GEOPOLYMERS: MORE THAN JUST A CEMENT W. M. Kriven1, J. Bell1, M. Gordon1 and Gianguo Wen2 1Department of Materials Science and Engineering 2Center for Microanalysis of Materials, Frederick Seitz Materials Research Laboratory University of Illinois at Urbana-Champaign Urbana, IL 61801, USA. ABSTRACT

Geopolymers (GPs) may be thought of as a type of chemically bonded ceramic, more specifically, as “alkali-bonded ceramics” or “ABCs”. They are a class of cementitious materials that do not depend on the presence of calcium. Geopolymers are amorphous, gel-like ceramics that are formed by the mixing of alkali-silicate solutions with aluminosilicate minerals or glasses. Two processing routes for the thermal conversion of potassium-based geopolymers into ceramics were carried out for this study. A low water content GP, having a molar composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O and reinforced with amorphous 500 nm silica spheres, was thermally converted into crystalline leucite of the same dehydrated composition, by heating above 900oC. Upon thermal conversion the dehydrated GP exhibited macroscopic cracking. The microstructure consisted of a glazed surface covered with spherical voids, caused by entrapped air. Ceramic powders derived by crushing GPs having a molar composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O were die-pressed into pellets and sintered at 1200oC into leucite. The resulting pellets exhibited no cracking upon conversion into leucite. SEM analysis also showed a similar glazed surface. The interior of the sample was compromised of ~3 µm size phases, which according to XRD were leucite grains, dispersed in an intergranular, amorphous phase. TEM of the amorphous matrix phase showed a stable amorphous phase having a remnant, typical geopolymer, microstructural texture. Selected area diffraction patterns of the continuous matrix phase suggested that it was generally amorphous, with numerous incipient nanocrystals of leucite forming. Further in-depth TEM studies are warranted. INTRODUCTION

Geopolymers are alkali-containing, aluminosilicate ceramics that are formed by the mixing of alkali-silicate solutions with aluminosilicate minerals or glasses. MacKenzie and Barbosa1 hypothesized that GPs are formed from cross-linked AlO4

- and SiO4 tetrahedra, where charge-balancing cations are provided by alkali metal cations, i.e. Li+, Na+, K+, Cs+ or Rb+. The formation and characteristics of GPs have been the subject of numerous studies.1-25 Applications of geopolymers have included ceramic matrix composites2-8, waste encapsulation8-

10 and as alternative cements.11-19

This work covers the thermal conversion of potassium-based geopolymers into ceramics and an investigation into the microstructure of thermally converted geopolymers.

EXPERIMENTAL PROCEDURES SiO2 reinforced geopolymer A concentrated solution of potassium hydroxide was mixed using deionized water and potassium hydroxide pellets (Fisher Scientific Inc., Pittsburgh, PA). It was assumed that ten weight percent of each potassium hydroxide pellet was water. Cab-o-sil EH-5 ® fumed silica (Cabot Corp., Wheaton, IL) was then dissolved into the potassium hydroxide solution to form a potassium silicate solution with a molar ratio of K2O • SiO2 • 7.5 H2O. After complete dissolution of the fumed silica, the potassium silicate solution was a transparent yellow liquid. The aluminosilicate source used for forming the geopolymer came from metakaolin, derived from Hydrite PXN® kaolin (Imerys, Dry Branch, GA). The kaolin was converted into metakaolin by calcining at 700oC for one hour using a heating and cooling ramp rate of 5oC/min. Amorphous silica spheres, 500 nm in diameter (Fiber Optic Center Inc., New Bedford, MA) were sieved through a 200 mesh (74 µm) sieve, in preparation for mixing. Using an ice bath, the potassium silicate solution, metakaolin, and SiO2 spheres were mixed together in a high-speed shear mixer. The molar ratios of the final geopolymer were K2O • Al2O3 • 4 SiO2 • 7.5 H2O. The resulting paste was applied to the surfaces of untreated glass microscope slides. The slides were then placed onto a perforated stainless steel plate suspended above water in a sealed plastic container. The samples were allowed to cure at 50oC for 50 days. After curing, the samples were slid off the untreated glass microscope slides. One of the samples was then crushed and ground using a porcelain mortar and pestle for X-ray diffraction (XRD) analysis.

