Role of Molar Concentration of Activators in Developing Engineered Geopolymer Cement

Pastes using Natural Pozzolanic Volcanic Ash: Microstructure and Chemical Analysis

Ben Goldsmith

Nyack High School

Goldsmith 1

Acknowledgements

I would like to thank my mentor Dr. Kunal Kupwade Patil from MIT, my teachers Ms.

Kleinman and Ms. Foisy, my parents, and my science research peers for assisting me with this

research.

Table of Contents

1. Introduction - page 2

2. Statement of Purpose

3. Methods and Materials

a. 3.1 Raw Material Analysis

b. 3.2 Methods

c. 3.2.1 Mixing

d. 3.3.1 X-Ray Diffraction

e. 3.3.2 27Al NMR Analysis

4. Results

a. 4.2 27Al NMR Analysis

5. Discussion and Conclusion

6. References

Goldsmith 2

List of Figures and Tables

● Figure 1 - page 5

● Figure 2 - page 6

● Table 1 - page 6

● Table 2 - page 7

● Figure 3 - page 9

● Figure 4 - page 10

● Figure 5 - page 12

● Figure 6 - page 13

● Figure 7 - page 14

● Figure 8 - page 15

● Table 3 - page 16

Goldsmith 3

Abstract

A microstructure and chemistry based study was performed on geopolymer cement paste

prepared with natural pozzolanic volcanic ash. The role of activator influencing the resulting

products was examined using 27Al Magic Angle Spinning (MAS) Nuclear Magnetic Resonance

Spectroscopy and powder X-Ray diffraction. Variation in zeolitic products was observed when

the activator concentration was increased from 10 to 14 M NaOH. Calcium

(alkali)-Sodium-Alumino-Silicate-Hydrate (C,N)-A-S-H gel was the predominant product that

was formed due to sodium based alkali activation. 27Al NMR analysis revealed the presence of

unreacted volcanic ash when geopolymer cement paste was prepared with 10M sodium

hydroxide. The deconvolutions analysis suggested high levels of cross-linking in the

(Calcium/Sodium)-Alumino-Silicate-Hydrate (C,N)-A-S-H gel. This study clearly shows the

careful consideration needs to be given in selecting the activators before preparing geopolymer

cement with natural volcanic ash.

Goldsmith 4

1.Introduction

Concrete is the second most consumed substance on earth after water 1. While the

pouring of concrete is necessary to build infrastructure in the modern era, it has dire effects on

our climate. For every ton of Portland cement, the major ingredient in concrete, produced, one

ton of CO2 is emitted into Earth’s atmosphere 2. The manufacturing of cement alone accounts for

approximately 6% of all greenhouse gas emissions, including carbon dioxide, methane,

chlorofluorocarbons, water vapor, and others. The production of this key ingredient in concrete is

wasteful because it must be produced in a kiln that is at least 1400oC. It takes about 400 lbs. of

coal to produce enough heat energy to make just one ton of cement 2. In addition to the emissions

from heat energy production, almost half of the total Green House Gas (GHG) emissions from

the cement industry are actually from limestone calcination, which is a chemical process that

occurs when limestone is heated. The calcination process breaks limestone down into calcium

oxide, as well as CO2 which is released into the atmosphere. Furthermore, the other ingredients

in concrete are gravel, sand, and water, which also take energy to gather, manufacture, and

transport.

Due to both the wastefulness of production and the scale at which it is being produced, it

makes sense to look into reducing the environmental impact of producing concrete. One solution

to this problem is the production of an alternative, which is geopolymer concrete. Geopolymer

concrete or alkali activated binder is an emerging technology, which has exhibited a wide range

of advantages such as high corrosion and temperature resistance, reduced CO2 emission

construction material, a minimum energy-consuming binder and also negligible carbon footprint

3, 4, 5, 6, 7. Davidovits and co-workers coined the term ‘‘geopolymers’’ to classify polymeric

Goldsmith 5

networks of synthetic aluminosilicate binders also known as inorganic polymers 8, 9. The

polymerization mechanism also involves a reaction of alumino-silicate minerals in the source

material with the alkali activators such as alkaline solution along with soluble silicates.

