Quantitative kinetic and structural analysis of geopolymers. Part 1. the activation of metakaolin with sodium hydroxide

Article abstract of DOI:10.1016/j.tca.2012.03.021

Isothermal conduction calorimetry (ICC) is used here to measure the kinetics of geopolymerisation of metakaolin by reaction with NaOH solution under a variety of conditions. Three exothermic peaks are observed in the calorimetric curve, and are assigned to the dissolution of metakaolin, the formation of geopolymer with disordered or locally ordered structure, and finally the reorganization and partial crystallization of this inorganic polymer gels. For the purpose of further quantifying the ICC data, the geopolymeric reaction products are assumed to have an analcime-like local structure, and their standard formation enthalpies are estimated from the available data for this structure. This assumption enables ICC to be used for the first time in a quantitative manner to determine the real reaction kinetics of geopolymerization. Increasing the NaOH concentration up to a molar overall Na/Al ratio of 1.1 is seen to enhance the reaction extent observed at 3 days, up to a maximum of around 40% in the high liquid/solid ratio systems studied here, and accelerates the crystallization process. However, further addition of NaOH does not give any additional reaction within this period, or any further acceleration. Raising the reaction temperature from 25 °C to 40°C increases the initial reaction rate but has little effect on the final reaction extent, particularly when Na/Al > 1.

Full text of DOI:10.1016/j.tca.2012.03.021

Thermochimica Acta 539 (2012) 23–33
Contents lists available at SciVerse ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
Quantitative kinetic and structural analysis of geopolymers. Part 1. The activation
of metakaolin with sodium hydroxide
Zuhua Zhanga,∗, Hao Wanga, John L. Provisb, Frank Bullena, Andrew Reidc, Yingcan Zhua
a Faculty of Engineering and Surveying, University of Southern Queensland, Toowoomba, QLD, 4350, Australia
b Department of Chemical & Biomolecular Engineering, University of Melbourne, VIC, 3010, Australia
c Halok Pty Ltd., West End, QLD, 4101, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 January 2012
Received in revised form 20 March 2012
Accepted 23 March 2012
Available online 31 March 2012
Isothermal conduction calorimetry (ICC) is used here to measure the kinetics of geopolymerisation of
metakaolin by reaction with NaOH solution under a variety of conditions. Three exothermic peaks are
observed in the calorimetric curve, and are assigned to the dissolution of metakaolin, the formation
of geopolymer with disordered or locally ordered structure, and finally the reorganization and partial
crystallization of this inorganic polymer gels. For the purpose of further quantifying the ICC data, the
geopolymeric reaction products are assumed to have an analcime-like local structure, and their standard
formation enthalpies are estimated from the available data for this structure. This assumption enables
ICC to be used for the first time in a quantitative manner to determine the real reaction kinetics of
geopolymerization. Increasing the NaOH concentration up to a molar overall Na/Al ratio of 1.1 is seen to
enhance the reaction extent observed at 3 days, up to a maximum of around 40% in the high liquid/solid
ratio systems studied here, and accelerates the crystallization process. However, further addition of NaOH
does not give any additional reaction within this period, or any further acceleration. Raising the reaction
temperature from 25 ◦C to 40 ◦C increases the initial reaction rate but has little effect on the final reaction
extent, particularly when Na/Al > 1.
Keywords:
Geopolymer
Isothermal conduction calorimetry
Metakaolin
Kinetics
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
final products [3,5–7]. To control properties such as the setting
time and rheology of the geopolymer slurry, and to better under-
Geopolymer binders and cements are generally formed by
reaction of an aluminosilicate powder (such as metakaolin or
fly ash) with an alkaline solution. The chemical reaction process
responsible for geopolymer formation, known as geopolymeriza-
tion, involves the dissolution of Si and Al atoms from the source
material, reorientation of precursor ions in solution, and conden-
sation reactions to form an inorganic polymer (geopolymer) [1].