A second sample of geopolymer was then heated to 700oC for one hour, using a heating and cooling ramp rate of 5oC/min. After heating, the sample remained intact. The sample was then heated to 1400oC for one hour using a heating and cooling ramp rate of 5oC/min. After firing, the sample was still intact, although several major cracks had developed on the sample surface. Some warping near the sample’s edges was also observed. Powdered samples of the geopolymer sample were heated to 900°, 1100° and 1400°C for one hour, using the same heating and cooling ramp rate. A section of the geopolymer that was heated to 1400 oC, was polished and sputter-coated with a gold-palladium alloy for scanning electron microscopy (SEM Model S-4700, Hitachi, Osaka, Japan). XRD patterns of the powdered geopolymer were taken, using a step size of 0.02 at a rate of 0.5o 2θ/min at an operating voltage of 45 kV and 20 mA from 5o to 75o 2θ using Cu Kα radiation (Rigaku D/Max, Tokyo, Japan). The XRD data was analyzed with the help of JADE® PC software (Minerals Data Inc., Livermore, CA). Powder pressed sample MetaMax EF® metakaolin (Engelhard Corp., Iselin, NJ) was added to a potassium silicate solution pre-prepared with a molar ratio of K2O • 2 SiO2 • 10 H2O to form a geopolymer having a molar ratio of K2O • Al2O3 • 4 SiO2 • 10 H2O, and mixed with a high-speed shear mixer. The geopolymer gel was then poured into an airtight plastic container and allowed to cure at 50oC for two days. After removal from the container, the cured geopolymer was dried at 50oC for two days. After drying, the geopolymer was powered using a mortar and pestle and sieved through a 200 mesh (74 µm) sieve, to eliminate particles > 74 µm. The powder was then pressed into pellets at 500 psi (3.45 MPa) with a minor addition of poly-vinyl alcohol. The pressed pellets were heated to 1200oC for ten hours at a heating and cooling rate of 5oC/min. The sintered pellets showed no signs of cracking. After heating, the pellets were prepared for either XRD or SEM analysis in a manner similar to the preparation of the SiO2 reinforced samples. TEM

preparation of the samples was done using standard, ion milling, preparation techniques. The samples were then carbon coated and examined by TEM at an operating voltage of 120 kV (Phillips CM12). RESULTS: SiO2 reinforced sample

The ABC remained amorphous or semi-crystalline between room temperature and 900oC as seen from the XRD plots in Fig.1. Above 900oC, however the ABC crystallized into single-phase, low temperature, tetragonal leucite (KAlSi2O6). Ceramic grains are visibly absent from the SEM micrographs of the heated and polished ABC sample and large spherical voids, on the order of 50 microns, covered the surface of the sample, as seen in Fig. 2.

Fig. 1. Ex-situ XRD patterns of an ABC with a composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O incorporating amorphous SiO2 spheres after being heated to 25oC (bottom pattern), 900oC, 1100oC and 1400oC (top pattern).

Fig.2. SEM image of an ABC with a composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O

10 20 30 40 50 60 70

Two-Theta (deg)

0

2500

5000

7500

Inte

nsity(C

ou

nts

)

[1-1-3-plus1-500nm-7pt5h2o-pxn-roomtemp.MDI] Comment=1-1-3-plus1-500nm-7pt5h2o-pxn-roomtemp <Psi=0.0>

[1-1-3plus1-PXN-500nmspheres-K-900C-5min-Up-Down.MDI] Comment=1-1-3plus1-PXN-500nmspheres-K-900C-5min-Up-Down <Psi=0.0>

[1-1-3plus1-PXN-500nmspheres-K-1100C-5min-Up-Down.MDI] Comment=1-1-3plus1-PXN-500nmspheres-K-1100C-5min-Up-Down <Psi=0.0>

[1-1-3+50umSF-1400C-5Cupdown-PXN-MK.MDI] Comment=1-1-3+500nmSF-1400C-5Cupdown-PXN-MK <Psi=0.0>

incorporating amorphous SiO2 spheres after being heated 1400oC for one hour. Powder Pressed Sample Similar to the SiO2 reinforced sample, the diffraction pattern of the heated ABC, (Fig. 3.) indicated that the heated ABC crystallized into leucite (K2O • Al2O3 • 4 SiO2). SEM analysis, of the pressed powder sample also showed no distinct ceramic grains, although there was a suggestion of a facetted phase dispersed in a continuous matrix phase (Fig. 4). As seen through large pores on the surface (Fig. 4(b), the internal microstructure of converted ABC consisted of fused agglomerates, approximately three microns in diameter. A converted leucite sample was further annealed at 1400°C/5h. The SEM micrograph again suggested the presence of large, possibly crystallized leucite grains, dispersed in a continuous matrix (Fig. 5).

