laboratory blanks ............................................................................................... 70
2.4.3 Calibration .......................................................................................................... 70
2.4.4 Limits of quantification ...................................................................................... 70
2.4.5 Isobaric interference ........................................................................................... 70
2.4.6 Matrix effects ..................................................................................................... 71
2.5 Degradation by hydrolysis ....................................................................................... 71
3. Results and discussion ...................................................................................................... 72
3.1 Environmental samples ............................................................................................ 72
3.2 Degradation by hydrolysis ....................................................................................... 79
3.2.1 Initial recoveries ................................................................................................. 79
3.2.2 Degradation kinetics ........................................................................................... 79
Conclusion ................................................................................................................................ 86
Acknowledgments .................................................................................................................... 87
References ................................................................................................................................ 87
Appendix A: Supplementary data ............................................................................................ 90
A.1 Discussion on the choice of containers for hydrolysis study as a function of pH ......... 90
A.2 Description of the identification of degradation products by HPLC-ESI-TOF ............ 95
Chapter 3
65
Abstract
The occurrence of a polyether trisiloxane surfactant in the ng L-1 range in German surface
waters is reported for the first time. The studied surfactant does not ubiquitously occur in the
aquatic environment in measurable concentrations but can reach surface waters on a local
scale. As a first step towards the understanding of the environmental fate, the hydrolysis was
studied according to the OECD guideline 111. It confirmed that the trisiloxane surfactant is
sensitive to hydrolysis and that the hydrolysis rate strongly depends on the pH and the
temperature. If one takes only into account the hydrolysis, the trisiloxane surfactant could
persist several weeks in river water (the half-life in water is approximately 50 days at pH 7,
25°C, and an initial concentration of 2 mg L-1). A degradation product, more polar than the
initial trisiloxane surfactant, was tentatively identified by high resolution mass spectrometry.
Chapter 3
66
1. Introduction
With 236, 000 tons produced in North America in 2009 (Oxford Economics for the Silicones
Environmental, Health and Safety Council of North America, 2010), silicones are an
important category of chemicals, represented in every segment of our everyday life. Food,
drugs, construction materials or personal care products contain silicones (Andriot et al.,
2007). This widespread use raises the question of their environmental impact. The assessment
of the environmental impact of silicones is a substantial task because of the variety of
compounds. The name “silicone” refers to all chemical substances in which silicon atoms are
linked via oxygen atoms, each silicon bearing one or several organic groups (Moretto et al.,
2000). Three categories of silicones have been suggested to have noteworthy environmental
loading (Figure 1): volatile methylsiloxanes (VMS), polydimethylsiloxanes (PDMS), and
polyethermethylsiloxanes (PEMS) (Allen et al., 1997).
PDMS
R = H, CH3, C(O)CH3
PEMS
VMS
Figure 1: Chemical structures of the silicones considered to have noteworthy environmental loading.
The two former categories, VMS and PDMS, have been studied extensively (Graiver et al.,
2003; Wang et al., 2013), but the knowledge on the environmental occurrence and fate of
PEMS is lacking (Powell and Carpenter, 1997). The aim of this paper is to fill this lack with a
study of the occurrence and fate of PEMS in the aquatic environment. In particular, we target
one trisiloxane surfactant (CAS: 27306-78-1) (x = 1, y = 0, 4 ≤ n ≤ 12, m = 0, R = CH3),
denoted MD´(-(CH2)3-(EO)n-OCH3)M. M stands for (CH3)3-Si-O1/2-, D’ for O1/2-Si(CH3)(R´)-
O1/2-, with R´ = -(CH2)3-(EO)n-OR and EO = OCH2CH2. Trisiloxane surfactants are
increasingly used in different sectors and recent studies raised concern about their effects on
non-target organisms: trisiloxane surfactants have been found to be toxic to mites (Cowles et
xx 5
x y
n m
OR
xx < 5
x
x < 7
Chapter 3
67
al., 2000) and to disturb the olfactory learning of honey bees (Ciarlo et al., 2012), one of the
multiple potential factors considered with regards to honey bees colony loss.
MD´(-(CH2)3-(EO)n-OCH3)M can enter the environment through various routes. It is used as
an adjuvant for pesticides because it improves the activity and the rainfastness of pesticides
(Venzmer, 2011). When spread on fields, MD´(-(CH2)3-(EO)n-OCH3)M may reach the soil
compartment and could reach natural waters by leaching or runoff. Another possible route for
MD´(-(CH2)3-(EO)n-OCH3)M to enter the environment is via the disposal of personal care
products containing MD´(-(CH2)3-(EO)n-OCH3)M (Floyd, 1999). After use, they are
discharged to wastewaters and MD´(-(CH2)3-(EO)n-OCH3)M reaches wastewater treatment
plants (WWTPs). If the elimination in WWTPs is not complete, MD´(-(CH2)3-(EO)n-OCH3)M
could reach natural waters via WWTP effluents. Also other entry routes to the aquatic
environment can be hypothesized: accidental contamination or improper disposal of
wastewaters coming from the washing of pesticide tanks. Once in the environment, the
partitioning of MD´(-(CH2)3-(EO)n-OCH3)M depends on its physico-chemical properties.
Hydrophilicity is frequently associated with mobility and preferential partition to the water
phase in the environment (Food and Agriculture Organization of the United Nations, 1989).
With a short hydrophobic part (three silicons) and a highly hydrophilic organic moiety (no
polyoxypropylene groups), MD´(-(CH2)3-(EO)n-OCH3)M is one of the most hydrophilic
PEMS. Based on the expected entry routes to the environment and its hydrophilicity,
MD´(-(CH2)3-(EO)n-OCH3)M is expected to reach the aquatic environment.
In the aquatic environment, hydrolysis is thought to be an important elimination mechanism.
This hypothesis is based on several studies showing that trisiloxane surfactants are sensitive to
hydrolysis. However, most of those studies have been focused on the stability in pesticide
formulations, under acidic conditions, and at high concentrations, from 0.1 g L-1 to 1 g L-1
(Knoche et al., 1991; Radulovic et al., 2009; Radulovic et al., 2010). Moreover, the hydrolysis
has been investigated by measuring the evolution of surface tension over time, but since
nothing is known on the surface activity of the by-products and degradation products, those
measurements are not specific and could have led to a wrong interpretation of the results.
Regarding the toxicity of trisiloxane surfactants to the aquatic environment, data are scarce
and only the acute toxicity has been investigated.: The No Observed Effect Concentration
(NOEC) has been found at 0.56 mg L-1 for Zebra Fish (96 h exposure time) and at 3.2 mg L-1
for Rainbow Trout (96 h exposure time); regarding the toxicity to algae, the NOEC to
Selenastrum capricornutum was determined at 1 mg L-1 (96 h exposure); and the NOECs were
10 mg L-1 and 25 mg L-1 for Daphnia similis and Daphnia magna, respectively (both 48 h
Chapter 3
68
exposure).Generally speaking, the knowledge on the environmental occurrence, fate and
toxicity of trisiloxane surfactants is scarce. The aim of this paper is to provide the first
measurements of a trisiloxane surfactant in surface waters, to investigate the importance of
hydrolysis as an elimination mechanism from the aquatic environment and to provide a
tentative identification of degradation products by high resolution mass spectrometry.
2. Experimental
2.1 Chemicals and standards
Reference substance of the trisiloxane surfactant was purchased from ABCR (Karlsruhe,
Germany). Stock solutions were prepared at 11 g L-1, 1 g L-1, 10 mg L-1, 1 mg L-1, and
0.1 mg L-1 (total surfactant concentration) in methanol and were stored at -18 °C. The
alkylphenol ethoxylate (Triton™ X-100 laboratory grade, average molecular weight:
625 g mol-1) was obtained from Sigma-Aldrich (Steinheim, Germany). All organic solvents
used for the analysis were HPLC-grade. Their purity was at least 99.8 %. Methanol and
dichloromethane were provided by Carl Roth GmbH (Karlsruhe, Germany), tetrahydrofurane
and isopropanol were purchased from Merck (Darmstadt, Germany) and acetonitrile was
obtained from Scharlau S.L. (Sentmenat, Spain). Ultrapure water was produced by an Arium
611 UV laboratory water purification system from Sartorius (Göttingen, Germany).
Ammonium acetate was purchased from Fluka (Steinheim, Germany, purity > 98.0 %),
sodium hydroxide, potassium chloride, potassium dihydrogene phosphate, boric acid, and
calcium chloride were obtained from Merck (Darmstadt, Germany) and had a purity of
minimum 99 %. All glassware was treated in a pyrolysis oven (model LHT6 / 120 from
Carbolite (Hope Valley, United Kingdom) at 550 °C for 60 min before use.
2.2 Sample preparation and analysis by liquid chromatography-mass
spectrometry
The sample enrichment and the analysis by HPLC-ESI-MS/MS were performed according to
Michel et al. (2012). Briefly, the sample was extracted by liquid-liquid extraction with
dichloromethane, the extract was blown down to dryness and reconstituted with 300 µL of a
mixture of 80 % ACN/10 % MeOH/10 % THF 10 mM ammonium acetate and 200 µL
aqueous 10 mM ammonium acetate solution. The HPLC unit is a 1200 HPLC system (Agilent
Technologies, Waldbronn, Germany) equipped with a solvent cabinet, a micro vacuum
degasser, a binary pump, a high-performance autosampler with 54 vial plates and a
temperature controlled column compartment. The separation was carried out on a PolymerX
Chapter 3
69
RP-1 (250 mm × 4.6 mm i.d., 5 µm, 100 Å) from Phenomenex with a flow rate of
800 µL min-1. The eluents were: A: a quaternary mixture including 40 % of aqueous 10 mM
ammonium acetate solution, 48 % ACN, 6 % THF, and 6 % MeOH and B: THF. The gradient
started with 100 % of eluent A over 8 min, decreased to 40 % over 16 min and was held for
3 min before going back to 100 %. The detection was achieved with a triple quadrupole mass
spectrometer API 4000 Q-Trap (Applied Biosystems/MDS Sciex Instruments, Concord,
Canada) equipped with an ESI source and used in multiple reaction monitoring (MRM) mode.
The mass spectrometer was operated in positive ionization mode and silicone surfactants were
detected as ammonium adducts. Data acquisition was performed with the Analyst software
(version 1.5.1). An extracted ion chromatogram was generated for every homologue and the
area obtained by integration of the peak was used for quantification.
2.3 Sampling method
The sample collection procedure followed the DIN standard 38402-15 (Deutsches Institut für
Normung, 2010). Due to their structures, surfactant molecules tend to accumulate at boundary
layers: water/air or water/sediments (González-Mazo et al., 2003; Knepper et al., 2003; Van
der Linden, 1992). Special care was therefore taken not to sample the water/air surface and
not to change the suspended solids fraction around the sampling point, for instance by moving
the bottom sediments. Samples were taken approximately at 20 cm from the surface by
introducing the bottle upside down in the water and by inverting it at 20 cm depth, allowing
the water to flow in. The bottles were filled without headspace and were immediately stored
in an ice-box. Generally speaking, it was avoided to collect samples where the water quality
was not typical of the water body (close to the river side, in stagnant zones, or, generally, in
any zone where the water is non-homogeneous). To fulfill this requirement, the samples were
taken directly in the middle of the water bed when possible or at least 3 m away from the river
bank by using a 3 m long arm. After arrival in the lab, the samples were stored at 4 °C and
were analyzed within a week. The stability of MD´(-(CH2)3-(EO)n-OCH3)M during storage of
river water samples was tested by analyzing one Neckar sample (Esslingen (10), 16.07.12,
Table 1 and Figure 2) directly after arrival in the lab and after one week of storage. For all
homologues, the two determined concentrations were not significantly different
(4.1 ± 0.1 ng L-1 and 4.2 ± 0.2 ng L-1 for example for the homologue n = 7), indicating that no
significant degradation occurs in this time scale under the applied storage conditions.
Chapter 3
70
2.4 Quality control
2.4.1 Calibration blanks
Every sequence of environmental samples began with two direct injections of deionized water
to check the absence of contamination in the solvents or in the HPLC-MS/MS system. A
blank sample (direct injection of deionized-water) was also always run after the calibration to
verify that no memory effect occurred. The absence of peak in any of the blanks confirms the
absence of contamination and memory effect in the HPLC-MS/MS system.
2.4.2 Laboratory blanks
For every set of environmental samples, two blanks samples of deionized water were prepared
simultaneously with river water samples and allowed verifying the absence of contamination
during sample preparation.
2.4.3 Calibration
The quantification of MD´(-(CH2)3-(EO)n-OCH3)M was performed by comparison with a 10-
points calibration in the range 2 ng L-1 - 250 ng L-1 (total surfactant concentration). The
calibration was performed over the whole sample preparation procedure so that the liquid-
liquid extraction recovery was taken into account.
2.4.4 Limits of quantification
The LOQs were defined according to the German standard DIN 32645 (Deutsches Institut für
Normung, 1994), based on the residual standard deviation of the calibration data.
2.4.5 Isobaric interference
The possibility of isobaric interference with an ethoxylated surfactant, Triton® X-100 was
tested. Certain homologues of Triton® X-100 (octylphenol ethoxylate with a number of
ethylene oxide groups between 6 and 14) have nearly the same molecular masses as the
targeted homologues of MD´(-(CH2)3-(EO)n-OCH3)M. Moreover, the structures of the two
ethoxylated surfactants present some similarities so that a similar retention time and
fragmentation pattern by MS/MS could be expected. A solution of Triton® X-100 at 1 mg L-1
was analyzed with the HPLC-MS/MS method described previously (obtained chromatogram
in supplementary material). A few peaks were detected with the transitions of
MD´(-(CH2)3-(EO)n-OCH3)M, but the retention times were several minutes different and not
all transitions recorded for MD´(-(CH2)3-(EO)n-OCH3)M were observable for Triton® X-100.
Chapter 3
71
The retention time and the presence of two transitions with a correct ratio ensure selectivity of
the analysis. The best proof of the identity of MD´(-(CH2)3-(EO)n-OCH3)M in the river water
samples would be measurements with high resolution mass spectrometry, for instance time-
of-flight mass spectrometry (TOF-MS), and confirmation by the exact mass and the isotopic
pattern. However, the lower sensitivity of the TOF compared to triple quadrupole mass
spectrometer did not allow for this confirmation in environmental samples.
2.4.6 Matrix effects
Since internal standards are currently not available for trisiloxane surfactants, the possibility
of signal suppression or enhancement due to matrix effects was investigated. The standard
addition method was applied to a Neckar sample (Esslingen (10), 16.07.12, Table 1 and
Figure 2). The sample was fortified with different volumes of a 0.1 mg L-1 stock solution of
MD´(-(CH2)3-(EO)n-OCH3)M to reach five different total surfactant concentrations
(0 ng L-1, 20 ng L-1, 40 ng L-1, 60 ng L-1, and 80 ng L-1).