Since geopolymer production reduces the need for high tempera-
ture calcination processes compared with Portland cement clinker
manufacture, geopolymers have high potential for use in ‘green’
concretes with less environmental impact than concretes based on
ordinary Portland cement (OPC), particularly when local industrial
wastes can be used [2]. In addition, geopolymers can possess other
desirable properties such as high thermal shock resistance [3] and
low permeability [4].
stand the fundamental characteristics of the final products, detailed
researchrelated to reactionkineticsis essential[8]. Geopolymeriza-
tion is a multistep process, involving dissolution, gel formation and
transformation reactions involving crystallographically disordered
phases, probably together with partial re-dissolution of the prod-
ucts as part of a dynamic equilibrium between gel and dissolved
species [9]. This means that it is difficult to develop a characteri-
zation technique that can provide a unified and all-encompassing
description of such a complex reaction sequence, although there
have been various studies which have provided data characterizing
different aspects of the reaction.
Calorimetry is a classic technique used in determining the
reaction kinetics of cement [10]. It has also been introduced
in studying alkali activation of slag [11], and in examining the
effects of various parameters on alkali activation of metakaolin
based geopolymers, such as reaction temperature [12–14], alkali
type and concentration [12,15–18], and the nature of the start-
ing kaolin/metakaolin precursor [19,20]. In these studies, which
have applied both isothermal and non-isothermal measurements,
the calorimetry provided helpful information regarding the process
of geopolymerization, however, it has not previously been able to
provide a quantitative determination of the real reaction extent,
Extensive studies have been carried out to understand the
effects of the type and concentration of alkaline activator on the
process of geopolymerization and the mechanical properties of the
∗
Corresponding author. Tel.: +61 7 46312549.
E-mail address: zuhua.zhang2@usq.edu.au (Z. Zhang).
0040-6031/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.tca.2012.03.021

24
Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
Table 1
Composition of metakaolin (MK) as determined by XRF (wt.%; LOI is loss on ignition at 1000 ◦C).
Oxide
SiO2
Al2O3
41.55
K2O
0.43
MgO
0.05
Fe2O3
0.56
TiO2
0.26
P2O5
0.23
Na2O
0.26
LOI
Composition
55.57
0.91
similar to the information that can be obtained in OPC systems. This
level of insight is able to be achieved for OPC because the hydration
degree of OPC can be determined from the ratio of cumulative heat
released to the total heat release expected at full hydration, which
can be theoretically calculated [21]. Up to this time, the lack of a
reliable measure of total heat release at ‘full geopolymerization’ has
limited the use of calorimetry results in quantifying geopolymer-
ization kinetics.
content since it is also reactive; this assumption will be justified in
Section 3 below. The quartz (estimated as 6.8 wt% by assuming all
Al is from kaolinite, and that the kaolinite is 1:1 stoichiometric) is
considered as being inactive in the geopolymerization process due
to its low reactivity.
2.2. Calorimetry
Alternative and novel characterization techniques have also
provided useful information regarding the process of geopoly-
merization. Rees et al. [22] and Hajimohammadi et al. [23]
used attenuated total reflectance Fourier transform infrared
spectroscopy (ATR-FTIR) to monitor geopolymerization reaction
systems at early age, providing the capability to monitor gel forma-
tion directly, but the application of this technique is often hindered
by the presence of silica in the activating solution. Provis et al.
[24–26] directly measured the reaction kinetics during the initial
setting period by in situ energy-dispersive X-ray diffractometry
(EDXRD) and alternating current impedance spectroscopy, and also
presented a reaction kinetic model to describe the different stages
of the reaction process. White et al. [27] developed a mathematical
model for the process of geopolymer gel growth from metakaolin
based on a coarse-grained Monte Carlo simulation using parame-
ters derived from density functional theory, and also applied in situ
neutron pair distribution function analysis to determine important
information regarding structural evolution at early age [28], but
quantification of the precise degree of reaction has also not yet
been possible by these techniques.
The geopolymer pastes were studied in a 3114/3236 TAM 83 Air
isothermal conduction calorimeter (Thermometric AB, Sweden) at
20, 25, 30, 35 and 40 ◦C by an internal mixing procedure. Prelimi-
nary tests showed that a homogeneous paste could be obtained by
this procedure only when the liquid/solid ratio was higher than
0.8 mL/g for the calcined kaolin used in this study. The pastes
studied were thus designed with a liquid/solid ratio of 1.0 mL/g
to ensure successful mixing. The MK powder was placed in an
ampoule containing a mini-blender, and NaOH solution was held
separately in an injector. Both the ampoule and injector were
maintained in the calorimeter at the reaction temperature until
a thermal balance condition was reached. Once the NaOH activa-
tor was injected, the mini-blender was started and the heat flow
immediately recorded at 10 s intervals. At each reaction tempera-
ture, the heat flow from a control paste prepared by mixing water
with MK under the same conditions was also recorded, to provide
a baseline to calibrate the result and to account for the heat due
to wetting and the mini-blender. To minimize possible heat distur-
bances due to the environment, the room was maintained at the
reaction temperature, except for the systems at 35 ◦C and 40 ◦C.