Fig. 3. Ex-situ XRD patterns of an ABC with a composition of K2O • Al2O3 • 4 SiO2 •10 H2O after being heated to 1200oC/1h.

Fig. 4. SEM images of ABC with a composition of K2O • Al2O3 • 4 SiO2 • 10 H2O after being heated to 1200oC/10 h.

a b

Fig. 5. Fracture surface SEM micrograph of an ABC heated to 1200°C/12h for conversion to leucite (K2O • Al2O3 • 4 SiO2), and then further annealed at 1400oC for 5h. There is an apparently facetted phase dispersed in a continuous matrix.

Sample preparation for TEM proved to be difficult as many of the large facetted grains seemed to have fallen out of the specimen. However, micrographs of the assumed to be remnant continuous phase were taken, and are illustrated in Figs. 6-8. TEM SAD micrographs indicated that the continuous matrix phase was amorphous, with possibly some incipient nanocrystals of leucite forming (Figs. 6 and 7). The microstructure consists of a lighter continuous matrix phase surrounding a few darker inclusions ~5 nm in diameter. The microstructure is reminiscent of spinodal decomposition such as might be formed by a miscibility gap. Increased micrograph exposure times in SAD mode showed faint diffraction spots in addition to amorphous diffuse diffraction. In Figs. 6(a) and 7(a) the dark spots are thought to be nanograins which happen to be in diffracting conditions in the TEM, as confirmed by the corresponding SAD patterns and DF images made from the diffracted spots, as seen in Fig. 8.

Fig. 9 is a higher magnification region of an ABC having a composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O after being heated to 1200oC for one hour. The micrograph is consistent with separate earlier observations made by gas phase porosimetry, that ABC’s prepared from natural metakaolins contain ~40% porosity by volume, having an average pore radius of ~3nm.26 The TEM micrograph of Fig. 9(a) coupled with the SAD pattern of Fig. 9(b) suggest that the dark regions of the micrograph may be such pores.

Fig.6. Bright-Field TEM micrograph of an ABC with a composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O after being heated to 1200oC for one hour (a). Corresponding selected area diffraction (SAD) of the region (b).

Fig. 7. Bright-Field TEM image of an ABC with a composition of K2O • Al2O3 • 4 SiO2 •10 H2O after being heated to 1200oC (left). Selected area diffraction of the region with one second exposure time (center), 30 second exposure time (right).

a b

Fig. 8. TEM micrographs taken in bright field (BF) (a) and corresponding dark field (DF) (b) conditions, confirming the presence of incipient nanocrystallites, thought to be leucite.

Fig. 9. TEM micrographs of an ABC with a composition of K2O • Al2O3 • 4 SiO2 • 7.5 H2O after being heated to 1200oC for one hour (a). Selected area diffraction (SAD) of the region (b).

a b a b BF DF

• Geopolymers are a hydrated, alumino-silicate, gel whose charge is compensated by alkali

cations • Geopolymers have an amorphous microstructure, with nanoporosity of the order of 1-3 nm • The nanoporosity constitutes ~41% by volume of the as-set geopolymer matrix • On ~1000°C annealing (Al2O3:M2O:4SiO2:10 H2O) compositions ceramic compounds can be formed (leucite, pollucite, kalsilite (1:1:2:10) • GPs show some potential as adhesives, particularly to glass and possibly some ceramic oxides. • After the conversion process, GPs appear to be a porous, amorphous gel-ceramic (cf. glass-ceramic) mixture, rather than a fully crystalline ceramic. • The ceramic component contains defects (e.g. remnant pores and twins)