2.5 Degradation by hydrolysis
The degradation by hydrolysis was carried out according to the OECD guideline 111
(Organisation for Economic Cooperation and Development, 2004). Two buffer solutions (pH
7 and pH 9) were prepared from sodium hydroxide, boric acid, potassium chloride, and
potassium phosphate in ultrapure water. Their pH values were adjusted at 7.0 ± 0.1 and 9.0 ±
0.1 and 100 mL of the buffered solutions were placed in polypropylene (PP) bottles (VWR,
Bruchsal, Germany): 8 bottles for pH 7 and 8 bottles for pH 9. The solutions were autoclaved
(Laboklav 55-195 from Steriltechnik AG, Detzel Schloss/Satuelle, Germany) at 121°C and
1.5 bars for 20 min. At the beginning of the experiment, 18 µL of a stock solution of
MD´(-(CH2)3-(EO)n-OCH3)M at 11 g L-1 in MeOH was spiked to reach a starting
concentration of 2 mg L-1 (total surfactant concentration) in all bottles except four blanks.
After homogenization, the buffered solutions of the test substance were sampled to determine
the initial concentration of MD´(-(CH2)3-(EO)n-OCH3)M. The test vessels were then kept in
the dark at different temperatures (12°C, 25°C, and 50°C) and the test was allowed to proceed
for 30 days or until 99 % of hydrolysis, whichever applied first. Samples were regularly taken
(every day during one week and then once a week) and analyzed to determine the
concentration of MD´(-(CH2)3-(EO)n-OCH3)M. To ensure a reproducible volume during
sampling, all sampled solutions were allowed to reach room temperature right before
sampling. To maintain sterility, the sampling was performed in a laminar flow cabinet (model
Chapter 3
72
Golden Line GL-130, Kojair, Vilppule, Finland). Aliquots (500 µL) were taken with a sterile
pipet tip, blown down to dryness under nitrogen, and reconstituted with 300 µL of 10 mM
ammonium acetate solution in 80 % ACN / 10 % THF / 10 % MeOH and 200 µL of aqueous
10 mM ammonium acetate. The HPLC-ESI-MS/MS analysis was performed according to
Michel et al. (2012). A new calibration was prepared from the stock solutions for every new
sequence of samples.
3. Results and discussion
3.1 Environmental samples
The measured concentrations of MD´(-(CH2)3-(EO)n-OCH3)M are summarized in Table 1.
The individual oligomeric concentrations range from below the LOQ up to 96 ng L-1
(homologue n = 7, Neckar Deizisau, 27.02.12, Figure 2). To the best of our knowledge, this is
the first time that a trisiloxane surfactant is detected in surface waters. The screening of
several German rivers was initiated in 2012 and revealed the occurrence of the trisiloxane
surfactant in the Neckar River, for the first time in February 2012. This first result was
confirmed by additional samples from the same area (Deizisau in March 2012, Poppenweiler
in May 2012 and Besigheim in April 2012). The repeated detection of
MD´(-(CH2)3-(EO)n-OCH3)M in the Neckar River suggested the presence of one or several
point source. To obtain more specific information on the occurrence of the trisiloxane
surfactant in this area, an additional sampling, specifically dedicated to this zone, was carried
out in July 2012. Samples were taken between Rottenburg am Neckar and Besigheim, at
sampling sites routinely used by our local partner for the water quality surveillance (Figure 2).
The trisiloxane surfactant was again positively detected. The identification of the origin of the
trisiloxane surfactant in the Neckar River was not possible because of the diversity and
density of activities in this area. The land utilization in this part of the state of Baden-
Württemberg is dominated by agriculture (54 %) (Ministry of the Environment, Climate
Protection and the Energy Sector Baden-Würtemberg, 2009). MD´(-(CH2)3-(EO)n-OCH3)M is
known to have applications in agriculture and the runoff from agricultural fields could be a
source of trisiloxane surfactant in the Neckar River. On the other hand, this region is heavily
industrialized and densely populated (910 inhabitants per km² against 235 inhabitants per km²
on average in Germany) (Baden-Würtemberg Umweltministerium, 2009). Several WWTPs
treating domestic and industrial wastewaters are located along the Neckar River and their
effluents could be a source of trisiloxane surfactant. The identification of the point source(s)
Chapter 3
73
would require a precise knowledge on the origin of the influents from each WWTP and a
closer sampling campaign.
One of the first question rising from the positive detection of a chemical in river water is to
know if it represents a threat to the aquatic environment. The toxicity of the trisiloxane
surfactant has been studied on several aquatic organisms. The No Observed Effect
Concentration (NOEC) has been found at 0.56 mg L-1 for Zebra Fish (96 h exposure time) and
at 3.2 mg L-1 for Rainbow Trout (96 h exposure time); regarding the toxicity to algae, the
NOEC to Selenastrum capricornutum was determined at 1 mg L-1 (96 h exposure); and the
NOECs were 10 mg L-1 and 25 mg L-1 for Daphnia similis and Daphnia magna, respectively
(both 48 h exposure).
If one refers to the acute toxicity data existing for MD´(-(CH2)3-(EO)n-OCH3)M, the
concentrations of trisiloxane surfactant measured in the Neckar should not be a threat for the
aquatic species. However, more specific information would be needed on the chronic toxicity
of MD´(-(CH2)3-(EO)n-OCH3)M and on its bioaccumulation factor.
Chapter 3
74
Table 1: Concentrations of individual homologues of MD (́-(CH2)3-(EO)n-OCH3)M in different environmental samples collected in 2012 and 2013.
Concentration in ng L-1 River
(sampling point number)
km Date GPS coordinates pH
value Conductivity
in µS/cm n = 4 n=5 n=6 n=7 n=8 n=9 n=10 n=11 n=12
LOQs (ng L-1) 0.4 1.0 0.7 1.6 2.7 5.0 7.1 5.7 3.4 Neckar Kiebingen (1) n.a. 16.07.12 N48°28.822’ E 8°58.288’ 8.01 696 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Tübingen (2) n.a. 16.07.12 N48°31.077’ E 8°04.431’ 8.18 703 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Pliezhausen (5) n.a. 16.07.12 N48°33.097’ E 9°12.276’ 8.14 724 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Wernau (8) n.a. 16.07.12 N48°42.086’ E 9°25.071’ 8.47 703 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Plochingen (9) n.a. 16.07.12 N48°42.781’ E 9°24.328’ 8.47 700 1 <LOQ 0.7 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Deizisau (6) 200 27.02.12 N48°43.248’ E 9°22.730’ n.a. n.a. 32 50 77 96 66 48 42 37 22 26.03.12 n.a. n.a. 8 10 12 9 6 <LOQ <LOQ <LOQ <LOQ 16.07.12 8.36 689 5 4 4 3 <LOQ <LOQ <LOQ <LOQ <LOQ 25.02.13 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Esslingen (10) n.a. 16.07.12 N48°43.106’ E 9°20.916’ 8.34 687 5 5 5 4 3 <LOQ <LOQ <LOQ <LOQ Poppenweiler (11)
165 21.05.12
N48°54.700’ E 9°15.054’ n.a. n.a. 2 4 5 4 4 <LOQ <LOQ <LOQ <LOQ
16.07.12 7.85 717 1 <LOQ 0.9 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Besigheim (12) 137 23.04.12 N49°00.052’ E 9°08.821’ n.a. n.a. 2 2 3 2 <LOQ <LOQ <LOQ <LOQ <LOQ 21.05.12 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 16.07.12 7.85 727 0.4 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 11.02.13 1 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Kochendorf 104 24.04.12 N49°12.940’ E 9°12.598’ n.a. n.a. 2 2 3 3 <LOQ <LOQ <LOQ <LOQ <LOQ 22.05.12 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 05.06.12 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 12.02.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Steinlach* Derendingen (4) n.a. 16.07.12 N48°29.052’ E 9°03.936’ 7.92 484 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Dußlingen (3) n.a. 16.07.12 N 48°27.260’ E 9°03.335’ 8.5 488 <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Fils* Plochingen (7) n.a. 16.07.12 N48°42.211’ E 9°25.550’ 8.65 620 14 10 8 5 <LOQ <LOQ <LOQ <LOQ <LOQ
Chapter 3
75
Koersch* Friedrichsmühle 022 20.03.12 n.a. n.a. n.a. 1 <LOQ 1 <LOQ <LOQ <LOQ <LOQ <LOQ <LOD 16.04.12 n.a. n.a. 2 2 2 2 <LOQ <LOQ <LOQ <LOQ <LOQ Kocher* Kochendorf 905 24.04.12 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Ruhr Dumberg n.a. 18.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 18.03.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Rhein Karlsruhe 359 04.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 21.02.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 04.03.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Worms 443 21.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Köln 685.8 12.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Düsseldorf 732.1 12.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Duisburg 757.9 19.03.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Dahme Berlin n.a. 31.07.12 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Müggelsee Berlin n.a. 31.07.12 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Donau Ulm 803 04.02.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 04.02.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ 04.03.13 n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ Elbe Riesa n.a. 13.03.13 n.a. n.a. n.a. <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ <LOQ
n.a. data not available for this sample *Neckar tributary. The concentrations of the homologues are the average of two replicates.
Chapter 3
76
Figure 2: Sampling sites on the Neckar River and two of its tributaries, between Rottenburg am Neckar
and Besigheim.
The standard addition method applied to sample 10 (Esslingen, 16.07.12, Table 1 and
Figure 2) showed a clear signal suppression in this sample. Figure 3 compares the individual
oligomeric concentrations obtained for this sample with an external calibration of the entire
method including sample extraction and with the standard addition method. The error bars
correspond to the standard deviation of two replicates.
Stuttgart
Reutlingen
Ludwigsburg
Esslingen am Neckar
Plochingen
Wernau
(Neckar)
Pliezhausen
Besigheim
Tübingen
Rottenburg
am Neckar
PlochingenDeizisau
Wernau
(Neckar)
Esslingen am
Neckar
Fils
Steinlach
Neckar
Poppenweiler
(1)
(2)
(3)
(5)
(6)
(7)(8)
(9)(10)
(11)
(12)
Chapter 3
77
Figure 3: Oligomeric concentrations of MD (́-(CH2)3-(EO)n-OCH3)M in the Neckar River at Esslingen
(sampled on 16.07.12) determined with external calibration of the entire method including sample
preparation and with standard addition.
The concentrations obtained by the standard addition method were between 28 % and 52 %
higher than those obtained with the external standard calibration. This could imply that the
concentrations measured in the Neckar River and given in Table 1 are underestimated and to
some extent the true concentrations are actually higher. As already stated in Michel et al.
(2012), without internal standards, the analytical method can only be considered as semi-
quantitative. Nevertheless, it gives first estimation on the environmental concentrations of
trisiloxane surfactants.
Examples of oligomeric distributions obtained in the river water samples collected in July
2012 are represented in Figure 4 and are compared to the oligomeric distribution in the
standard of the trisiloxane surfactant. For the graphical representation, all values with a signal
to noise ratio higher than 10 have been included.
Number of ethylene oxide groups
2 4 6 8 10 12 14
Con
cent
rati
on i
n ng
/L
0
2
4
6
8
10External calibrationStandard addition
Chapter 3
78
Figure 4: Oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M in the standard solution at 75 ng L-1 ( )
and in the river water samples collected on the 16.07.12: ( ) Neckar at Deizisau, ( ) Neckar at
Esslingen, ( ) Neckar at Poppenweiler, and ( ) Fils at Plochingen.
The oligomeric distribution in the standard of MD´(-(CH2)3-(EO)n-OCH3)M has been
determined by mathematical treatment of the Poisson distribution, full scan MS and NMR
(Michel et al., 2012) and was centered on 7.5 EO groups. In river water samples, the shape of
the oligomeric distribution was different. As confirmed by the standard addition method
(Figure 3), and by repeated measurements of the same sample, the maximum was no longer
on 7.5 EO units but on lower molecular weight homologues. A tentative explanation for the
observed shape of the oligomeric distribution is suggested by comparison with nonylphenol
ethoxylates (NPEO), another type of ethoxylated surfactants. It has been observed for NPEOs
that the sorption capacity increases with the number of EO groups (Lara-Martìn, et al., 2008).
If this is true for the trisiloxane surfactant, it means that the homologues with a long EO chain
would be more likely to be removed by adsorption to sludge, sediments or soil than the low
molecular weight homologues. Such a different adsorption could explain the change of
oligomeric distribution between the commercial product and the environmental samples.
Given the lack of data on the ecotoxicity, bioaccumulation factors, and persistence of the
different homologues, it is difficult to predict the environmental consequences of this change
of oligomeric distribution. For NPEO, it is known that the homologues with a short EO chain
are more toxic to aquatic organisms that the homologues with a long EO chain (Blasco, et al.,
2003). Nothing allows affirming that the same is true for trisiloxane surfactants.
Number of ethylene oxide groups
2 4 6 8 10 12 14
Con
cent
rati
on i
n ng
/L
0
2
4
6
8
10
12
14
16
18
Chapter 3
79
3.2 Degradation by hydrolysis
3.2.1 Initial recoveries
After first experiments showing a non-negligible adsorption of MD´(-(CH2)3-(EO)n-OCH3)M
on glass vials at 2 mg L-1 (total surfactant concentration), an alternative material was
investigated (detailed description in supplementary material Text S1, Table S1, and Figure
S1). Polypropylene bottles of 125 mL were found to be the best alternative. The initially
measured concentrations were very close to the expected concentrations (recoveries higher
than 80 %, except for the homologues n = 4 and n = 5 in buffer solution at pH 7).
3.2.2 Degradation kinetics
Hydrolysis is commonly described by a first-order kinetics and the following equations apply:
C = C0 e-kt (1)
or ln(C/C0)= -kt (2)
C is the remaining concentration of the test substance after a certain time (in mg L-1), C0 is the
initial concentration of the analyte, k is the hydrolysis rate constant (in d-1), t is the time (in d).
In Figure 5, ln(C/C0) is plotted as a function of time for all conditions applied in this study.
The curves are given for the homologues n = 6, n = 8, n = 10, and n = 12, as examples. The
curves obtained for the other homologues are given in supplementary material (Figure S2). At
pH 9 and 50 °C, more than 90 % of the initial test substance was already hydrolyzed after
1 day. The corresponding curves could therefore not be represented in Figure 5. The
hydrolysis rate was taken as the slope of the linear regression applied to the plots
ln(C/C0) = f(t) and the reaction half-time is calculated with: t1/2 = (ln 2)/k . The activation
energy (Ea in kJ mol-1) of the hydrolysis reaction was calculated based on the following
equation:
2T
1
1T
11kln 2kln
RaE
(3)
T is the absolute temperature, R is the gas constant (8.314 J mol-1 K-1). The hydrolysis rate
constant, the reaction half-life, the corresponding correlation coefficients and the activation
energy are summarized in Table 2.