It is clear that only a combination of multiple characterization
techniques can give full insight into both the reaction extent and
early-age products formed during geopolymerization. In this inves-
tigation, isothermal conduction calorimetry (ICC) is combined with
X-ray diffraction (XRD), infrared spectroscopy (FTIR) and scanning
electron microscopy (SEM) methods to monitor the geopolymer-
ization process in the simplified metakaolin–NaOH system, and to
examine the reaction products. Two important reaction parameters
for geopolymerization – reaction temperature and alkali concentra-
tion – are investigated in terms of their influence on kinetics in a
quantitative sense.
2.3. Characterization of geopolymers
After geopolymerization, a scanning electron microscope (JEOL
Ltd., Japan) was used to analyze the morphology of selected speci-
mens. Fracture surfaces were coated with gold, and imaged under
high vacuum conditions with an accelerating voltage of 15 kV.
Powder samples were prepared for X-ray diffraction (XRD) and
FTIR analysis by grinding the geopolymer reaction products in ace-
tone followed by oven-drying at 65 ◦C for 6 h. The XRD analysis was
conducted on an ARL 9900 Series X-ray workstation (Thermo Scien-
tific) with Co K␣ radiation operated at 40 kV and 40 mA at a scanning
speed of 2.4◦ min−1. The FTIR spectra were collected in the range
4000–400 cm−1 using a Nexus 670 FTIR spectrometer (Nicolet) in
transmittance mode, with a sensitivity of 2 cm−1. The spectra were
converted into relative absorbance values, and deconvoluted in the
range 800–1300 cm−1 using the Peakfit (Version 4.12) software
with Gaussian peak shapes and variable peak widths. The fitting
process, including setting the numbers and positions of component
bands and adjusting the band shape, was performed in accordance
with procedures described in the literature [22,29,30], combined
with the self-fitting function of the software. The main princi-
ple used in the fitting procedure was to minimize the number of
2. Experimental
2.1. Geopolymer paste formulation
The metakaolin (MK) used in this study was obtained by calci-
nation of kaolin powder at 900 ◦C (Taojinfeng New Materials Co.
Ltd., Fujian, China). Table 1 gives the composition of the MK as
determined by X-ray fluorescence (XRF, Rh radiation). The parti-
cle size distribution is represented by the D10, D50 and D90 values
determined by laser diffraction, which were 1.4 ␮m, 5.9 ␮m, and
17.0 ␮m, respectively, and the specific surface area as measured by
N2 sorption and the Brunauer–Emmett–Teller (BET) method was
14.1 m2/g. The activator used was sodium hydroxide solution, pre-
pared by dissolving granular NaOH (chemical grade) in distilled
water.
Table 2
Composition of geopolymerization systems, defined as molar ratios.
NaOH (mol/L)
H2O:Na
Na:Al
Al:Si
Table 2 gives the design of geopolymer pastes. It should be noted
that the residual kaolinite (6.5 wt.% as calculated by assuming that
the loss on ignition (LOI) of the metakaolin is due to the dehydration
of kaolinite) is treated as part of the amorphous aluminosilicate
6
8
10
12
9.26
6.94
6.17
4.63
0.74
0.98
1.10
1.47
1
1
1
1

Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
25
40
30
20
10
0
50
a
c
40
Peak I
30
Peak II
20
Peak I
Peak II
10
10
12
6
10
12
12
8
6
0
0
12
24
36
48
60
72
0
12
24
36
48
60
72
Reaction time (h)
Reaction time (h)
50
40
30
20
10
0
b
d
40
30
20
10
0
6M
6M
Peak I
8M
8M
10M
12M
10M
12M
Peak I
Peak II
Peak II
12
10
10
Peak III
6
12
8
12
6
8
0
12
24
36
48
60
72
0
12
24
36
48
60
72
Reaction time (h)
Reaction time (h)
50
e
Peak I
40
30
20
Peak II
10
Peak III
60
6
12
10
0
0
12
24
36
48
72
84
96
Reaction time (h)
Fig. 1. Effects of NaOH concentration on heat evolution rate of MK geopolymerization at (a) 20 ◦C, (b) 25 ◦C, (c) 30 ◦C, (d) 35 ◦C and (e) 40 ◦C.