One massive advantage of geopolymer concretes in general is that they are normally

more durable than typical Portland cement concretes4. The increased durability can be owed to

the pozzolanic properties of most ashes, meaning the material is not cementitious by itself, but

gains cementitious properties when exposed to certain activators including sodium hydroxide.

Pozzolanic materials can do this because the presence of water allows sodium hydroxide

to act as a bonding agent for the pozzolanic particles. It is important to note, however, that only

acidic volcanic ashes have shown to have pozzolanic properties. This means that most volcanic

ashes found in the eastern hemisphere will not work in geopolymer cement, because they don’t

have pozzolanic properties10, 11, 12. The pozzolanic properties of the volcanic ash helps create

strong chemical bonds in the geopolymer concrete. This advantage to geopolymer concrete is

essential to its success, because in order to displace Portland cement production in the future,

viable solutions such as geopolymer concretes will need to contain enough benefits over typical

concrete to encourage their usage. Volcanic ash is a resource naturally found in many areas on

Earth, including areas with no active volcanoes, thus lessening the need for the transport of

materials. One goal of this study is to use volcanic ash to make concrete. If volcanic ash

geopolymer concrete is utilized, less Portland cement will have to be produced, and global

greenhouse gas emissions will consequently be reduced as well.

Goldsmith 6

2. Statement of Purpose

Currently, fly ash geopolymer concrete is used more commonly than volcanic ash concrete,

particularly because it’s already considered a “green” concrete. However the environmental

advantage of volcanic ash concrete is that it does not rely upon coal burning for its production.

Fly ash will eventually stop being produced once the world transitions from fossil-fuel energy to

renewable energy. As a result, volcanic ash geopolymer concrete is a superior long-term solution

compared to fly ash concretes. In a recent study, researchers found evidence that supported that

volcanic ash could, in fact, be used as a viable substitute for fly ash in concrete 10. The study

looked into the micro and pore structural developments at different stages of curing. The

researchers determined that volcanic ash could be an effective substitute for fly ash based on data

that showed that the addition of volcanic ash did not cause any fatal flaws in the concrete.

The current study looks into the chemical and microstructure effects of molar

concentration of the activator on volcanic ash based geopolymer cement paste (GCP) (Refer to

Figure 1). Advanced characterization tools such as Magic Angle Spinning (MAS), Nuclear

Magnetic Resonance (NMR), and X-Ray Diffraction (XRD) were used to examine the

microstructure and chemistry of the resulting structure. The molecular structure of the resulting

volcanic ash geopolymer cement is extremely important because it determines characteristics

such as the strength, as well as corrosion and wear resistance 3.

Goldsmith 7

Fig 1. Multi-scale scale experimental approach analyzing geopolymer cements using Powder

X-Ray Diffraction and 27Al NMR (Concept adapted from Ashbrook et al, 2014 13)

3. Materials and Methods

3.1 Raw Material Analysis Regular ground volcanic ash (IP) originating from the Kingdom of

Saudi Arabia (KSA) was procured from Pozzolan Product Factory located in Jeddah, KSA.

Particle size distribution (PSD) was conducted at Micromeritics, Norcross GA on volcanic ash

and OPC by suspending them in isopropyl alcohol using a Laser Light Scattering technique with

a Micromeritics Saturn DigiSizer 5205.The particle size distributions of volcanic ash as received

from the manufacturer is shown in Figure 2. The average means, medians, and modes of the

diameters and the diameters at select volume percentiles for these three materials are shown in

Table 1.

Goldsmith 8

Table 1. Particle Size Analysis with two different particle sizes

Binder Type Mea

n

(μm)

Media

n

(μm)

Mod

e

(μm)

Diameter for selected percentiles by

volume

D 90

(μm)

D 50

(μm)

D10

(μm)

Volcanic

Ash

17.1

4

10.00 13.27 42.46 10.00 1.50

X-Ray Fluorescence (XRF) spectroscopy was performed using an Energy Dispersive

X-Ray Fluorescence (EDXRF) spectrometer; an electrically-cooled silicon drift detector with a

30mm x 1mm active area, a 8.8 mm 50KV X-Ray beam collimator, and a 50W Rhodium target

X-Ray tube. The XRF analysis provided the mass percent distribution of oxides (Table 2). The

net proportion of silicon oxide (SiO2), aluminum oxide (Al2O3), and ferric oxide (Fe2O3)

Goldsmith 9

components for the raw volcanic ash is 74.3%, indicating that this material is a Class C type of

ash according to ASTM C618 14.