The high correlation coefficients obtained at pH 9 confirm the good agreement with a first
order kinetics. At pH 7, 12°C and 25°C, the low correlation coefficients indicate that the
hydrolysis does not follow a first order kinetics law.
Chapter 3
80
Table 2: Hydrolysis rate constants, reaction half-life, correlation coefficients, and activation energy for the hydrolysis of MD (́-(CH2)3-(EO)n-OCH3)M under the
different conditions tested in this study; na: not applicable.
pH T in °C
T in K
k in d-1 t1/2 in d R² Ea in kJ mol-1
n=6 n=8 n=10 n=12 n=6 n=8 n=10 n=12 n=6 n=8 n=10 n=12 n=6 n=8 n=10 n = 12
pH 9 12 285 0.33±0.02 0.31±0.02 0.29±0.02 0.29±0.01 2 2 2 2 0.99 0.99 0.99 0.99
57 58 63 60 pH 9 25 298 0.9±0.2 0.9±0.2 0.9±0.2 0.9±0.1 0.7 0.8 0.7 0.8 0.94 0.94 0.94 0.96
pH 9 50 323 n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a.
pH 7 12 285 0.005±0.007 0.004±0.005 0.002±0.005 0.003±0.005 151 198 289 210 0.08 0.07 0.04 0.06
89 75 90 77 pH 7 25 298 0.02±0.01 0.014±0.008 0.013±0.007 0.014±0.0008 29 50 55 51 0.48 0.34 0.35 0.35
pH 7 50 323 0.9±0.1 0.46±0.03 0.33±0.01 0.31±0.01 1 1 2 2 0.96 0.99 0.99 0.99
River water
pH 8 25 298 0.087±0.004 0.081±0.001 0.081±0.001 0.083±0.003 8 9 9 8 1.00 1.00 1.00 1.00 n.a. n.a. n.a. n.a.
River water filtrated
pH 8 25 298 0.105±0.005 0.101±0.002 0.102±0.002 0.102±0.002 7 7 7 7 0.99 1.00 1.00 1.00 n.a. n.a. n.a. n.a.
Chapter 3
81
Figure 5: Graphical representation of ln(C/C0) as a function of time for the homologues n=6, n=8, n=10,
and n=12, for pH 7 and 9 and temperature tested in the study: ( ) 12°C, ( ) 25°C, and ( ) 50°C.
pH 7, n=6
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n=8
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n=10
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n=12
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 6
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 8
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 10
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 12
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
Chapter 3
82
The hydrolysis rate clearly depends on the temperature and on the pH. The hydrolysis is faster
at pH 9 than at pH 7 and is also much faster at 50°C than at 25°C and 12°C. This confirms
that the trisiloxane surfactant is unstable at high temperatures and under alkaline conditions
but is rather stable at neutral pH and low temperatures. Those results are in good agreement
with previous observations that the surface tension of an aqueous solution of
MD´(-(CH2)3-(EO)n-OCH3)M at 0.1 % (6 < pH < 8) does not increase during at least 40 days,
indicating that no significant degradation occurs during this time period (Knoche et al., 1991).
Figure 6: Comparison of the hydrolysis rates of MD (́-(CH2)3-(EO)8-OCH3)M in buffers at pH 7 and pH 9
and in sterilized river water (filtrated and non-filtrated)
To be closer to environmental conditions, the hydrolysis was studied in sterilized river water
(Neckar river, pH 8) and sterilized filtrated river water. The curves obtained in deionized
water and river water are compared in Figure 6. In the samples of river water (pH 8), the
hydrolysis was faster than at pH 7 and slower than at pH 9, confirming the important
influence of the pH value on the hydrolysis rate. No significant difference was observed
between the filtrated and non-filtrated samples. Generally, in German rivers, the pH is around
8 and the temperature lies in the range 0°C - 25°C. Based on the results obtained in this study,
if one takes only into account the hydrolysis as an elimination mechanism, the trisiloxane
surfactant could persist several weeks in river water.
Additional experiments were conducted to check the influence of the concentration of
MD´(-(CH2)3-(EO)n-OCH3)M on hydrolysis rate. The hydrolysis was followed at pH 7 and
Time in days
0 10 20 30
ln C
/C0
-3
-2
-1
0
1Pure water pH 7Pure water pH 9River water non filtratedRiver water filtrated
Chapter 3
83
pH 9 and at 12°C, 25°C and 50°C, at 1 g L-1 in glass containers. Although not strictly
performed under sterile conditions (no use of a laminar flow cabinet), this test showed that the
degradation was slower at 1 g L-1 than at 2 mg L-1. At 1 g L-1, the concentration of
MD´(-(CH2)3-(EO)n-OCH3)M is higher than the critical micellar concentration (CMC is
80 mg L-1 (Knoche et al., 1991)), which means that micelles and free molecules exist together
in the solution. In a micelle in aqueous solution, the hydrophilic moiety points toward the
outside of the aggregate while the hydrophobic part is in the center of the micelle, protected
from nucleophilic attacks. The fact the siloxane chain, sensitive to hydrolysis, is protected in
the center of the micelle could explain why the hydrolysis is slower when the concentration is
higher than the CMC (Lundberg and Stjerndahl, 2011). In the aquatic environment, the
concentrations of MD´(-(CH2)3-(EO)n-OCH3)M are very unlikely to be above the CMC and
this phenomenon should not be relevant at environmental concentrations.
Under certain conditions, the hydrolysis of silicones is reversible (Spivack et al., 1997) and
this feature could explain certain experimental observations. At 1 g L-1, pH 7 and 12 °C, the
measured concentrations of MD´(-(CH2)3-(EO)n-OCH3)M were not regularly decreasing as
expected but were very variable, decreasing and then increasing again with regular time
intervals. This phenomenon was also observed by other authors during the degradation of
another silicone-based surfactant at 1 g L-1 in an aqueous solution (Laubie et al., 2013). It
seems that at those high concentrations and under certain conditions of pH and temperature,
the hydrolysis of the siloxane chain is reversible and a combination of hydrolysis and
condensation takes place. However, one can assume that the condensation requires relatively
high concentrations and is therefore not relevant under environmental conditions where the
typical concentrations are in the order of a few ng L-1.
The oligomeric distribution of MD´(-(CH2)3-(EO)n-OCH3)M was followed along the duration
of the test and no significant changes were observed (Figure 7).
Chapter 3
84
Figure 7: Evolution of the oligomeric distribution of MD (́-(CH2)3-(EO)n-OCH3)M during the hydrolysis
test at pH 9, 12°C.
At pH 9, the hydrolysis rate constants are similar for all homologues. All homologues are
hydrolyzed simultaneously and the shape of the distribution is not much modified: the center
stays at 7.5 EO groups, as illustrated by Figure 7. At pH 7, the shape of the oligomeric
distribution seemed to be modified during the experiment. The center was at 7.5 EO in the
initial solution and at 9 EO groups after 15 days at pH 7 and 50°C. However, as described in
details in the supplementary material, this effect is most likely due to the adsorption of the
homologues with n = 4 and n = 5 on the PP bottles. The hydrolysis rate constants determined
for these two homologues are consequently not representative of the hydrolysis itself and
were dismissed from our dataset. Those observations indicate that the difference of oligomeric
distribution between the river water samples and the commercial trisiloxane surfactant does
not come from different hydrolysis rates between the homologues of
MD´(-(CH2)3-(EO)n-OCH3)M.
Number of ethylene oxide groups
2 4 6 8 10 12 14
Con
cent
rati
on i
n m
g/L
0.0
0.1
0.2
0.3
0.4Day 0Day 1Day 2Day 3Day 4Day 7Day 15Day 23Day 30
Chapter 3
85
Figure 8: Proposed hydrolysis reaction of MD (́-(CH2)3-(EO)n-OCH3)M and relative evolution of the
concentrations of the trisiloxane surfactant homologue MD (́-(CH2)3-(EO)4-OCH3)M ( ) and the identified
degradation product ( ) over time for the experiment at pH 9, 12°C.
One degradation product formed by hydrolysis could be identified (product (1) in Figure 8) by
high resolution HPLC-(ESI)-TOF measurements (experimental settings in supplementary
material) and confirms the results previously obtained by Bonnington (2000). Since no
analytical standard was available for the degradation product (1), only the qualitative increase
of concentration of the degradation product (1) was followed during the study of hydrolysis.
The decay of MD´(-(CH2)3-(EO)4-OCH3)M and the qualitative increase of degradation
product (1) during the experiment at pH 9 and 12°C are shown as example in Figure 8. The
degradation product (1) is more polar than the initial trisiloxane surfactant and is therefore
suspected to be more mobile in the environment. The identified degradation product is in
good agreement with the results obtained by Laubie et al. (2013) when studying the
degradation of a polydimethylsiloxane-graft-polyethylene oxide (a silicone surfactant
containing various number of siloxane units and several pendant polyethylene oxide groups)
Time in days
0 5 10 15 20 25 30
C/C
0 fo
r M
-D´(
EO
4-O
CH
3)M
in
%
0
20
40
60
80
100
Pea
k ar
ea /
Ini
tial
pea
k ar
ea f
or
degr
adat
ion
prod
uct
(1)
in %
0
1000
2000
3000
4000
5000
6000
2H2O 2
n n
(1) (2)
Chapter 3
86
by NMR. The degradation product (2), trimethylsilanol, is known to be eliminated by
adsorption and by oxidation by OH radicals to give finally silica. The fate of the degradation
product (1) has not been studied yet and no data exist on its toxicity.
Conclusion
The measurements reported in this study indicate that the investigated trisiloxane surfactant is
not ubiquitously in the aquatic environment in measurable concentrations but can reach
surface waters on a local scale. The measured concentrations, in the low ng L-1 range, should
not represent a threat for the aquatic environment. However, more specific data on the chronic
ecotoxicity of MD´(-(CH2)3-(EO)n-OCH3)M would be necessary to draw a conclusion on its
effects in the aquatic environment.
To provide information on the importance of hydrolysis as an elimination mechanism of the
trisiloxane surfactant from the aquatic environment, the hydrolysis was studied under relevant
conditions of pH and temperature. Several factors influencing the hydrolysis rate could be
highlighted: the pH, temperature, and concentration. The obtained results indicate that under
environmental conditions (0 - 25°C, pH close to neutrality), taking only into account the
hydrolysis, the trisiloxane surfactant could persist several weeks. To formulate a definitive
statement on its persistence in the aquatic environment, it would be necessary to study the
biodegradation additionally.
One degradation product of MD´(-(CH2)3-(EO)n-OCH3)M could be tentatively identified by
high resolution mass spectrometry. It was found to be more polar than the trisiloxane
surfactant and it is thus expected to be more mobile in the environment. Data on the
environmental occurrence, fate, and ecotoxicity of the identified degradation product are
missing. In order to fully characterize the environmental impact of trisiloxane surfactants,
more research would be necessary to characterize the environmental occurrence, fate and
toxicity of their degradation products.
Chapter 3
87
Acknowledgments
This work was financially supported by the German Association for Gas and Water, a
registered technical-scientific association (Deutscher Verein des Gas- und Wasserfaches e.V.
DVGW), project W 2-01-10. We thank Dr. Marco Scheurer for the TOF measurements and
for the review of the manuscript.
References
Allen, R.B., Kochs, P., and Chandra, G., 1997. Industrial organosilicon materials, their environmental entry and predicted fate, pp. 1, in: Organosilicon materials, Chandra, G. (Ed.), Springer, Berlin, Heidelberg.
Andriot, M., Chao, S.H., Colas, A., Cray, S., de Buyl, F., DeGroot, J.V., Dupont, A., Easton, T., Garaud J.L., Gerlach, E., Gubbels, F., Jungk, M., Leadley, S., Lecomte, J.P., Lenoble, B., Meeks, R., Mountney, A., Shearer, G., Stassen, S., Stevens, C., Thomas, X., and Wolf, A.T., 2007. Silicones in industrial applications, pp. 61, in: Inorganic Polymers, De Jaeger, R., Gleria, M. (Eds.), Nova Publishers, Hauppauge.
Blasco, J., Hampel, M., and Moreno-Garrido, I., 2003. Toxicity of surfactants, pp. 827, in: Analysis and fate of surfactants in the aquatic environment, Knepper, T.P., et al. (Eds.), Elsevier, Amsterdam.
Bonnington, L.S., 2000. Analysis of organosilicone surfactants and their degradation products. PhD thesis, University of Waikato, Hamilton, New Zealand.
Ciarlo, T.J., Mullin, C.A., Frazier, J.L., and Schmehl, D.R., 2012. Learning impairment in honey bees caused by agricultural spray adjuvants. Plos One 7, 1.
Cowles, R.S., Cowles, E.A., McDermott, A.M., and Ramoutar, D., 2000. “Inert” formulation ingredients with activity: toxicity of trisiloxane surfactant solutions to twospotted spider mites (Acari: Tetranychidae). J. Econ. Entomol. 93, 180.
Deutsches Institut für Normung e.V., 2010. DIN 38402-15, German standard methods for the examination of water, waste water and sludge – General information (group A) – Part 15: Sampling from running waters (A 15), Beuth, Berlin.
Deutsches Institut für Normung e.V., 1994. DIN 32645, Chemical analysis; decision limit; detection limit and determination limit; estimation in case of repeatability; terms, methods, evaluation, Beuth, Berlin.
Floyd, D.T., 1999. Silicone surfactants: applications in the personal care industry, pp. 181, in: Silicone Surfactants, Surfactants science series, Vol 86, Hill, R.M. (Ed.), Marcel Dekker, New York.
Food and Agriculture Organization of the United Nations, 1989. Revised guidelines on environmental criteria for the registration of pesticides. http://www.fao.org/fileadmin/templates/agphome/documents/Pests_Pesticides/Code/Old_guidelines/ENVICRI.pdf (Accessed October 2013)
Chapter 3
88
González-Mazo, E., Knepper, T.P., Ventura, F., and de Voogt, P., 2003. Sample handling, pp. 393, in: Analysis and fate of surfactants in the aquatic environment. Knepper, T.P., et al. (Eds.), Elsevier, Amsterdam.
Graiver, D., Farminer, K.W., and Narayan, R., 2003. A review of the fate and effects of silicones in the environment. J. Polym. Environ. 11, 129.
Knepper, T.P., Eichhorn, P., and Bonnington, L.S., 2003. Environmental processes, pp. 525, in: analysis and fate of surfactants in the aquatic environment, Knepper, T.P. (Ed.), Elsevier, Amsterdam.
Knoche, M., Tamura, H., and Bukovac, M.J., 1991. Performance and stability of the organosilicone surfactant L-77 - effect of pH, concentration, and temperature. J. Agric. Food Chem. 39, 202.