DISCUSSION: SiO2 reinforced sample The added silica spheres were successfully incorporated into the geopolymer paste, such that upon heating the final geopolymer crystallized into leucite. The presence of spherical pores, two orders of magnitude larger than the added SiO2 spheres, were the result of entrapped air in the ABC paste. The absence of grains on the surface of the heated ABC was most likely the result of an amorphous, silica-rich phase coating the surface of the converted ABC. During heating and prior to crystallization, the ABC may have formed an amorphous layer on the outer surface, effectively acting as a glaze. Despite the low molar water ratio used (H2O / K2O = 7.5) and reinforcement with silica, cracking of the SiO2 reinforced sample still occurred upon heating. Powder pressed sample As in the SiO2 reinforced sample, no ceramic grains were visible on the surface of the sample. This is due to the surface being covered with an amorphous phase. The agglomerations in the interior of the sample are thought to be the crystalline leucite grains indicated by XRD, covered with an intergranular amorphous phase. This type of microstructure is reminiscent of glass-ceramics. TEM selected area diffraction patterns of the continuous matrix phase suggested that it was generally amorphous, with occasional incipient nanocrystals of leucite forming. HRTEM observations were consistent with separate measurements of nano porosity of ~3 nm made by gas phase porosimetry. Further detailed TEM studies of crystallized ABC resulting from high temperature heat treatments are planned to identify the exact microstructure of heated and annealed ABCs at the nanometer level.

The process of forming ABCs can be utilized to hydrothermally synthesize amorphous phases or “glasses” at low temperatures, which can subsequently be crystallized into leucite. Depending on the choice of alkali and the amount of silica present, the ABC processing route could be extended to forming amorphous precursors to nepheline (NaAlSiO4), kalsilite (KAlSiO4) and pollucite (CsAlSi2O6) at low temperatures. CONCLUSIONS

Alkali-bonded ceramics (geopolymers) can be converted into single-phase ceramic leucite at high temperatures. Crystallization of potassium-based ABCs into leucite (K2O • Al2O3 • 4 SiO2) occurs between 900 and 1000oC. SEM analyses of the heated samples indicate that the converted ABCs are reminiscent of glass ceramics. TEM of the amorphous, matrix phase of an ABC heated to 1200oC shows a two-phase microstructure, reminiscent of a gel or a spinodal decomposition product. TEM selected area diffraction of the amorphous phase of the leucite-based ABC showed diffuse scattering characteristic of amorphous phase, as well as some reflections due to incipient nanocrystals, as confirmed by TEM DF imaging. Powders derived from ABCs can be used in traditional ceramic processing techniques to produce crack-free ceramics. Further in-depth TEM studies are warranted. ACKNOWLEDGEMENTS

This work was supported by the AFOSR, under STTR Stage II, Grant number F49620-02 C-010. The work was carried out in part, in the Center for Microanalysis of Materials, University

of Illinois, which is partially supported by the US Department of Energy under Grant DEFG02-91-ER45439. REFERENCES 1 V.F.F. Barbosa, K.J.D. MacKenzie, “Thermal Behaviour of Inorganic Geopolymers and Composites Derived from Sodium Polysialate.” Materials Research Bulletin, 38 [2] 319-331 (2003). 2 R.E. Lyon, P.N Balaguru, A. Foden, U. Sorathia, J. Davidovits and M. Davidovits, “Fire-Resistant Aluminosilicate Composites,” Fire Materials, 21 [2] 67-73 (1997). 3 J. Davidovits, M. Davidovits and N. Davodivits, “Geopolymer, Ultra High-Temperature Tooling Material for the Manufacture of Advanced Composites,” International SAMPE Symposium and Exhibition, 36 [2] 1939-1949 (1991). 4 C.G. Papakonstantinou, P.N. Balaguru and R.E. Lyon, “Comparative Study of High Temperature Composites,” Composites: Part B, 32 637-649 (2001). 5 S.C. Förster, T.J. Graule and L.J. Gauckler, “Strength and Toughness of Reinforced Chemically Bonded Ceramics,” Ceramic Transactions, 40 247-256 (1994). 6 J. Davidovits, M. Davidovits and N. Davodivits, “Geopolymer, Ultra High-Temperature Tooling Material for the Manufacture of Advanced Composites,” International SAMPE Symposium and Exhibition, 36 [2] 1939-1949 (1991). 7 W. M. Kriven, J. L. Bell and M. Gordon, “Microstructure and Microchemistry of Fully-Reacted Geopolymers and Geopolymer Matrix Composites,” Ceramic Transactions, 153 227-250 (1994). 8 J.G.S. Van Jaarsveld, J.S.J. Van Deventer, L. Lorenzen, “The Potential Use of Geopolymeric Materials to Immobilize Toxic Metals: Part I. Theory and Applications,” Minerals Engineering 10 [7] 659-669 (1997). 9 J.G.S. Van Jaarsveld, J.S.J. Van Deventer, A. Schwartzman, “ The Potential Use of Geopolymeric Materials to Immobilise Toxic Metals: Part II. Material and Leaching Characteristics,” Minerals Engineering, 12 [1] 75-91 (1999). 10 J.W. Phair, J.S.J. Van Deventer, J.D. Smith, “Effect of Al Source and Alkali Activation on Pb and Cu Immobilisation in Fly-ash Based "Geopolymers",” Applied Geochemistry, 19 [3] 423-434 (2004). 11 H. Xu, Van Deventer J.S.J., “Geopolymerisation of Multiple Minerals,” Minerals Engineering, 15 [12] 1131-1139 (2002). 12T.W. Cheng, J.P. Chiu, “Fire-resistant Geopolymer Produced by Granulated Blast Furnace Slag,” Minerals Engineering, 16 [3] 205-210 (2003).