Table 2. X-ray Fluorescence (XRF) Spectroscopy of Volcanic Ash

Binder

Type

Mass % as Oxide

CaO SiO2 Al2O3 MgO SO3 TiO2 K2O Fe2O3 Na2

O

P2O5

Volcanic

Ash

9.29 47.0 14.80 7.95 - 2.41 1.40 12.5 3.54 0.61

3.2 Methods

3.2.1 Mixing

GCP was prepared using volcanic ash and was alkali activated by sodium silicate and

sodium hydroxide solution. An activator/binder ratio of 0.55 was maintained throughout the

experiment. The activator solution consisting of sodium hydroxide: sodium silicate was

maintained 1:1, while the volcanic ash was the binder. A 10 and 14 M sodium hydroxide and

sodium silicate manufactured by P Q corporation (45% by weight and SiO2 to Na2O ratio of 2:1)

was used in the preparation of the geopolymer paste. The GCP was prepared by addition of the

sodium silicate to the fly ash and mixing it for 1 min. The mixture was then alkali activated by

Goldsmith 10

adding the sodium hydroxide. Cylindrical cement paste specimens, 10.16 cm height by 5.08 cm

diameter, were cast and cured at 80℃ for 48 h.

3.3.1 X-Ray Diffraction

X-ray diffraction (XRD) was conducted using PANalytical X'Pert PRO using a Cu Kα

radiation nickel foil filter. The ground sample was placed on Bragg-Brentano geometry optics on

a flat plate sample geometry. Fixed divergence slit of 1/16 degree was chosen to limit beam

overflow on the samples at small angles of 2θ. Incident and diffracted-side soller slits of 0.02

rad were used for this experiment. The X'Celerator high-speed linear detector was used for XRD

experiment with an active length 2.122 degrees 2θ, a step size of 0.0167113 degrees 2θ and scan

range from 4 to 90 degrees.

3.3.2 27 Al NMR Analysis

Solid-state 27Al MAS NMR spectra were collected at 94.7 MHz on a Tecmag Redstone

363 MHz NMR spectrometer using a Doty Scientific probe for 5 mm outer diameter zirconia

rotors and a spinning speed of 10.0 kHz. The 27Al MAS experiments (1-pulse Block decay)

employed a pulse width of 1 s, a relaxation delay of 0.3 s and 7,500 to 10,000 scans. All spectra

were collected with a pulse angle of 45° (solids).

4. Results

3.1 X-Ray Diffraction

(XRD) XRD analysis of GCP prepared with 10 and 14 M samples are shown in Figure 3 and 4,

respectively. The XRD phases revealed some changes in mineralogy and the GCP with 10M

specimen showed the presence of Gismondine-Ca which is related to the zeolitic phase and

Goldsmith 11

attributed to Calcium alumino silicate hydrate (C-A-S-H). The Carncrinite phase belongs to the

family of Sodium alumino silicate hydrate (N-A-S-H)15. The geopolymers which are activated

with sodium show the presence of crystalline and amorphous N-A-S-H gels. The N-A-S-H gel is

the binding glue for geopolymers. Since the volcanic ash was Class C type and had higher

concentration of CaO therefore it resulted in Crystalline C-A-S-H gel phase of Gismondine. In

addition to zeolitic phases, Ettringite and Cristobalite was observed among 10 M GCP specimen.

Fig 3. X-Ray Diffraction Spectrum for GCP prepared with 10M activator

The XRD analysis of 14M GCP specimen revealed slightly different phases, since the

sodium hydroxide concentration was increased from 10 to 14M concentration. Faujasite-Na and

Heulandite-Ca were detected which are attributed to the N-A-S-H and C-A-S-H gel phases,

Goldsmith 12

respectively. Furthermore, Hibschite belong to the Hydrogarnett class of family and was formed

due to presence of Al content in the volcanic ash. Carbonation was also detected with the

presence of Aragonite.