Lara-Martìn, P.A., Gómez-Parra, A., and González-Mazo, E., 2008. Reactivity and fate of synthetic surfactants in aquatic environments. Trends Anal. Chem. 27, 684.
Laubie, B., Bonnafous, E., Desjardin, V., Germain, P., and Fleury, E., 2013. Silicone-based surfactant degradation in aqueous media. Sci. Total Environ. 454-455, 199.
Lundberg, D., and Stjerndahl, M., 2011. Hydrolysis of surfactants containing ester bonds: modulation of reaction kinetics and important aspects of surfactant self-assembly. J. Chem. Educ. 88, 1274.
Michel, A., Brauch, H.-J., Worch, E., and Lange, F.T., 2012. Development of a liquid chromatography tandem mass spectrometry method for trace analysis of trisiloxane surfactants in the aqueous environment: An alternative strategy for quantification of ethoxylated surfactants. J. Chromatogr., A. 1245, 46.
Ministry of the Environment, Climate Protection and the Energy Sector Baden-Würtemberg, 2009. Bewirtschaftungsplan Bearbeitungsgebiet Neckar. http://www.um.baden-wuerttemberg.de/servlet/is/63580/Bewirtschaftungsplan_Neckar_26_11_09.pdf (Accessed October 2013)
Moretto, H.H., Schulze, M., and Gebhard, W., 2000. Silicones, in: Ullmann's Encyclopedia of Industrial Chemistry. Wiley-VCH, Weinheim.
Organisation for Economic Cooperation and Development (OECD) 2004. Guideline for the testing of chemicals, Test No. 111: Hydrolysis as a function of pH. http://www.oecd-ilibrary.org/docserver/download/9711101e.pdf?expires=1381235928&id=id&accname=guest&checksum=85BE2401C159EB0366E51E8EA0AEFB4E (Accessed October 2013)
Oxford Economics for the Silicones Environmental, Health and Safety Council of North America, 2010. The socio-economic impact of silicones in North America. http://sehsc.americanchemistry.com/Silicone-Uses/The-Socio-economic-Impact-of-Silicones-in-North-America.pdf (Accessed October 2013)
Powell, D.E., and Carpenter, J.C., 1997. Polyethermethylsiloxanes, pp. 225, in: Organosilicon materials, Chandra, G., (Ed.), Springer, Berlin, Heidelberg.
Radulovic, J., Sefiane, K., and Shanahan, M.E.R., 2009. Spreading and wetting behaviour of trisiloxanes. J. Bionic Eng. 6, 341.
Chapter 3
89
Radulovic, J., Sefiane, K., and Shanahan, M.E.R., 2010. Ageing of trisiloxane solutions. Chem. Eng. Sci. 65, 5251.
Spivack, J.L., Pohl, E.R., and Kochs, P., 1997. Organoalkoxysilanes, organosilanols, and organosiloxanols, pp. 105, Organosilicon materials, in: Chandra, G. (Ed.), Springer, Berlin, Heidelberg.
Van der Linden, W.E., 1992. The analytical chemistry of silicones. Anal. Chim. Acta 262, 348.
Venzmer, J., 2011. Superspreading - 20 years of physicochemical research. Curr. Opin. Colloid Interface Sci. 16, 335.
Wang, D.G., Norwood, W., Mehran, A., Byer, J.D., and Brimble, S., 2013. Review of recent advances in research on the toxicity, detection, occurrence and fate of cyclic volatile methyl siloxanes in the environment. Chemosphere 93, 711.
Chapter 3
90
Appendix A: Supplementary data
A.1 Discussion on the choice of containers for hydrolysis study as a function
of pH
The study was initially planned to be done in brown glass bottles, but the initial recoveries,
determined on day 0, right after spiking and homogenization, were not satisfying (less than
70 %, cf. Table S1 ). Given the surface active properties of the trisiloxane surfactants,
partition to interfaces was suspected and confirmed by several tests: i) when acetonitrile was
added to the aqueous solution before sampling (100 mL ACN for 100 mL solution), the
recoveries raised to almost 100 %, suggesting that the low recoveries were due to sorption on
interfaces (air/water or glass/water) rather than a fast degradation of
MD´(-(CH2)3-(EO)n-OCH3)M. ii) The possibility of loss by evaporation during sample
preparation was eliminated since less than 10 % difference was observed between samples
spiked before and after evaporation. Different containers were then tested to avoid sorption on
interfaces and to obtain satisfying initial recoveries. The results are shown in Table S1.
Sair/water stands for the contact surface between air and the solution and Scontainer/water refers to
the contact surface between the solution and the walls of the container.
Chapter 3
91
Table S1: Characteristics of the tested containers and initial recoveries (average over two replicates ± standard deviation) for aqueous solutions of
MD (́-(CH2)3-(EO)n-OCH3)M at 2 mg L-1.
Characteristics of the tested containers Recoveries in % Diameter
cm Height
cm Sair/water
cm² Scontainer/water
cm² Total S
cm² n=4 n=5 n=6 n=7 n=8 n=9 n=10 n=11 n=12
Average over all homologues
Brown glass 7 1.7 38 37 76 63±1 58.4±0.3 61±6 66±6 67±2 68.5±0.3 62.6±0.2 67±3 66±9 64
LDPE 1) 3.7 3.7 11 43 54 86±2 84.2±0.3 89±2 89±2 91±5 97±1 85±3 91±1 87±5 89
PFA 2) 4 3.7 13 46 59 97±16 92±9 90±5 89±5 95±11 87±0 87±5 84±8 84±3 89
PP 3) white cap square
section 5 2.1 20 33 53 81±4 86±3 85±6 91±3 95±3 100±2 94±3 96±2 98±2 92
PP colored cap round
section
2.6 9 5 73 79 109±14 104±4 100±1 94±1 96±7 95±10 89±10 96±16 92±6 97
PP white cap round section buffer pH 9
4.8 5.5 18 83 101 81±15 93±13 91±12 90±10 88±8 85±6 83±5 78±5 78±4 85
PP white cap round section buffer pH 7
4.8 5.5 18 83 101 63±4 79±7 82±7 84±6 84±6 82±6 81±6 77±6 78±5 79
1) Low Density Polyethylene; 2) Perfluoroalkoxy polymer; 3) Polypropylene.
Chapter 3
92
It seemed that the size and the shape of the container should be taken into account in the
selection of the appropriate container. Especially, a correlation could be found between the
recovery and the air/water contact area (Sair/water), as shown in Figure S1, suggesting a
partition of the surfactant molecules at the water/air interface.
Figure S1: Initial recoveries plotted over the air/water contact surface.
The possibility of adsorption on the walls of the PP bottles was also investigated. For that
purpose, a PP bottle previously filled with an aqueous solution of
MD´(-(CH2)3-(EO)n-OCH3)M at 2 mg L-1 (total surfactant concentration) was rinsed with
water and then extracted with a few mL of methanol. The extract was blown down to dryness,
reconstituted with the HPLC buffers, and analyzed by HPLC-MS/MS. The amount found on
the walls were respectively 50 %, 16 %, and 6 % of the initial concentration for the
homologues with n = 4, 5, and 6 and less than 2 % for the homologues with n > 6. This
proved that a significant adsorption of the homologues with n = 4 and 5 takes place on the PP
bottles. Those observations seem to confirm that in purely aqueous solution, the surfactant
molecules tend to migrate to the interfaces air/solution and container/solution. This creates an
unhomogeneity which is not favorable to a precise determination of the concentration in the
bulk of the solution. To minimize this phenomenon, the container type is chosen to minimize
the air/solution interface. In order to be adapted to the study, several other conditions were
required. The container should be big enough to contain 100 mL, be autoclavable, and be
possible to close tightly. Taking all those criteria, polypropylene bottles with a round section
appeared as a good compromise (good recoveries for MD´(-(CH2)3-(EO)n-OCH3)M with n > 5
Air/water contact surface in cm²
0 10 20 30 40
Rec
over
y in
%
50
60
70
80
90
100
110
Brown glassLDPEPFAPP white cap square sectionPP colored cap round sectionPP white cap round section
Chapter 3
93
in an aqueous solution, possibility to be autoclaved, resistance to temperatures up to 135°C,
availability in different volumes and shapes, low price) and were selected to carry out the
study.
The partition seems to depend also on:
- the concentration: no low initial recoveries were observed for aqueous solution in brown
glass bottles at 1 g L-1 (total surfactant concentration).
- the composition of the aqueous phase: initial recoveries were higher for the aqueous
solution buffered at pH 9 than at pH 7.
The possible influence of the sorption on glass on the determined concentrations in
environmental samples was verified: For that purpose, Neckar water was taken in different
containers (brown glass, PP, and PFA) and was spiked with MD´(-(CH2)3-(EO)n-OCH3)M to
reach 50 ng L-1 (total surfactant concentration). The samples were prepared and analyzed with
the described method and the obtained peak areas were compared. No differences higher than
15 % were observed, indicating that for river water in the ng L-1 range, the type of container
used for sampling does not significantly influence the determined concentration.
Chapter 3
94
Figure S 2: Graphical representation of ln(C/C0) as a function of time for the homologues n=5, n=7, n=9,
and n=11, for all conditions of pH and temperature tested in the study: ( ) 12°C, ( ) 25°C, and ( ) 50°C.
pH 7, n = 5
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n = 7
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n = 9
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 7, n = 11
Time in days
0 10 20 30
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 5
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 7
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 9
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
pH 9, n = 11
Time in days
0 2 4 6 8 10
ln(C
/C0)
-3
-2
-1
0
Chapter 3
95
A.2 Description of the identification of degradation products by HPLC-ESI-
TOF
The identification of degradation products was performed by high resolution mass
spectrometry, on a HPLC-ESI-TOF system. The HPLC unit was a 1290 high performance LC
system (Agilent Technologies, Waldbronn, Germany) equipped with a solvent cabinet, a
micro vacuum degasser, a binary pump, a high-performance autosampler with two 54 vial
plates, and a temperature controlled column compartment. The separation was carried out on a
PolymerX RP-1 (250 mm × 4.6 mm i.d., 5 µm, 100 Å) from Phenomenex. The series 1290
system was coupled with a 6540 UHD Q-TOF mass spectrometer (Agilent Technologies,
Waldbronn, Germany). A one year old solution of MD´(-(CH2)3-(EO)n-OCH3)M at 1 g L-1 in
MilliQ water was analysed by HPLC-MS. A very high number of masses were observed.
Most of the masses belonged to series, where the distance between two peaks is 44 Da, which
corresponds to one CH2CH2O group. The most intense series were investigated and three
structures could be attributed thanks to the exact mass and the isotopic pattern (Figure S 3). In
the case of silicon based molecules, the high resolution mass spectrometry is especially useful
to discriminate between Si-O and CH2CH2O groups which both have a mass of 44 Da.
Figure S 3: Identified degradation products of MD (́-(CH2)3-(EO)n-OCH3)M.
In order to be detected by HPLC-ESI-MS, the degradation products have to be ionized in the
source of the mass spectrometer. Only molecules with an ethylene oxide moiety were able to
be ionized by forming NH4+ adducts. The number of degradation products that can be
identified by ESI-MS is therefore limited and other techniques would be necessary. The
corresponding masses were followed during the hydrolysis study and only the degradation
product (1) showed a significant increase of its intensity during the time scale of the study.
n n
n
(1) (2) (3)
Chapter 3
96
Figure S 4: a: HPLC-MS/MS chromatograms of direct injections of MD (́-(CH2)3-(EO)n-OCH3)M at
20 µg L-1 (Figure S4 a) and Triton X-100 at 1 mg L-1 (Figure S4 b).
a
2 4 6 8 10 12 14 16 18 20 22 24 26 0.0
4.0e3
8.0e3
1.2e4
1.6e4
2.0e4
2.4e4
2.8e4
Time in min
n = 4, 488/103 488/381
n = 6, 576/103 576/147
664/103
n = 8 664/221
752/103 n = 10 752/221
840/103 n = 12 840/221
Si O Si O Si
C H 2 C H 2
C H 2 (EO) n OCH 3
n = 8, 576/147
C H 3
C H 3 C H 3
C H 3
C H 3
O (EO) n OH
b
2 4 6 8 10 12 14 16 18 20 22 24 26 28
0.0
2.0e3
4.0e3
6.0e3
8.0e3
1.0e4
1.2e4
1.4e4
1.6e4
1.8e4
Time in min
n = 7, 532/147
n = 9, 620/147
n = 13, 796/147
Inte
nsit
y in
cps
Inte
nsit
y in
cps
Chapter 4
97
Chapter 4
Mobility of a polyether trisiloxane surfactant on
soil: soil/water distribution coefficients and
leaching in a soil column
Chapter 4
98
TABLE OF CONTENTS
Abstract .................................................................................................................................... 99
1. Introduction .................................................................................................................... 100
2. Materials and methods ................................................................................................... 102
2.1 Chemicals .................................................................................................................. 102
2.2 HPLC-MS/MS analysis ............................................................................................ 102
2.3 Soil extraction ........................................................................................................... 103
2.4 Sorption experiment .................................................................................................. 103
2.4.1 Soils .................................................................................................................. 103
2.4.2 Preliminary experiment .................................................................................... 104
2.4.3 Sorption kinetics ............................................................................................... 104
2.4.4 Sorption isotherms ............................................................................................ 104
2.5 Leaching in soil column ............................................................................................ 104
3. Results and discussion .................................................................................................... 106
3.1 Sorption using a batch equilibrium method .............................................................. 106
3.1.1 Preliminary experiment .................................................................................... 106
3.1.2 Sorption kinetics ............................................................................................... 106
3.1.3 Sorption isotherms ............................................................................................ 109
3.2 Leaching in soil column ............................................................................................ 114
3.2.1 Acesulfame as a tracer ...................................................................................... 114
3.2.2 The trisiloxane surfactant HANSA ADD 1055 ................................................ 115
Conclusion .............................................................................................................................. 119
Acknowledgments .................................................................................................................. 119
References .............................................................................................................................. 120
Appendix A ............................................................................................................................ 122
References .............................................................................................................................. 132
Chapter 4
99
Abstract
Polyether trisiloxane surfactants are widespread used as agricultural adjuvants because they
increase the activity and the rainfastness of pesticides. On the contrary to pesticides, the
environmental fate of agricultural adjuvants has not been much investigated, yet. Especially
for trisiloxane surfactants, the knowledge on their environmental fate is scarce. To fill this
gap, the mobility of a polyether trisiloxane surfactant on soil was studied. With a sorption
batch equilibrium method, distribution coefficients between water and soil (Kd, Koc, and Kclay)
were estimated for two standard soils (loam and sandy loam) and for every homologue of the
trisiloxane surfactant. The obtained values for Kd were between 15 L kg-1 and 135 L kg-1,
indicating that the trisiloxane surfactant is only slightly mobile in soil. The leaching in soil
column was studied in a worst case scenario where the application of the trisiloxane surfactant
was done on quartz sand and was immediately followed by a heavy rainfall. At the end of the
experiment, less than 0.01 % of the initially applied trisiloxane surfactant has leached through
20 cm of quartz sand. Based on the Kd values and on the results of the leaching in soil
column, the studied trisiloxane surfactant seems to be unlikely to reach ground water after
application as agricultural adjuvant.