13 H Xu, J.S.J. Van Deventer, “Effect of Source Materials on Geopolymerization,” Industrial & Engineering Chemistry Research, 42 [8] 1698-1706 (2003). 14 J.G.S. Van Jaarsveld, J.S.J. Van Deventer, “Effect of the Alkali Metal Activator on the Properties of Fly Ash-based Geopolymers,” Industrial and Engineering Chemistry Research, 38 [10] 3932-3941 (1999). 15 H. Xu, J.S.J Van Deventer, G.C. Lukey, “Effect of Alkali Metals on the Preferential Geopolymerization of Stilbite/Kaolinite Mixtures,” Industrial & Engineering Chemistry Research, 40 [17] 3749-3756 (2001). 16 H. Xu, J.S.J.Van Deventer, “The Geopolymerisation of Alumino-silicate Minerals,” International Journal of Mineral Processing, 59 [3] 247-266 (2000). 17 V.F.F. Barbosa, K.J.D. MacKenzie, C. Thaumaturgo, “Synthesis and Characterisation of Materials Based on Inorganic Polymers of Alumina and Silica: Sodium Polysialate Polymers,” International Journal of Inorganic Materials 2 [4] 309-317 (2000). 18 V.F.F. Barbosa, K.J.D. MacKenzie, “Synthesis and Thermal Behaviour of Potassium Sialate Geopolymers.” Materials Letters, 57 [9-10] 1477-1482 (2003). 19 J. Davidovits, “ Geopolymers - Inorganic Polymeric New Materials,” Journal Of Thermal Analysis, 37 [8] 1633-1656 (1991). 20 H. Rahier, B. Van Mele, M. Biesemans, J. Wastiels and X. Wu, “Low-Temperature Synthesized Aluminosilicate Glasses. Part I. Low-Temperature Reaction Stoichiometry and Structure of a Model Compound,” Journal of Materials Science, 31 [1] 71-79 (1996). 21 H. Rahier, B. Van Mele, J. Wastiels, “Low-Temperature Synthesized Aluminosilicate Glasses. Part II. Rheological Transformations During Low temperature Glasses,” Journal of Materials Science, 31 [1] 80-85 (1996). 22 Rahier H, Simons W, VanMele B, Biesemans M. “Low-temperature Synthesized Aluminosilicate Glasses. Part III. Influence of the Composition of the Silicate Solution on Production, Structure and Properties,” Journal of Materials Science, 32 [9] 2237-2247 (1997). 23 H. Rahier H, J.F. Denayer, B. Van Mele, “Low-temperature Synthesized Aluminosilicate Glasses. Part IV. Modulated DSC Study on the Effect of Particle Size of Metakaolinite on the Production of Inorganic Polymer Glasses,” Journal of Materials Science, 38 [14] 3131-3136 (2003). 24 H. Xu, J.S.J Van Deventer, “Microstructural Characterisation of Geopolymers Synthesized from Kaolinite/stilbite Mixtures using XRD, MAS-NMR, SEM/EDX, TEM/EDX, and HREM,” Cement And Concrete Research, 32 [11] 1705-1716 (2002).

25 J. P. Hos, P. G. McCormick, L. T. Byrne, “Investigation of a Synthetic Aluminosilicate Inorganic Polymer,” Journal of Materials Science, 37 [11] 2311-2316 (2002). 26 W. M. Kriven, P. Sarin, M. Gordon and J. Bell, to be published.