Fig 4. X-Ray Diffraction Spectrum for GCP prepared with 14M sample

XRD analysis clearly indicates that by increasing the concentration of activator, different

types of crystalline N-A-S-H and C-A-S-H gels are formed. Furthermore, it clearly suggests that

higher alkali uptake and higher calcium concentration of the volcanic ash lead to the formation

of these gels. The volcanic ash contained intermediate to high levels calcium, sodium and

alumina and the mixture was activated using Na based activator which lead to the formation of

Calcium (alkali) aluminosilicate hydrate (C-(N-)-A-S-H gels4. The “N” reflects the activators

Goldsmith 13

used in the alkali activated system, in this study a combination of sodium hydroxide and sodium

silicate were used as activators. Even still, a challenge remains in understanding the interaction

of these products and the stresses which are developed due to this interaction of these products.

Additional studies are required using small angle neutron/X-ray scattering to examine the

evolution of structure of these synthetic gels. Additionally, more studies are needed to determine

the significance and implications of the presence of certain crystalline N-A-S-H and C-A-S-H

gels.

4.2 27Al NMR Analysis

27Al NMR analysis is a powerful technique that can be used to analyze the amorphous

and crystal structure of geopolymeric gels. NMR analysis not only provides structure, but also

gives detailed insight into the interactions and dynamics, which can predict the rate of the

reaction 13 (see Figure 5). The testing offers valuable data on both the physical and chemical

components of a material. For advanced understanding of NMR analysis of alumino silicate

materials the readers are referred to texts by Engelhardt and Michel, 198716 and Mackenzie and

Smith, 200217. For basic understanding, however, it is important to note that a Nuclear Magnetic

Resonance machine works by relying on the magnetic properties of the nuclei of some atoms. By

reading the resonance frequency of a material, the machine can help determine the physical and

chemical makeup of a material with accuracy 16.

Goldsmith 14

Fig 5. Nuclear Magnetic Resonance (NMR) Applications (Adapted with permission from

Ashbrook et al. 2014, 13)

27Al NMR spectra for GCP paste prepared with 10M and 14M is shown in Figure 6.

Three distinct Al environments (AlIV, AlV and AlVI) are observed in the 27Al MAS NMR spectra

for chemical shift (δ) = 50–80 ppm (i.e., the observed chemical shift), 30–40 ppm and 0–20 ppm,

respectively16. The Al in the hydrated phase is exclusively in six-fold coordination, making 27Al

NMR convenient for analyzing the effect of hydration when pure alumina phases are present in

cements. The volcanic ash used in the current study had higher trace of Al2O3 content (14.80 %)

which lead to the formation of alumina rich phases. The 10M GCP specimen showed resonance

peak at 52.28 ppm indicating the presence of unreacted volcanic ash which was not involved in

the geopolymerization process. The peak can be assigned to the tetrahedral arrangement of

aluminum (AlT). For the 14M GCP specimen, a shift in resonance to 54.32 ppm was detected

when compared to 10M samples (52.28 ppm). A peak shift of ~2 ppm resonance can be

attributed the degree of geopolymerization indicating higher the chemical shift greater the

dissolution process of the volcanic ash which leads to the formation of the geopolymer network.

Goldsmith 15

Fig 6. 27Al Magic Angle Spinning (MAS), Nuclear Magnetic Resonance (NMR) analysis for

GCP prepared with 10 and 14 M.

Base spectrum and the deconvoluted spectra for GCP prepared with 10M and 14 M are

shown in Figures 7 and 8. A broad spectrum between 50 to 80 ppm was detected on both the

specimens. The peak broadening has been associated with quadrupolar interactions caused by the

tetrahedral distortion that was influenced by the changes in viscosity during the curing process18.

The deconvolution analysis for peak value, percent area, fraction lorentzian and peak assignment

is shown in Table 3. The resonances in the range of 50-80 ppm are associated to

Goldsmith 16

tetra-coordinated AlIV atoms, whereas the resonances in the range of 30-40 ppm are related to the

penta-coordinated AlV atoms.

Fig 7. (A) Base spectrum of 27Al Geopolymer paste prepared with 10M concentration (B)

Deconvoluted spectrum analysis for geopolymer paste prepared with 10 M concentration

Goldsmith 17

Fig 8. (A) Base spectrum of 27Al Geopolymer paste prepared with 10M concentration (B)

Deconvoluted spectrum analysis for geopolymer paste prepared with 10 M concentration

Goldsmith 18

Table 3. 27Al NMR Deconvolution analysis for Geopolymer paste prepared with 10M and

14M samples.