Chapter 4
100
1. Introduction
Trisiloxane surfactants are increasingly used as agricultural adjuvants in combination with
herbicides, fungicides, insecticides, nutrients, or growth regulators because of their ability to
improve the activity and application characteristics of plant protection products. They reduce
the surface tension of water from 72 mN m-1 to 22 mN m-1 (at 24°C and 0.1 %) and promote a
rapid and important spreading on hydrophobic surfaces like leaves (Penner et al., 1999;
Knoche et al., 1991). This effect gave them the name of superspreaders or superwetters. The
chemical structures of the three main trisiloxane surfactants used in agriculture are
represented in Figure 1, together with an illustration of spreading on a leaf, with and without
silicone surfactant.
Figure 1: (a) Chemical structure of the three main trisiloxane surfactants used in agriculture. M stands
for (CH3)3-Si-O1/2-, D’ for O1/2-Si(CH3)(R )́-O1/2-; (b) Wetting of a leaf by water with (lower part) and
without (upper part) trisiloxane surfactant.
When used as agricultural adjuvants, trisiloxane surfactants may reach the soil compartment
and could enter the aquatic environment by leaching to groundwater or by runoff from land
into surface waters (Cornejo et al., 2000; Organisation for Economic Co-operation and
Development, 2000). Knowledge of their behavior on soil is needed to estimate their potential
for leaching and runoff and therefore to assess the potential exposure of the aquatic
environment to trisiloxane surfactants. This knowledge is particularly important since data on
the toxicity of trisiloxane surfactants to the aquatic environment are scarce. The toxicity of the
trisiloxane surfactant has been studied on several aquatic organisms: The No Observed Effect
Concentration (NOEC) has been found at 0.56 mg L-1 for Zebra Fish (96 h exposure time) and
at 3.2 mg L-1 for Rainbow Trout (96 h exposure time); regarding the toxicity to algae, the
NOEC to Selenastrum capricornutum was determined at 1 mg L-1 (96 h exposure); and the
Chapter 4
101
NOECs were 10 mg L-1 and 25 mg L-1 for Daphnia similis and Daphnia magna, respectively
(both 48 h exposure). No data on the chronic toxicity of the trisiloxane surfactant to aquatic
organisms are available. On the contrary to pesticides, the behavior of adjuvants on soil has
not been much investigated yet, partly due to the lower application rates compared to the
pesticides themselves. Among non-ionic surfactants used as agricultural adjuvants, most of
the existing studies focus on alcohol ethoxylates, alkylamine ethoxylates, or alkylphenol
ethoxylates (Krogh et al., 2003), but the higher efficiency of trisiloxane surfactants to reduce
surface tension compared to carbon-based surfactants raises the question of their particular
behavior on soil. A study already exists on the fate of one trisiloxane surfactant
(R = C(O)CH3) on soil but the lack of a specific analytical method led to ambiguous results
(Griessbach et al., 1997; Griessbach et al., 1998). The test substance was a trisiloxane
surfactant, 14C-radiolabelled at the terminal methyl group of the hydrophilic moiety, and the
detection was performed by liquid scintillation counting. The sorption/desorption experiments
concluded that the trisiloxane surfactant is immobile on soil (KF between 8 µg1-1/n(cm³)1/ng-1
and 75 µg1-1/n(cm³)1/ng-1) (Griessbach et al., 1997) while a soil column leaching experiment
revealed the presence of 14C-radiolabelled substance in the leachate, without possible
discrimination between the trisiloxane surfactant or its degradation products (Griessbach et
al., 1998).
The degree of sorption on soil depends on the properties of the soil (organic matter, clay
content, metallic oxides, for example) and of the sorbate (Navarro et al., 2007). Due to the
surfactant nature of the trisiloxane, the prediction of the behavior on soil is difficult.
Surfactants contain two distinct parts, each having different physico-chemical properties and
different affinities for the soil components. To gain information on the behavior of a
trisiloxane surfactant on soil, two studies have been carried out and are reported in this paper:
sorption/desorption experiments according to the OECD guideline 106 (Organisation for
Economic Co-operation and Development, 2000) and leaching in soil column experiments.
The aim of the sorption/desorption experiments was not to elucidate the sorption mechanism
of the trisiloxane surfactant but to provide representative values of the sorption coefficients on
agricultural soils. The aim of the lysimeter experiments was to estimate its potential for
leaching throughout the soil in a worst case scenario where the application of the trisiloxane
surfactant is done on quartz sand and immediately followed by a simulated heavy rainfall. If
no leaching would be observed in this case, one could reasonably estimate that the chances for
trisiloxane surfactants to reach ground water are very low. This work focuses on one
Chapter 4
102
trisiloxane surfactant, with R = CH3 (CAS: 27306-78-1), Figure 1, denoted MD´(-(CH2)3-
(EO)n-OCH3)M in this paper.
2. Materials and methods
2.1 Chemicals
The reference substance of the trisiloxane surfactant used for the sorption experiments was
purchased from ABCR (Karlsruhe, Germany) and had a purity of 90 %. Stock solutions were
prepared in methanol and were stored at -18 °C. The soil column experiments were performed
with a commercial trisiloxane surfactant, HANSA ADD 1055, provided by CHT R. Beitlich
(Tübingen, Germany). The acesulfame potassium used as a tracer in the soil column
experiment was purchased from Fluka (Buchs, Switzerland) and the standard of acesulfame
potassium (purity > 99 %) used for analysis was purchased from Fluka (Steinheim, Germany).
All organic solvents used for the analysis were of HPLC grade. Their purity was at least
99.8 %. Methanol and dichloromethane were provided by Carl Roth GmbH (Karlsruhe,
Germany), tetrahydrofurane and isopropanol were purchased from Merck (Darmstadt,
Germany) and acetonitrile was obtained from Scharlau S.L. (Sentmenat, Spain). Ultrapure
water was produced by an Arium 611 UV laboratory water purification system from Sartorius
(Göttingen, Germany). Ammonium acetate was purchased from Fluka (Steinheim, Germany,
purity > 98.0 %) and calcium chloride was obtained from Merck (Darmstadt, Germany).
2.2 HPLC-MS/MS analysis
The analysis of MD´(-(CH2)3-(EO)n-OCH3)M was performed by HPLC-MS/MS, according to
Michel et al. (2012), either by direct injection or after liquid-liquid extraction by
dichloromethane. The sample preparation by liquid-liquid extraction with dichloromethane
was carried out as described in Michel et al. (2012). For direct injection, 500 µL of sample
were transferred in a HPLC vial, blown down to dryness under nitrogen, and reconstituted
with eluent A. The HPLC unit was a 1200 HPLC system (Agilent Technologies, Waldbronn,
Germany) equipped with a solvent cabinet, a micro vacuum degasser, a binary pump, a high-
performance autosampler with 54 vial plates and a temperature controlled column
compartment. The conditions applied for the chromatographic separation are summarized in
the Table S 2 of the supplementary material. The detection was achieved with a triple
quadrupole mass spectrometer API 4000 Q-Trap (Applied Biosystems/MDS Sciex
Instruments, Concord, Canada) equipped with an ESI source and used in multiple reaction
monitoring (MRM) mode. The mass spectrometer was operated in positive ionization mode
Chapter 4
103
and silicone surfactant homologues were detected as ammonium adducts. Data acquisition
was performed with the Analyst software (version 1.5.1). All glassware was treated in a
pyrolysis oven model LHT6 / 120 from Carbolite (Hope Valley, England) at 550 °C for
60 min after use. The limits of quantifications were calculated for each homologue from the
residual standard deviation of the linear calibration function. They were between 0.9 ng L-1
and 3.1 ng L-1 for the samples analyzed after liquid-liquid extraction and between 4 µg L-1
and 20 µg L-1 for the samples analyzed by direct injection. The individual values for each
homologue are given in supplementary material (Table S 1). The analytical method is
specific, sensitive and allows obtaining information on the oligomeric distribution of the
trisiloxane surfactant.
The analysis of acesulfame was performed on the same HPLC-MS/MS system as the
trisiloxane surfactant. The chromatographic conditions are summarized in the Table S 2 of the
electronic supplementary material. All samples for acesulfame were analyzed by direct
injection: to 500 µL of sample, 50 µL of internal standard (acesulfame–d4) at 0.1 mg L-1 were
added, the sample was blown down to dryness and reconstituted with 100 µL of eluent B’ and
400 µL of eluent A’ before injection of 10 µL on the HPLC-MS/MS.
2.3 Soil extraction
The soil to be extracted was weighted in an aluminum plate (Carl Roth, Karlsruhe, Germany)
and freeze-dried (Christ Beta 2-16, Osterode am Harz, Germany). The water content of the
soil was determined as the weight difference of soil sample before and after freeze-drying. For
extraction, 2 g of dried soil were weighted in a 50 mL glass flask and were sonicated for
10 min with 10 mL of methanol. The system was centrifuged and the supernatant solvent was
collected. The extraction was repeated three times and all three extracts were combined. The
final extract was blown down to dryness and reconstituted with the eluent A followed by
analysis by HPLC-MS/MS. For each soil analysis, triplicates were prepared.
2.4 Sorption experiment
2.4.1 Soils
The three standard soils, LUFA No. 2.1, 2.4, and 5M, used for the sorption experiment were
purchased from Landwirtschaftliche Untersuchungs- und Forschungsanstalt Speyer (LUFA
Speyer, Germany). The characteristics of the soils are given in supplementary material,
Table S 3.
Chapter 4
104
2.4.2 Preliminary experiment
The preliminary experiment was carried out on soils 2.1, 2.4, and 5M, and for each soil, three
soil/solution ratios were tested: 1/10: 10 g of soil; 1/25: 4 g of soil; 1/50: 2 g of soil. The
soil/solution systems were prepared by weighting the amount of soil and adding 100 mL of
CaCl2 solution at 0.01 M. After overnight equilibration (overhead shaker C. Gerhardt GmbH
& Co. KG, Königswinter, Germany), the systems were spiked with MD´(-(CH2)3-(EO)n-
OCH3)M to reach 2 mg L-1, the aqueous phases were sampled and analyzed after 4 h, 6 h,
24 h, 48 h. The aim of the preliminary experiment was to determine the best soil/solution ratio
and the time required for sorption equilibrium.
2.4.3 Sorption kinetics
According to the results of the preliminary experiment, the sorption kinetics was studied at
2 mg L-1, on soils 2.4 and 5M, and with a soil/solution ratio 1/25. The parallel method was
adopted for the study, which means that for each soil, 10 samples with the same soil/solution
ratio were prepared, as many as time intervals chosen for the study. At each sampling point
(after 2 h, 4 h, 6 h, 8 h, and 24 h), duplicates were centrifuged for 15 min at 1500 rounds min-1
(8K centrifuge from Sigma, Osterode am Harz, Germany) to separate the aqueous phase and
the solid phase. The aqueous phase was sampled and analyzed by direct injection on HPLC-
MS/MS and the solid phase was extracted and analyzed according to the procedure described
above. This procedure was repeated at each sampling point.
2.4.4 Sorption isotherms
The sorption isotherms were obtained on soils 2.4 and 5M, with a soil/solution ratio of 1/25
and using five different concentrations of trisiloxane surfactant: 2 mg L-1, 1.5 mg L-1,
1 mg L-1, 0.5 mg L-1, 0.1 mg L-1. Two replicates were prepared for each concentration. Both
the solid phase and the aqueous phase, were analyzed separately. The concentrations of
MD´(-(CH2)3-(EO)n-OCH3)M measured in each phase were used to build the sorption
isotherms.
2.5 Leaching in soil column
The leaching in soil column was studied on a lab lysimeter from EcoTech Umwelt-
Messsysteme GmbH, Bonn, Germany. The lysimeter consisted of a polypropylene cylinder of
25 cm length and 30 cm diameter, covered on top by a sprinkling head equipped with stainless
steel needles of 0.40 mm diameter, and at the bottom with a polyamide membrane with a pore
Chapter 4
105
size of 0.45 µm. The cylinder was filled with 20 cm of medium sized quartz sand from Carlo
Bernasconi AG, Zurich, Switzerland. Characteristics of this sand are given in the
supplementary material, Table S 4. A short soil column (20 cm) was chosen for this study
based on the results of the sorption studies, indicating a strong sorption of the trisiloxane
surfactant on soil. The column was equipped with pH meters and pF meters located at 10 cm,
15 cm, and 20 cm depth. Before the beginning of the experiment, the column was filled with
quartz sand, conditioned with deionized water to reach the maximal water holding capacity,
and was then allowed to drain off by gravity for 24 h. The pore volume was calculated from
the weight difference between the water saturated column and the dry column and was
estimated at 3.42 L. The porosity, calculated as the ratio of pore volume over the total volume
of sand in the column has been estimated at 0.24. 7923 µg of HANSA ADD 1055 and 454 µg
of acesulfame were applied as 113 mL of a solution of HANSA ADD 1055 (70 mg L-1) and
acesulfame (4 mg L-1) in deionized water. The application was done with a sprayer of 500 mL
(Bürkle GmbH, Bad Bellingen, Germany). The amount of trisiloxane surfactant applied
corresponded to 1.1 L ha-1 which is approximately ten times more than the normal application
rate of HANSA ADD 1055 on agricultural fields. Immediately after application, artificial
rainfall was simulated by a peristaltic pump delivering deionized water at a constant flow rate
of 9.5 mm h-1, which corresponds to a heavy rainfall. A peristaltic pump connected at the
bottom of the lysimeter and set at -190 hPa allowed the water to leach through the column
with a constant flow. Pore water samples were taken at 5 cm, 10 cm, and 15 cm depth by
means of PTFE tubes connected to 10 mL polypropylene syringes. The leachate, i.e. the water
collected at the bottom of the column, was quantitatively collected in a 10 L polypropylene
bottle and regularly sampled: every hour during the first day, and then once or twice per day
for 6 days. Samples taken between 5 cm and 15 cm, were analyzed by direct injection on
HPLC-MS/MS, while the leachate samples were concentrated by liquid-liquid extraction
before analysis by HPLC-MS/MS. After the experiment, the soil was removed and separated
in seven segments: 0-2 cm, 2-4 cm, 4-6 cm, 6-8 cm, 8-10 cm, 10-15 cm, 15-20 cm. The sand
corresponding to each segment was carefully homogenized and was stored in freezer at -18°C
until extraction and analysis according to the method described above.