In the reaction systems at 20–30 ◦C there are two distinct peaks
in heat flow, as marked in Fig. 1: peak I is sharp and appears imme-
diately after mixing, followed by a broad peak II. Peak III is only
found in systems 12 M-35 ◦C (Fig. 1d), 12 M-40 ◦C and 10 M-40 ◦C
(Fig. 1e), although it appears in the last of these after more than
72 h. The nature of the reaction processes responsible for each peak
will be discussed below. The time required for peak I to reach a
maximum in the rate of heat evolution is constant at around 5 min
for all the systems studied, regardless of the reaction temperature.
In contrast, the time corresponding to the maximum of peak II
varies notably accordingly to the reaction temperature. Increasing
the reaction temperature significantly shortens the time before the
second maximum in heat evolution; this time period may be
component bands while retaining a regression coefficient r2 value
above 0.999.
3. Results
3.1. ICC testing
Fig. 1 shows the heat evolution rates of MK activated with NaOH
solution at 20–40 ◦C in the first 72 h of reaction. As mentioned
in Section 2.2, the heat evolution rate data are corrected by sub-
tracting the heat evolution of MK mixed with water, meaning that
the data presented correspond specifically to the chemical reaction
process.

26
Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
Table 3
Cumulative heat release of exothermic peaks I and II in NaOH-activated metakaolin geopolymers in the first 72 h of reaction (or until the onset of peak III in some systems,
where this time is noted).
NaOH (mol/L)
Heat release (J/g MK)
20 ◦
C
25 ◦
C
30 ◦
C
35 ◦
C
40 ◦
C
6
8
10
12
461
544
588
571
326
421
580
625
414
541
584
624
436
556
622
442
578
626 (54 h)
656 (24 h)
655(40.5 h)
compared with the ‘induction period’ observed in Portland cement
hydration, although the physicochemical processes taking place
will be very significantly different in geopolymerization. When
viewed together with previous work [15], it is likely that there
would also be a peak III in other systems with lower temperature
and NaOH concentration, but that it would be weak, and would
appear later than the 72-h period of the experiments here.
most exothermic aspect of the reaction process, as seen qualita-
tively from the high early heat release (peak I) in all data sets in
Fig. 1. Quantitative analysis in support of this attribution will be
presented later in this paper.
3.2. XRD analysis
The amplitude of each heat release peak varies significantly as
the reaction temperature is changed from 20 to 40 ◦C. For peak
I, the maximum heat release rate increases monotonically with
increasing NaOH concentration and temperature. For peak II, it
is interesting to note that the maximum rate is usually higher in
systems activated with 8 M or 10 M NaOH than in systems with
a higher activator concentration of 12 M. This is attributed to the
more extensive polymerization between Al and Si species taking
place in the systems with slightly lower activator concentrations
than at this highest alkalinity, as the solubility of aluminosilicate
s increases with increasing pH; further discussion on this point will
be presented later in this paper. The trend that the maximum heat
evolution rate of peak II is much lower than peak I is consistent
with the results of previous studies of metakaolin geopolymeriza-
tion [15,31]. However, this trend can be altered by calcium addition
[31], the use of a silicate activator [12] or the use of different alu-
minosilicate precursors [32].