Sample Type Peak value (ppm) Fraction

Lorentzian

%

Area

Assignment

10M-GPC 50.976 0.255 60.7 Al IV

37.409 0 39.3 Al V

14M-GPC 52.124 0.105 60.65 Al IV

40.104 0 39.35 Al V

The resonances related to ~30 ppm and at 60.7 ppm have been related to the zeolitic

phase of faujasite ((Na2,Ca,Mg)3.5[Al7Si17O48]·32(H2O) 17, 19 (Refer to Figure 6). Faujasite is

commonly observed in sodium hydroxide activated geopolymers. In the current study, volcanic

ash, which was rich in Ca, Mg, and Na, was used as an aluminosilicate and was activated with

sodium hydroxide. The crystalline form of sodium alumino-silicate hydrate (N-A-S-H) gel is

usually assigned to the faujasite phase of the zeolitic mineral. Faujasite was detected in these

samples since the samples were cured at 80oC and commonly this phase appears during the early

stage of crystallization. Studies20 show that when geopolymeric gel is formed at elevated

temperature, thermodynamically stable crystalline structures are produced such as sodalite,

faujasite and gismondine that were observed in the XRD analysis.

Furthermore, resonance around ~60.00 ppm have been associated to the formation of

sodalite phase, where the Al becomes confined inside the sodalite cages in the geopolymeric

structure, where it remains in tetrahedral coordination21. Additionally, the deconvolutions

Goldsmith 19

significant at δ = 60 ppm are attributed to the coordinated Al, suggesting high levels of

cross-linking in the (Calcium/Sodium)-Alumino-Silicate-Hydrate (C,N)-A-S-H gel22. This

suggests an increase in the molar concentration of the activator causes chemical shift forming

higher AlIV environments. Further analysis is required to find the amount of concentration of

(C,N)-A-S-H gel formed when pozzolanic volcanic ash is used as a precursor. Experiments using

tri-calcium silicate C3S and volcanic ash are required to find the structure by linking small angle

neutron/x-ray (SANS/SAXS) scattering to NMR studies.

5. Discussion and Conclusion

A laboratory scale study was performed to examine the role of activators in preparation

of GCP using natural pozzolanic volcanic ash. 10 and 14 M concentration of sodium based

activator was used to prepare geopolymer cement and micro and chemical analysis was

performed by powder X-Ray diffraction and 27Al NMR. Variation in zeolitic products was

observed when the activator concentration was increased from 10 to 14 M concentration.

(C,N)-A-S-H gel was the predominant product that was formed due to sodium based alkali

activation. 27Al MAS NMR revealed the presence of tetra-coordinated AlIV atoms. The difference

in the chemical shift revealed that unreacted volcanic ash was present in 10M activated samples.

The deconvolution analysis showed high levels of cross-linking in the

(Calcium/Sodium)-Alumino-Silicate-Hydrate (C,N)-A-S-H gel. This suggests an increase in

molar concentration of the activator which slightly influences the chemical shift, causing higher

AlIV environments to form. Additionally, discovered unreacted volcanic ash in the 10M sample

indicates that 10M is not an ideal molarity. This is because there should not be any unreacted

Goldsmith 20

material in a geopolymer concrete. Overall, this study clearly shows that careful consideration

needs to be given to selecting the activators before preparing geopolymer cement using volcanic

ash. The drastic difference in physical and chemical makeup between the 10M and 14M

concentrations shows that the molarity of the activators can have an effect on not just the amount

of bonds formed, but the types of bonds formed as well. Rather than a traditional strength

approach, a detailed microstructure based approach needs to be adopted in the future in order to

study the complex and heterogeneous nature of the geopolymer binders.

Disclaimer:

Massachusetts Institute of Technology owns the copyright and the intellectual property rights for

this work.

Goldsmith 21

5. References

1. P. C. Aïtcin and S. Mindess, "Sustainability of Concrete." Taylor & Francis, (2011).

2. M. Rubenstein, "Emissions from the cement industry," The Earth Institute of Columbia

University (2012).

3. P. Duxson, A. Fernández-Jiménez, J. Provis, G. Lukey, A. Palomo, and J. van Deventer,

"Geopolymer technology: the current state of the art," J Mater Sci, 42[9] 2917-33 (2007).