Chapter 4
106
3. Results and discussion
3.1 Sorption using a batch equilibrium method
3.1.1 Preliminary experiment
A comprehensive description of the results is given in the supplementary material
(Figure S 1). On soils 2.4 and 5M, the sorption equilibrium was reached after 24 h. On soil
2.1, no sorption equilibrium could be obtained, even after 48 h. Given that the soil 2.1 is
acidic (pH 5.2) (United States Department of Agriculture, 2013) and that
MD´(-(CH2)3-(EO)n-OCH3)M is known to be very sensitive to hydrolysis, one can assume
that the lack of sorption equilibrium reflects the degradation of MD´(-(CH2)3-(EO)n-OCH3)M.
This result suggests that in contact with wet acidic soils, the degradation of MD´(-(CH2)3-
(EO)n-OCH3)M by hydrolysis is quite fast.
3.1.2 Sorption kinetics
The kinetics of sorption on soils 2.4 and 5M is detailed in the supplementary material, Figures
S 2 and S 3. The distribution of all homologues between the aqueous phase and the soil at
sorption equilibrium (after 24 h) is shown in Figure 2, together with the mass balance.
Chapter 4
107
Figure 2: Distribution of every homologue of MD (́-(CH2)3-(EO)n-OCH3)M between the aqueous phase
and the soil together with the mass balance on soil 2.4 and 5M.
Both for soils 2.4 and 5M, a satisfying mass balance was obtained (between 80 % and 110 %,
except for the homologue n = 4). The total amount of test substance initially applied was
recovered, either in the soil or in the aqueous phase which means that
MD´(-(CH2)3-(EO)n-OCH3)M did not degrade during the time of the experiment. As already
described in a previous paper (Michel et al., 2014), adsorption of the homologue n = 4 on PP
Chapter 4
108
container is possible under certain conditions. Due to the experimental uncertainty resulting
from this phenomenon, MD´(-(CH2)3-(EO)4-OCH3)M was dismissed from the data set for the
estimation of the sorption coefficients.
It appears clearly on Figure 2 that the extent of sorption increases with the number of ethylene
oxide groups, especially for the soil 5M. This result can at first be considered as
contradictory. Since the hydrophilicity of MD´(-(CH2)3-(EO)n-OCH3)M increases with the
number of EO, it is expected that the homologues with a longer EO chain partition rather to
the aqueous phase. However, it is known from the study of other ethoxylated surfactants that
the sorption on soil is the result of different sorption mechanisms (Lara-Martín et al., 2008):
hydrophobic interactions between the hydrophobic moiety of the surfactant and the particulate
phase and hydrophilic interactions through hydrogen bonds between the EO chain and the
clay mineral of the solid phase (Krogh et al., 2003). In the case of
MD´(-(CH2)3-(EO)n-OCH3)M, all homologues have the same hydrophobic moiety. The
difference of sorption between the homologues therefore highlights the existence of
hydrophilic interactions between the EO chain and the soil.
The distribution coefficients Kd (in L kg-1) were calculated based on the following equation:
(eq)ads aq
C
(eq)adss
C
dK (1)
Csads(eq) is the content of substance sorbed on the soil at sorption equilibrium (in µg g-1) and
Caqads (eq) is the mass concentration of the substance in the aqueous phase at sorption
equilibrium (in mg L-1).
The results reported in this paper do not allow determining which mechanisms are involved in
the sorption of trisiloxane surfactant and which factor has a dominating influence. However, it
is probable that the extent of sorption depends on the organic content and the clay content of
the soil. For this reason both the organic carbon normalized coefficient (Koc) and the clay
content normalized coefficient (Kclay) (Cornejo et al., 2000) have been calculated, according
to the following equations:
%oc
100dKocK (2)
%clay
100dKclayK (3)
The obtained values of Kd, Koc, and Kclay are summarized in Table 1. The higher the Kd value,
the less likely a chemical will move throughout the soil. On soil 2.4 and 5M, log Koc for
Chapter 4
109
MD´(-(CH2)3-(EO)n-OCH3)M is between 3 and 4, which corresponds to a slightly mobile
compound (Food and Agriculture Organization of the United Nations, 2000).
3.1.3 Sorption isotherms
According to the Freundlich sorption equation:
(eq)adsaqC log
adsn
1
FadsK log(eq)ads
sC log (4)
Where Csads (eq) is the concentration of substance sorbed on the soil at sorption equilibrium
(in µg g-1), Caqads (eq) is the concentration of the substance in the aqueous phase at sorption
equilibrium (in µg mL-1), KadsF is the Freundlich sorption coefficient (in µg1-1/n(cm³)1/ng-1),
and nads is the regression constant. The Freundlich coefficient gives an indication on the extent
of sorption: it represents the concentration of test substance in the soil when the concentration
in the aqueous phase is equal to 1 µg mL-1. 1/nads takes into account the non-linearity of the
sorption and generally ranges between 0.7 and 1.0. KadsF, 1/nads, and the correlation
coefficients (Rads²) were determined for all homologues by fitting the experimental data to the
Freundlich equation and the obtained values are summarized in Table 1. The sorption
isotherm for the homologue n = 7 is represented as an example in Figure 3. All other sorption
isotherms are given in the supplementary material (Figure S 4 and S 5).
The correlation coefficients for the sorption isotherms are almost all above 0.99, showing that
the sorption of the trisiloxane surfactant follows the Freundlich model. The KadsF values
obtained on soil 2.4 were higher than the one obtained on soil 5M. The soil 2.4 has higher
organic carbon content and clay content, both factors which promote sorption.
If the coefficients 1/nads would be equal to one, KFads would be equal to Kd and the plot of
Csads as a function of Caq
ads would be linear. For the trisiloxane surfactant studied, 1/nads is
close to one.
The Freundlich sorption coefficient had already been measured for another trisiloxane
surfactant, MD´(-(CH2)3-(EO)7-OC(O)CH3)M, on five different soils and was found to be
between 7.4 µg1-1/n(cm³)1/ng-1 and 75.2 µg1-1/n(cm³)1/ng-1 (Griessbach et al., 1997). Direct
comparison with the current results is not possible given that the soils and the test substances
are different. However, one can notice that the values of KadsF for MD´(-(CH2)3-(EO)n-
OCH3)M and MD´(-(CH2)3-(EO)7-OC(O)CH3)M are in the same order of magnitude.
Furthermore, in the work from Griessbach et al (1997), no homologue specific Freundlich
sorption coefficients had been determined. It was therefore not possible to observe the
influence of the EO chain length on sorption.
Chapter 4
110
The measured Kd coefficients were compared to the Kd coefficients predicted from the Kow
values of the trisiloxane surfactant (cf chapter 1, section 4.2.1) according to the following the
equation (Gerstl and Mingelgrin, 1984):
log Kom = 4.4 + 0.72 · log Kow (5)
Where Kom = 1.724
Kd ·
%oc
100
The predicted Kd values for MD´(-(CH2)3-(EO)n-OCH3)M on the soils 2.4 and 5M were in the
range 2 104 L kg-1 – 6 105 L kg-1, several orders of magnitude away from the measured values.
The equation (5) is based on the assumptions that the organic matter of the soil is the main
factor influencing the sorption of the test substance and that the involved interactions are
mainly non-polar. The significant difference between the measured and predicted Kd values
may be a hint that the initial assumptions of the equation (5) are not applicable to the
trisiloxane surfactant. Organic matter of the soil may not be the main factor influencing the
sorption of the trisiloxane surfactant on soil and interactions others that non-polar interactions
may be involved.
Chapter 4
111
Figure 3 Sorption isotherm of MD (́-(CH2)3-(EO)7-OCH3)M on soil 2.4 and 5M.
To ensure reliable results in spite of the lack of internal standard, the extraction recovery and
the matrix effect were evaluated for each soil and at each concentration applied in this study.
For the extraction recovery, triplicates of fresh soil samples were spiked before extraction
with MD´(-(CH2)3-(EO)n-OCH3)M to reach the desired concentration. After freeze-drying and
extraction, the obtained samples were analyzed by HPLC-MS/MS and compared to an
external standard. The extraction yield was calculated as the ratio of peak area in the sample
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm soil 2.4
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-1
0.01
0.1
1
10
Sorption isotherm soil 5M
Chapter 4
112
spiked before extraction over the peak area in the external standard. The obtained extraction
recoveries were between 70 % and 90 % and were included in the calculation of the amount
of MD´(-(CH2)3-(EO)n-OCH3)M adsorbed on soil. To test the matrix effect for the analysis of
solid samples, triplicates of fresh soil samples were extracted with the previously described
procedure and MD´(-(CH2)3-(EO)n-OCH3)M was spiked after extraction at a level of
2 mg L-1. The obtained signal was compared to an external standard. Less than 10 %
difference was observed, indicating that no significant matrix effect occurs. Matrix effect for
the analysis of the aqueous phase of the soil/solution systems was tested by equilibrating
overnight a soil/solution system, separating the aqueous and solid phase by centrifugation and
by spiking the aqueous phase with MD´(-(CH2)3-(EO)n-OCH3)M at five levels of
concentrations (2 mg L-1, 1.5 mg L-1, 1 mg L-1, 0.5 mg L-1, 0.1 mg L-1) corresponding to the
five concentrations used to obtain the sorption isotherms. The spiked samples were compared
to external standards and less than 10 % difference was observed confirming the absence of
matrix effect.
Chapter 4
113
Table 1: Sorption coefficient (Kd in L kg-1), organic carbon normalized sorption coefficient (Koc in L kg-1), clay content normalized sorption coefficient (Kclay in L kg-1),
Freundlich sorption (KadsF µg1-1/n(cm³)1/ng-1), regression constants (nads), and correlation coefficients (Rads²) for all homologues of MD (́-(CH2)3-(EO)n-OCH3)M on both
soils studied.
Soil 2.4 Soil 5M
n Kd at 2 mg L-1 Koc 2 mg L
-1 Kclay 2 mg L-1 Kads
F 1/nads Rads² Kd at 2 mg L-1 Koc 2 mg L
-1 Kclay 2 mg L-1 Kads
F 1/nads Rads²
5 59 2687 227 46.2 0.9485 0.99 22 2368 196 20.1 0.9658 1.00
6 49 2214 187 39.7 0.9307 0.99 20 2092 173 18.5 0.9571 1.00
7 43 1930 163 36.4 0.9225 0.99 21 2165 179 18.2 0.9182 1.00
8 44 1998 169 36.1 0.9163 0.99 25 2656 219 21.6 0.9195 1.00
9 52 2334 197 38.2 0.8912 0.99 36 3759 311 27.2 0.8763 1.00
10 64 2896 244 42.7 0.8692 0.99 51 5345 442 33.4 0.8424 1.00
11 85 3867 326 51.0 0.8577 1.00 84 8844 731 47.1 0.8333 1.00
12 133 6039 509 58.8 0.8341 0.99 132 13862 1145 59.3 0.8178 1.00
Chapter 4
114
3.2 Leaching in soil column
3.2.1 Acesulfame as a tracer
In order to evaluate the leaching ability of the trisiloxane surfactant, the use of a conservative
substance applied simultaneously with MD´(-(CH2)3-(EO)n-OCH3)M was desired. In order to
be a good reference, the chosen substance must meet several criteria: it should not be retained
by the soil, it should be stable during the time scale of the experiment, and it should not
interact with the test substance. Ionic compounds are commonly used for that purpose.
However, since MD´(-(CH2)3-(EO)n-OCH3)M forms complexes with several cations (K+, Na+,
or NH4+ for instance (Bonnington, 2000)), the simultaneous application of
MD´(-(CH2)3-(EO)n-OCH3)M with a cation might change its leaching behavior. As a
consequence, acesulfame was chosen as an alternative reference. Because of its persistence
and mobility in soil, acesulfame is considered as a tracer of wastewater impact on
groundwater (Robertson et al., 2013; Van Stempvoort et al., 2011a; Van Stempvoort et al.,
2011b). One can assume that at the scale of our experiment, and with the type of soil used in
this study, acesulfame will not be degraded and will not be retained by the soil. Moreover, the
low limits of quantification reached by HPLC-MS/MS allow applying a small quantity of
acesulfame, reducing the eventual interactions with the trisiloxane surfactant. Two results
confirmed the applicability of acesulfame as a tracer in this soil column experiment: i) a
preliminary study on a quartz sand column of 70 cm depth showed that acesulfame and CsCl
leach simultaneously (graph in supplementary material, Figure S 6), ii) 100 % of the
acesulfame applied initially is found in the leachate.
The leaching of acesulfame through the column could be followed by taking side samples at
5 cm, 10 cm, and 15 cm at regular time intervals. Based on the obtained curves
(Supplementary material, Figure S 7), the water flow rate through the column was estimated
between 0.20 cm min-1 and 0.25 cm min-1.
Relative and cumulative breakthrough curves
The relative and cumulated breakthrough curves for acesulfame were determined based on the
concentrations measured in the leachate.
Chapter 4
115
Figure 4: Relative breakthrough curve (Figure 4a) and cumulated breakthrough curve (Figure 4b) for
acesulfame.
The mass of acesulfame recovered in the leachate is around 80 %. Taking into account the
mass of acesulfame taken away when sampling at 5 cm, 10 cm, and 15 cm, the mass balance
for acesulfame reaches 100 %.
3.2.2 The trisiloxane surfactant HANSA ADD 1055
The concentrations of MD´(-(CH2)3-(EO)n-OCH3)M in the side samples taken at 5 cm, 10 cm,
and 15 cm were lower than the LOQs (direct injection). In the leachate,
Pore volumes
0 2 4 6 8 10 12 14 16
Per
cent
age
of i
niti
ally
app
lied
ac
esul
fam
e fo
und
in t
he l
each
ate
0
5
10
15
20
25
Pore volumes
0 2 4 6 8 10 12 14 16
Cum
ulat
ive
perc
enta
ge o
f in
itia
lly
appl
ied
aces
ulfa
me
foun
d in
the
lea
chat
e
0
20
40
60
80
100
a
b
Chapter 4
116
MD´(-(CH2)3-(EO)n-OCH3)M could be quantified and the relative and cumulative
breakthrough curves could be built (Figure 5).
Relative and cumulative breakthrough curves
Figure 5: Relative breakthrough curve (Figure 5a) and cumulated breakthrough curve (Figure 5b) for
MD (́-(CH2)3-(EO)7-OCH3)M.