4. S. A. Bernal and J. L. Provis, "Durability of Alkali-Activated Materials: Progress and

Perspectives," Journal of the American Ceramic Society, 97[4] 997-1008 (2014).

5. P. J. and J. S. J. Davidovits, "Alkali Activated Materials: State-of-the-art Report, Rilem TC

224-AAM," pp. 388. in. Springer/RILEM, Dordrecht, 2014.

6. K. Kupwade-Patil and E. Allouche, "Examination of Chloride-Induced Corrosion in

Reinforced Geopolymer Concretes," Journal of Materials in Civil Engineering, 25[10]

1465-76 (2013).

7. M. Sufian Badar, K. Kupwade-Patil, S. A. Bernal, J. L. Provis, and E. N. Allouche, "Corrosion

of steel bars induced by accelerated carbonation in low and high calcium fly ash

geopolymer concretes," Construction and Building Materials, 61[0] 79-89 (2014).

8. J. Davidovits, "Synthetic Mineral Polymer Compound of the Silicoaluminates." in., 1984.

9. J. Davidovits, "Geopolymer Chemistry and Applications." Institut Géopolymère, (2011).

10. K. Kupwade-Patil, A. F. Al-Aibani, M. F. Abdulsalam, C. Mao, A. Bumajdad, S. D.

Palkovic, and O. Büyüköztürk, "Microstructure of cement paste with natural pozzolanic

volcanic ash and Portland cement at different stages of curing," Construction and

Building Materials, 113 423-41 (2016).

Goldsmith 22

11. E. Gotti, J. Oleson, L. Bottalico, C. Brandon, R. Cucitore, and R. Hohlfelder, "A comparison

of the chemical and engineering characteristics of ancient roman hydraulic concrete with

a modern reproduction of vitruvian hydraulic concrete," Archaeometry, 50[4] 576-90

(2008).

12. M. Jackson, G. Vola, D. Všianský, J. Oleson, B. Scheetz, C. Brandon, and R. Hohlfelder,

"Cement Microstructures and Durability in Ancient Roman Seawater Concretes," pp.

49-76. in Historic Mortars, Vol. 7. RILEM Bookseries. Edited by J. Válek, J. J. Hughes,

and C. J. W. P. Groot. Springer Netherlands, 2012.

13. S. E. Ashbrook, D. M. Dawson, and V. R. Seymour, "Recent developments in solid-state

NMR spectroscopy of crystalline microporous materials," Physical Chemistry Chemical

Physics, 16[18] 8223-42 (2014).

14. ASTM, "ASTM C 618 Standard Specification for Coal Fly Ash and Raw or Calcined Natural

Pozzolan for Use as a Mineral Admixture in Concrete," Annual Book of ASTM Standard,

Section, 4 (2004).

15. J. Provis, "Geopolymers and other alkali activated materials: why, how, and what?," Mater

Struct, 47[1-2] 11-25 (2014).

16. G. Engelhardt and D. Michel, "High-Resolution Solid-State Nmr of Silicates and Zeolites."

John Wiley & Sons, Limited, (1987).

17. K. J. D. MacKenzie and E. Smith, "Multinuclear Solid-State Nuclear Magnetic Resonance of

Inorganic Materials." Elsevier Science, (2002).

18. R. E. Wasylishen, S. E. Ashbrook, and S. Wimperis, "NMR of Quadrupolar Nuclei in Solid

Materials." Wiley, (2012).

Goldsmith 23

19. D. Freude and H. Behrens, "Investigation of 27Al NMR chemical shifts in zeolites of the

faujasite type," Crystal Research and Technology, 16[3] K36-K38 (1981).

20. J. L. Provis and J. S. J. Van Deventer, "Geopolymers: Structure, Processing, Properties and

Industrial Applications." Woodhead Publishing Limited, (2009).

21. P. P. Man and J. Klinowski, "Quadrupole nutation 27Al NMR studies of isomorphous

substitution of aluminium in the framework of zeolite Y," Chemical Physics Letters,

147[6] 581-84 (1988).

22. R. J. Myers, S. A. Bernal, J. D. Gehman, J. S. J. van Deventer, and J. L. Provis, "The Role of

Al in Cross-Linking of Alkali-Activated Slag Cements," Journal of the American

Ceramic Society, 98[3] 996-1004 (2015).