Peaks corresponding to MD´(-(CH2)3-(EO)n-OCH3)M were observed in the different fractions
of the leachate, indicating that a small amount of MD´(-(CH2)3-(EO)n-OCH3)M did leach
through the column. A first peak comes around 0.5 pore volumes, and a second peak comes
after 8 pore volumes. The observation of an early peak, of trisiloxane surfactant in the
Pore volumes
0 2 4 6 8 10 12 14 16
Per
cent
age
of i
niti
ally
ap
plie
d M
D´(
-(C
H2)
3-(E
O) 7-
OC
H3)
M
foun
d in
the
lea
chat
e
0.0000
0.0002
0.0004
0.0006
0.0008
0.0010
0.0012
0.0014
0.0016
0.0018
Pore volumes
0 2 4 6 8 10 12 14 16
Cum
ulat
ed p
erce
ntag
e of
ini
tial
ly a
ppli
ed M
D´(
-(C
H2)
3-(E
O) 7-
OC
H3)
M
foun
d in
the
lea
chat
e
0.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
a b
Chapter 4
117
leachate, coming simultaneously with acesulfame, could come from the leaching of the test
substance sorbed to colloids. It is indeed known that the migration of colloids in soil can
enhance the leaching of strongly sorbing contaminants (Laegdsmand et al., 2007). However
the amount of MD´(-(CH2)3-(EO)n-OCH3)M found in the leachate represents less than 0.01 %
of the initially applied MD´(-(CH2)3-(EO)n-OCH3)M. The strong sorption of the trisiloxane
surfactant on soil was confirmed by the analysis of the soil segments.
Distribution of MD (́-(CH2)3-(EO)n-OCH3)M between soil and water
Unlike acesulfame, almost all the MD´(-(CH2)3-(EO)n-OCH3)M was recovered from the soil
(between 75 % for MD´(-(CH2)3-(EO)12-OCH3)M and 96 % for
MD´(-(CH2)3-(EO)8-OCH3)M. The distribution of MD´(-(CH2)3-(EO)7-OCH3)M in the
different segments of the soil column is shown in Figure 6.
Figure 6: Distribution of MD (́-(CH2)3-(EO)7-OCH3)M in different segments of the soil column. Duration
of the experiment: 6 d, volumetric flow rate: 11.2 mL min-1, total amount of water applied: 51 L.
50 % of the initially applied MD´(-(CH2)3-(EO)7-OCH3)M was found in the first 6 cm of soil.
Between 80 % and 95 % are contained in the first 15 cm and less than 1 % in the segment
15 cm to 20 cm. The leaching distance, i.e. the deepest soil segment in which 0.5 % of the
applied test substance was found, is in this experiment located between 10 cm and 15 cm. The
error bars represent the standard deviation over triplicate samples.
Percentage of initially applied MD´(EO7OMe)M
in different soil segments
0 5 10 15 20 25
Dep
th i
n cm
15-20
10-15
8-10
6-8
4-6
2-4
0-2
Chapter 4
118
Based on the results of the soil column experiment, a retardation factor Rd, defined as the ratio
of the travel velocity of water over the travel velocity of the test compound, was calculated.
The travel velocity of water in the soil column was calculated at 0.07 cm min-1 based on the
following equation: porositysectioncolumnSoil
rateflowVolumetricwV
(6)
The travel velocity of the trisiloxane surfactant was estimated from its leaching distance in the
soil at the end of the experiment. After 8460 min, MD´(-(CH2)3-(EO)7-OCH3)M has reached
15 cm (Figure 6): the maximal travel velocity of MD´(-(CH2)3-(EO)7-OCH3)M in the soil
column is therefore 0.0018 cm min-1. This value corresponds to the travel velocity of the
leaching front and thus represents a maximal travel velocity. The retardation factor for
MD´(-(CH2)3-(EO)7-OCH3)M, given here as an example, is thus 38. This means that under the
conditions applied in this experiment MD´(-(CH2)3-(EO)7-OCH3)M is at least 38 times slower
that the water flow.
The soil/water distribution coefficient Kd can be calculated from the retardation factor
according to the following equation:
dKερ
1d
R (7)
Where ρ is the sand volumetric mass density (g cm-³) and ε is the porosity (dimensionless).
The organic carbon normalized distribution coefficient Koc calculated from Rd with the
equations (2) and (7) is 2901 L kg-1. The order of magnitude of this value is in good
agreement with the Koc values determined by the batch equilibrium method (1930 L kg-1
calculated with soil 2.4 and 2165 L kg-1 calculated with soil 5M).
Chapter 4
119
Conclusion
The study presented here provides important information on the mobility of a trisiloxane
surfactant in the environment. The soil/water distribution coefficients obtained for one
trisiloxane surfactant on two standard soils (average Kd over all homologues 66 cm³ g-1 on
loam and 49 cm³ g-1 on sandy loam) indicate that the trisiloxane surfactant is only slightly
mobile in soil. This result was confirmed with a lab scale experiment simulating the leaching
of the trisiloxane surfactant after application on an agricultural field. Even in the worst case
scenario considered here (quartz sand as a test soil, heavy rainfall immediately after
application) the trisiloxane surfactant stayed in the first 15 cm of soil (after applying 51 L of
water i.e. almost 15 times the pore volume). The corresponding retardation factor (38 for the
homologue with n = 7) indicates that the trisiloxane surfactant is considerably retarded
compared to the water flow. The calculated sorption coefficients together with the results of
the lab scale simulation of leaching in soil indicate that the studied trisiloxane surfactant is
unlikely to reach ground water after application on agricultural fields.
Once adsorbed on soil, MD´(-(CH2)3-(EO)n-OCH3)M is probably degraded by hydrolysis of
the siloxane chain. At least two degradation products can be predicted (Michel et al., 2014):
trimethylsilanol and polyethersilandiol. Both predicted degradation products are more polar
and could eventually leach through soil. This interpretation would explain the observations of
Griessbach et al. (1998), who detected 14C-radiolabelled substance in leachate when studying
the leaching of 14C-radiolabelled trisiloxane surfactant through a soil column. Further research
on the environmental fate of trisiloxane surfactants should focus on degradation products,
especially the polar polyetherdisilanol.
Acknowledgments
This work was financially supported by the German Association for Gas and Water, a
registered technical-scientific association (Deutscher Verein des Gas- und Wasserfaches e.V.
DVGW), project W 2-01-10. We are grateful to CHT R. Beitlich for their cooperation and for
providing HANSA ADD 1055 for the soil column experiments. We thank Dr. Oliver Happel
for the picture of superspreading and Florian Storck for the fruitful discussions during the
elaboration of this manuscript.
Chapter 4
120
ReferencesBonnington, L.S., 2000. Analysis of organosilicone surfactants and their degradation products. PhD thesis, University of Waikato, Hamilton, New Zealand.
Cornejo, J., Jamet, P., and Mansour, M., 2000. Pesticide/soil interactions: some current research methods. Institut national de la recherche agronomique, Paris
Food and Agriculture Organization of the United Nations, 2000. Assessing soil contamination: a reference manual
Gerstl, Z., and Mingelgrin, U., 1984. Sorption of organic substances by soils and sediments. J. Environ. Sci. Health B 19, 297.
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze, P., 1997. Mobility of a siliconepolyether studied by adsorption and desorption isotherms and soil thin-layer chromatography. Sci. Total Environ. 201, 89.
Griessbach, E.F.C., Copin, A., Deleu, R., and Dreze, P., 1998. Mobility of a siliconepolyether studied by leaching experiments through disturbed and undisturbed soil columns. Sci. Total Environ. 221, 159.
Krogh, K.A., Halling-Sörensen, B., Mogensen, B.B., and Vejrup, K.V., 2003. Environmental properties and effects of nonionic surfactant adjuvants in pesticides: a review. Chemosphere 50, 871.
Knoche, M., Tamura, H., and Bukovac, M.J., 1991. Performance and stability of the organosilicone surfactant L-77: effects of pH, concentration, and temperature. J. Agric. Food Chem. 39, 202.
Laegdsmand, M., Moldrup, P., and De Jonge, L.W., 2007. Modelling of colloid leaching from unsaturated, aggregated soil. Eur. J. Soil Sci. 58, 692.
Lara-Martín, P.A., Gómez-Parra, A., and González-Mazo, E., 2008. Reactivity and fate of synthetic surfactants in aquatic environments. TrAC, Trends Anal. Chem. 27, 684.
Michel, A., Brauch, H.-J., Worch, E., and Lange, F.T., 2012. Development of a liquid chromatography tandem mass spectrometry method for trace analysis of trisiloxane surfactants in the aqueous environment: an alternative strategy for quantification of ethoxylated surfactants. J. Chromatogr. A 1245, 46.
Michel, A., Brauch, H.-J., Worch, E., and Lange, F.T., 2014. Homologue specific analysis of a polyether trisiloxane surfactant in German surface waters and study on its hydrolysis. Environ. Pollut. 186, 126.
Navarro, S., Vela, N., and Navarro, G., 2007. Review. An overview on the environmental behaviour of pesticide residues in soils. Span. J. Agric. Res. 5:357-375.Venzmer J (2011) Superspreading - 20 years of physicochemical research. Curr. Opin. Colloid Interface Sci. 16, 335.
Chapter 4
121
Organisation for Economic Cooperation and Development (OECD), 2000. OECD Guideline for the testing of chemicals, Test No. 106: Adsorption - Desorption Using a Batch Equilibrium Method.
http://www.oecd-ilibrary.org/docserver/download/9710601e.pdf?expires=1381145969&id=id&accname=guest&checksum=5B06EEC083852D5BA81BBA88A66257BD (Accessed October 2013)
Powell, D.E., and Carpenter, J.C., 1997. Polyethermethylsiloxanes, pp. 224, in: Organosilicon materials, Chandra, G. (Ed.) Springer, Berlin, Heidelberg.
Penner, D., Burow, R., and Roggenbuck, F.C., 1999. Use of organosilicone surfactants as agrichemical adjuvants, pp. 241, in: Silicone Surfactants, Surfactants science series, Vol 86, Hill R.M. (Ed.), Marcel Dekker Inc, New York.
Robertson, W.D., Van Stempvoort, D.R., Solomon, D.K., Homewood, J., Brown, S.J., Spoelstra, J., and Schiff, S.L., 2013. Persistence of artificial sweeteners in a 15-year-old septic system plume. J. Hydrol. 477, 43.
United State Department of Agriculture, Nature Resources Conservation Service, 2013. Soil survey manual.
Van Stempvoort, D.R., Roy, J.W., Brown, S.J., and Bickerton, G., 2011a. Artificial sweeteners as potential tracers in groundwater in urban environments. J. Hydrol. 401, 126.
Van Stempvoort, D.R., Robertson, W.D., and Brown, S.J., 2011b. Artificial Sweeteners in a large septic plume. Ground Water Monit. Rem. 31, 95.
Chapter 4
122
Appendix A
Table S 1: Limits of quantification obtained by direct injection on the HPLC-MS/MS system and after
liquid-liquid extraction.
n LOQs direct injection
in µg L-1
LOQs liquid-liquid extraction
in ng L-1
4 4 0.9
5 8 1.1
6 7 1.5
7 10 2.8
8 20 2.5
9 8 3.1
10 10 0.1
11 4 2.4
12 4 0.1
Table S 2: Parameters of the chromatographic separations performed during analysis of MD (́-(CH2)3-
(EO)n-OCH3)M and acesulfame by HPLC-MS/MS.
MD (́-(CH2)3-(EO)n-OCH3)M Acesulfame
Eluent A (for MD´(-(CH2)3-(EO)n-OCH3)M) Eluent A’(for acesulfame)
40 % of aqueous 10 mM ammonium acetate solution, 48 % ACN, 6 % THF, and 6 % MeOH
20 mM aqueous ammonium acetate solution
Eluent B (for MD´(-(CH2)3-(EO)n-OCH3)M) Eluent B’(for acesulfame)
THF 20 mM ammonium acetate solution in MeOH
Flow rate 800 µL min-1 800 µL min-1 Column PolymerX RP-1
4.6 mm × 250 mm , 5 µm, 100 Å from Phenomenex
Zorbax Eclipse XDB-C8, 4.6 × 150 mm, 5 µm, from Agilent
Gradient expressed as a percentage of eluent A
0 min : 90 % 2 min : 90 % 6 min : 25 % 12 min : 25 % 17 min: 90 %
0 min : 100 % 8 min : 100 % 24 min : 40 % 27 min : 40 % 32 min: 100 %
Chapter 4
123
Table S 3: Characteristics of the soils used in the sorption experiments.
Standard soil type # Soil 2.1 Soil 2.4 Soil 5M
Organic carbon in % C 0.66 ± 0.10 2.21 ± 0.46 0.95 ± 0.10
Nitrogen in % N 0.05 ± 0.01 0.23 ± 0.04 0.11 ± 0.02
pH value (0.01 M CaCl2) 5.2 ± 0.3 7.2 ± 0.2 7.3 ± 0.1
Cation exchange capacity (meq / 100 g) 4.1 ± 0.6 32.2 ± 4.4 17.1 ± 3.3
Particle size (mm) distribution according to German DIN (%)
<0.002 2.4 ± 0.4 26.5 ± 1.9 10.8 ± 1.2
0.002 – 0.006 1.6 ± 0.4 8.1 ± 1.0 4.4 ± 0.9
0.006 – 0.02 3.6 ± 0.4 14.8 ± 1.1 9.3 ± 1.0
0.02 – 0.063 7.0 ± 0.5 23.0 ± 1.0 21.8 ± 1.1
0.063 – 0.2 27.2 ± 0.5 19.0 ± 0.3 37.9 ± 1.5
0.2 – 0.63 55.7 ± 1.5 6.9 ± 2.2 14.5 ± 1.8
0.63 – 2.0 2.5 ± 0.4 1.7 ± 0.2 1.3 ± 0.2
Soil type Silty sand (uS) Clayey loam (tL) Loamy sand (IS)
Particle size (mm) distribution according to USDA (%)
<0.002 2.8 ± 1.0 26.2 ± 1.9 11.5 ± 1.0
0.002 – 0.05 10.5 ± 1.9 40.6 ± 0.9 30.7 ± 1.5
0.05 – 2.0 86.7 ± 1.3 33.2 ± 1.8 57.8 ± 1.4
Soil type Sand Loam Sandy Loam
Water holding capacity (g/100 g) 30.8 ± 1.9 43.8 ± 1.3 39.2 ± 2.8
Volumetric mass density (g/1000 mL) 1474 ± 28 1289 ± 32 1314 ± 68
Table S 4: Characteristics of the quartz sand used for the leaching in soil column experiments.
Volumetric mass density (kg m-³) 1500
Particle size (mm) distribution (%)
<0.100 1
0.100 - 0.315 23
0.315 - 0-6.30 74
>0.630 2
Chemical composition (%)
SiO2 97.4
Al2O3 1.35
K2O + Na2O 0.85
Loss on ignition 0.20
TiO2 0.07
CaO + MgO 0.05
Fe2O + TiO2 0.03
Chapter 4
124
Adsorption/desorption study: preliminary experiments
For the preliminary experiment, the indirect method was used, i.e. the amount of test
substance sorbed on soil is calculated as the difference between the amount of test substance
initially applied present and the amount remaining in the aqueous phase at time t. Hence the
percentage of MD´(-(CH2)3-(EO)n-OCH3)M sorbed on soil was calculated according to the
following equation:
soilon )MOCH-(EO)-)MD´(-(CHof% 3n32 0m
100(t)adssm
Where m0 is the mass of test substance at the beginning of the test and msads is the mass of test
substance sorbed on soil at time t. msads is calculated from the concentration of test substance
in the aqueous phase at time t (Caqads (t)) following the equation:
msads = m0 – Caq
ads (t) ×V0
Soil 2.1:
Figure S 1: MD (́-(CH2)3-(EO)n-OCH3)M sorption kinetics on soils 2.1, 2.4, and 5M at three soil/solution
ratios (1/10, 1/25, and 1/50) as determined during the preliminary experiments.
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
n=4n=5n=6n=7n=8 n=9n=10n=11n=12
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Soil 2.1, soil/solution ratio 1/10 Soil 2.1, soil/solution ratio 1/25
Soil 2.1, soil/solution ratio 1/50
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
Chapter 4
125
Soil 2.4:
Soil 5M:
Figure S 1 (continued): MD (́-(CH2)3-(EO)n-OCH3)M sorption kinetics on soils 2.1, 2.4, and 5M at three
soil/solution ratios (1/10, 1/25, and 1/50) as determined during the preliminary experiments.
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
Soil 2.4, soil/solution ratio 1/10 Soil 2.4, soil/solution ratio 1/25
Soil 5M, soil/solution ratio 1/10 Soil 5M, soil/solution ratio 1/25
Soil 5M, soil/solution ratio 1/50
Soil 2.4, soil/solution ratio 1/50
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
Time in h
0 10 20 30 40 50 60
% o
f M
D´(
EO
n-O
Me)
M i
n th
e so
il
0
20
40
60
80
100
% o
f M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M i
n th
e so
il
Chapter 4
126
As noted in the OECD guideline, the percentage of sorption should be above 20 % and
preferably above 50 % (Boesten, 1990; Organisation for Economic Co-operation and
Development, 2000), in order to minimize the relative error on the sorption coefficient. Based
on the results of the preliminary experiments, the soil/solution ratio 1/25 was selected. The
time required to reach the adsorption equilibrium was 24 h. For soil 2.1, no adsorption
equilibrium could be reach, probably due to the degradation of MD´(-(CH2)3-(EO)n-OCH3)M.
Chapter 4
127
Sorption study: Sorption kinetics
Figure S 2: Partition of MD (́-(CH2)3-(EO)n-OCH3)M between the aqueous phase and the soil and mass
balance for all homologues on soil 2.4.
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 5 n = 6
n = 9
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoil Mass balance
n = 8
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 10
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
%
of
init
iall
y ap
plie
d M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
n = 7
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 11 n = 12
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Chapter 4
128
Figure S 3: Partition of MD (́-(CH2)3-(EO)n-OCH3)M between the aqueous phase and the soil and mass
balance for all homologues on soil 5M.
n = 5
n = 7
n = 9
n = 11
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 6
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 8
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 10
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
n = 12
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
%
of
init
iall
y ap
plie
d M
D´(
-(C
H2)
3-(E
O) n
-OC
H3)
M
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Time in h
0 5 10 15 20 25
% o
f in
itia
lly
appl
ied
MD
´(E
On-
OM
e)M
0
20
40
60
80
100
120
140
160Aqueous phaseSoilMass balance
% o
f in
itia
lly
appl
ied
MD
´(-(
CH
2)3-
(EO
) n-O
CH
3)M
Chapter 4
129
Sorption isotherms:
Figure S 4: Sorption isotherms for all homologues of MD (́-(CH2)3-(EO)n-OCH3)M on soil 2.4.
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.001
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.001
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs i
n µ
g g-1
0.01
0.1
1
10
100
Sorption isotherm
n = 5 n = 6
n = 7 n = 8
n = 9 n = 10
n = 11 n = 12
Chapter 4
130
Figure S 5: Sorption isotherms for all homologues of MD (́-(CH2)3-(EO)n-OCH3)M on soil 5M.
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.001
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
Caq in µg cm-3
0.0001 0.001 0.01 0.1 1
Cs
in µ
g g-
1
0.01
0.1
1
10
Sorption isotherm
n = 5 n = 6
n = 7 n = 8
n = 9 n = 10
n = 11 n = 12
Chapter 4
131
Leaching in soil column:
Figure S 6: Acesulfame concentration and conductivity measured on side samples at 5 cm and 35 cm
depth during a preliminary experiment with a 70 cm long quartz sand column.
The conductivity is proportional to the concentration of CsCl. As represented on Figure S 6,
the increase of concentration of acesulfame is simultaneous with the increase of concentration
of CsCl.
Figure S 7: Acesulfame concentration in the side samples taken at 5 cm, 10 cm, and 15 cm.
Time in h
1 2 3 4 5 6 7
Concentr
ation o
f acesulfam
e in µ
g/L
0
100
200
300
400
Con
du
ctivity in
µS
/cm
0
1000
2000
3000
4000
5000
6000Acesulfame 5 cm
CsCl 5 cm
Acesulfame 35 cm
CsCl 35 cm
Time in min
0 50 100 150 200 250 300
Ace
sulf
ame
conc
entr
atio
n in
µg
L-1
0
200
400
600
800
1000
1200
5 cm10 cm15 cm
Ace
sulf
ame
conc
entr
atio
n in
µg
L-1
400
300
200
100
0
1 2 3 4 5 6 7
1000
2000
3000
4000
5000
6000
0
Con
duct
ivit
y in
µS
cm
-1
Time in h
Acesulfame 5 cm CsCl 5 cm Acesulfame 35 cm CsCl 35 cm
Chapter 4
132
References
Boesten JJTI (1990) Influence of solid/liquid ratio on the experimental for of sorption coefficients in pesticide/soil systems. Pestic. Sci. 30,31.
Organisation for Economic Co-operation and Development (OECD), 2000. OECD Guideline for the testing of chemicals: Adsorption - Desorption Using a Batch Equilibrium Method. http://www.oecd-ilibrary.org/docserver/download/9710601e.pdf?expires=1381145969&id=id&accname=guest&checksum=5B06EEC083852D5BA81BBA88A66257BD (Access October 2013)
Summary and conclusion
133
Summary and conclusion
Summary and conclusion
134
The current work provides new knowledge on the environmental occurrence and fate of
silicone surfactants.
As a prerequisite to any further research, an analytical method for the trace analysis of
trisiloxane surfactants in aqueous environment was developed and validated. The method,
based on liquid-liquid extraction and HPLC-MS/MS reaches limits of quantification in the
ng L-1 range and recoveries higher than 80 %. The main challenge to overcome during the
development of this new method was the lack of a proper analytical standard for trisiloxane
surfactants. Each trisiloxane surfactant is a mixture of homologues, having different unknown
concentrations. Based on the mathematical model of a Poisson distribution, MS spectrum and
NMR measurements, the oligomeric composition of the standard was characterized and the
obtained standard was used to quantify every homologue individually. This new approach is
reliable, simple, avoids a time-consuming synthesis of individual homologues, avoids wrong
estimations of the concentrations, and allows obtaining information on the shape of the
oligomeric distribution in a sample. Because of its simplicity and reliability, this approach
could be directly used for the quantification of other ethoxylated surfactants.
The second part of the work focuses on trisiloxane surfactants in surface waters. The newly
developed analytical method was applied to river water samples and one trisiloxane surfactant
was detected around 50 ng L-1 in the Neckar River. The results showed that the studied
trisiloxane surfactant does not ubiquitously occur in the aquatic environment in measurable
concentrations but can reach surface waters on a local scale. Based on their chemical
structures, trisiloxane surfactants are generally considered to be sensitive to hydrolysis.
Hydrolysis is therefore thought to be an important elimination mechanism of trisiloxane
surfactants from the aquatic environment. This assumption was tested by studying the
hydrolysis of one trisiloxane surfactant in lab conditions at different temperatures, pH, and
concentrations. The half-lifes at pH 7 and 2 mg L-1 were found to be between 29 days and
55 days at 25°C and between 151 days and 289 days at 12°C. If one takes only into account
the hydrolysis (abiotic degradation), the obtained results indicate that the trisiloxane
surfactant could persist several weeks in surface waters.
The comparison of the measured environmental concentrations and the NOEC values of one
trisiloxane surfactant indicates that at these concentrations, the trisiloxane surfactant does not
Summary and conclusion
135
represent an acute toxicological risk for the aquatic environment. The No Observed Effect
Concentration (NOEC) has been found at 0.56 mg L-1 for Zebra Fish (96 h exposure time) and
at 3.2 mg L-1 for Rainbow Trout (96 h exposure time); regarding the toxicity to algae, the
NOEC to Selenastrum capricornutum was determined at 1 mg L-1 (96 h exposure); and the
NOECs were 10 mg L-1 and 25 mg L-1 for Daphnia similis and Daphnia magna, respectively
(both 48 h exposure). However no data are available on the chronic toxicity of the trisiloxane
surfactant to aquatic organisms.
The third part of the work addresses the behavior of one trisiloxane surfactant on soil. With a
sorption batch equilibrium method, distribution coefficients between water and soil (Kd, Koc,
and Kclay) were estimated for two standard soils (loam and sandy loam) and for every
homologue of the trisiloxane surfactant. The obtained values for Kd were between 15 L kg-1
and 135 L kg-1, indicating that the trisiloxane surfactant is only slightly mobile in soil. To
further investigate the possibility of leaching to ground water after application on agricultural
fields, the leaching in soil was simulated in the lab in a soil column. The experimental settings
were designed to correspond to a worst case scenario where the application of the trisiloxane
surfactant is done on quartz sand and is immediately followed by a heavy rainfall. Even in
these conditions, less than 0.01 % of the initially applied trisiloxane surfactant has leached
through 20 cm of quartz sand. Based on the Kd values and on the results of the leaching in the
soil column, the studied trisiloxane surfactant is considered to be unlikely to leach to ground
water after application as an agricultural adjuvant.
Overall, the current work answered the first key questions regarding trisiloxane surfactants in
the environment: their occurrence in surface water, their degradation by hydrolysis, their
behaviour on soil and the possibility to leach to ground water.
Several further research needs for trisiloxane surfactants could be identified.
To gain better knowledge on their persistence in the aquatic environment, it would be
necessary to study their biodegradation. The study of the hydrolysis reported here was carried
out under sterile conditions, in order to differentiate chemical and biological degradation.
However a degradation study including biodegradation would be more representative of
environmental conditions.
Summary and conclusion
136
Regarding the behavior on soil, another important point to assess would be the influence of
the trisiloxane surfactant on the mobility of other substances, especially pesticides. When
applied together with plant protection products, surfactants have been found to modify the
sorption of the active substances, insecticide, herbicide or fungicide (Hua et al., 2008; Wang,
et al., 2009; Abu-Zreig et al., 1999, Krogh et al., 2003). Depending on the type of soil and the
physico-chemical properties of the substances involved (surfactant and pesticide), the
mobility of the pesticide has been found to be enhanced or hampered. Since the trisiloxane
surfactant have more “powerful” surface active properties that traditional carbon-based
surfactants, one can expect a stronger influence on the mobility of pesticides. Further research
should be dedicated to this issue.
The hydrolysis of the studied trsiloxane surfactant produces more polar degradation products.
No data are available on the environmental fate of those degradation products. Given the
broad range of application of trisiloxane surfactants, from pesticides to personal care
products, their degradation products are expected to be widely found in the environment.
More research would be necessary to understand the environmental fate of these compounds.
Organosilicon compounds represent a very important category of chemicals with high volume
of production and diverse applications. Very little information is however available regarding
the overall amount of silicones in the environment. The work carried out by Pellenbarg
between 1979 and 1997, suggests the long term persistence of siloxanes in sediments
(Pellenbarg, 1979a; Pellenbarg, 1979b; Pellenbarg, 1982; Pellenbarg and Tevault, 1986;
Pellenbarg, 1988; Pellenbarg et al., 1997). By analyzing the extractable organic silicone
content of sediments and comparing it to 210Pb dating, the evidence of a siloxane horizon
could be shown, with a clear appearance of silicone in the sediments from the 1950’ s, which
corresponds to the commercial introduction of silicones. The results suggest the persistence of
siloxanes on sediments over several decades and lead the authors to propose siloxanes as a
tracer of anthropogenic activities. The high number of different forms of silicones makes their
individual analysis tedious and a surrogate parameter would be required. Several such
surrogate parameters already exist: the total organic carbon (TOC), the adsorbable organic
halogens (AOX), or more recently the adsorbable organic fluorine (AOF) (Wagner et al.,
2013) for example. Such a method for silicones would provide valuable information on the
overall amount of silicones in the environment and would allow continuing further the work
from Pellenbarg on silicones as tracers of anthropogenic contamination.
Summary and conclusion
137
ReferencesAbu-Zreig, M., Rudra, R.P., Dickinson, W.T., Evans, L.J., 1999. Effects of surfactants on
sorption of atrazine by soil, J. Contam. Hydrol., 36, 249.
Hua, R., Spliid, N.H., Heinrichson, K., Laursen, B., 2009, Influence of surfactants on the
leaching of bentazone in sandy loam soil, Pest Manag. Sci., 65, 857.
Krogh, K.A., Halling-Sørensen, B., Mogensen, B.B., Vejrup, K.V., 2003. Environmental
properties and effects of nonionic surfactants adjuvants in pesticides: a review, Chemosphere,
50, 871.
Pellenbarg, R., 1979a. Silicones as tracers for anthropogenic additions to sediments. Mar.
Pollut. Bull., 10, 267.
Pellenbarg, R., 1979b. Environmental poly(organosiloxanes) (silicones), Environ. Sci.
Technol., 13, 565.
Pellenbarg, R.E., 1982. Silicones in Chesapeake bay sediments. Mar. Pollut. Bull., 13, 427.
Pellenbarg, R.E. and Tevault, D.E., 1986. Evidence for a sedimentary siloxane horizon.
Environ. Sci. Technol., 20, 743.
Pellenbarg, R.E., 1988. Historical record of silicones in sediment, Sci. Total Environ., 73, 11.
Pellenbarg, R.E., DeCarlo, E.C., Boyle, M.E., and Lamontagne, R.A., 1997. Sedimentary
siloxanes: a geochronological study. Appl. Organomet. Chem., 11, 345.
Wagner, A., Raue, B., Brauch, H.J., Worch, E., Lange, F.T., 2013. Determination of
adsorbable organic fluorine from aqueous environmental samples by adsorption to
polystyrene-divinylbenzene based activated carbon and combustion ion chromatography, J.
Chromatogr. A, 1295, 82.
Wang, P., Keller, A.A., 2009. Partitioning of hydrophobic pesticides within a soil-water-
anionic surfactant system, Water res., 43, 706.