The cumulative heat release of the first and second peaks in all
systems is summarized in Table 3. The results fall in the same range
as the systems measured by Alonso and Palomo [12] and by Mun˜iz-
Villarreal et al. [13]. The slow reaction at 20 ◦C results in a higher
total heat release than that at 25 ◦C in all systems except with 12 M
NaOH. Higher NaOH concentrations or higher temperatures, being
conditions which enhance the solubility of metakaolin, lead to a
higher total heat release, as the dissolution of metakaolin is the
Fig. 2 shows the XRD data for MK and the geopolymer samples
obtained after 72 h of reaction at 30 ◦C. The MK used is largely amor-
phous, with a small quantity of quartz as an impurity phase, and
minor kaolinite and halloysite. The broad peak observed from 15
to 35◦ 2ꢀ in MK has been broadened up to 40◦ 2ꢀ after reaction
with NaOH solution, with the center of gravity shifted to higher
angle, as is well known to occur during geopolymer formation from
metakaolin. This shift demonstrates the formation of new amor-
phous phases, usually described as geopolymeric gels [14,33]. The
weakening of the diffraction peaks of kaolinite and halloysite (a
tubular polymorph of kaolinite) suggests that these phases have
participated in the reactions, which justifies their inclusion as reac-
tive components in the mix design calculations in Section 2.1. In
general the products are amorphous, and only in the 12 M system
small amounts of hydroxysodalite (Powder Diffraction File (PDF)
#46-1425) and sodium carbonate (PDF # 1-1166) are formed. The
sodium carbonate may be formed due to carbonation of excess
NaOH during sampling, crushing and XRD analysis. Comparing the
relative intensities of the broad diffraction peaks in Fig. 2, the inten-
sity between 30 and 40◦ 2ꢀ (attributed to geopolymer) is seen to
be higher than the intensity between 20 and 30◦ 2ꢀ (attributed
to MK) in the samples with higher NaOH concentration, implying
increased formation of geopolymeric gels. This is particularly visi-
ble when comparing the intensities at around 33◦ 2ꢀ and 27◦ 2ꢀ in
Fig. 2, where this ratio increases with NaOH concentration. How-
ever, the exact proportions of residual MK and newly formed gels
are unable to be determined by XRD; this is an acknowledged short-
coming of this characterization method in analyzing materials with
multiple distinct disordered (‘X-ray amorphous’) phases [24].
Fig. 3 shows the XRD patterns of the geopolymerization prod-
ucts from systems in which peak III appears (Fig. 1). Before peak
III appears, for systems activated with NaOH concentration of
12 mol/L (Fig. 4a-36 h and b-18 h), the products are mainly amor-
phous. The small peaks appearing at 2ꢀ = 28.16◦, 31◦, 34◦ and 35.04◦
are possibly due to crystalline nano-sized particles of zeolite A (PDF
#39-0272); the same peaks were observed, but more intense and
sharper, in the system with 10 mol/L at 40 ◦C (Fig. 4c-54 h), indi-
cating a higher extent of crystallization and a larger crystallite size
in this system. However, all systems studied contained predomi-
nantly disordered rather than crystalline aluminosilicate phases.
It is interesting to note that when comparing before and after
peak III, the XRD patterns did not change significantly, indicating
that this peak is not able to be attributed to a specific crystalliza-
tion event in the gel systems. There is an increased diffraction
intensity of zeolitic materials, but this is only slight, indicating
that peak III is probably caused by reorganization within the poly-
mer/zeolite precursor materials present, rather than a process of
crystallization of more polymer gels. The distinct difference in the
hydroxysodalite
Sodium carbonate
12 M
10 M
8 M
6 M
halloysite
MK
kaolinite
quartz
10
20
30
40
50
60
70
80
Fig. 2. Co K␣ radiation XRD patterns of MK and its geopolymerization prod-
ucts after activation with NaOH solution of 6–12 mol/L at 30 ◦
for 72 h
C
(kaolinite, Al2Si2O5(OH)4: powder diffraction file (PDF) #01-0527; halloysite,
Al2Si2O5(OH)4·2H2O: PDF #09-0451; quartz, SiO2: PDF #05-0490; hydroxysodalite,
Na8Al6Si6O24(OH)2·2H2O: PDF #46-1425; sodium carbonate, Na2CO3: PDF #1-
1166).

Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
27
kaolinite
halloysite
quartz
a
b
kaolinite
halloysite
quartz
o
A
12M-40C
o
A
12M-35C
A zeolite A
A zeolite A
A
A A
A
A
A
A
72 h
54 h
18 h
36 h
10
20
30
40
2θ (o)
50
60
70
80
10
20
30
40
2θ (o)
50
60
70
80
A
c
o
quartz
10M-40C
A
zeolite-A
A
A
A
A
A
A
A
A
A
A
A
A
A A
A
A
A
108 h
A
A
A
A
A
54 h
70
10
20
30
40
2θ (o)
50
60
80
Fig. 3. Co K␣ radiation XRD patterns of geopolymerization products of MK activated with NaOH solution for different periods: (a) 12 mol/L at 35 ◦C, (b) 12 mol/L at 40 ◦C and
(c) 10 mol/L at 40 ◦C (zeolite A, Na96Al96Si96O384·216H2O: PDF #39-0272). Each of these systems shows a third exothermic peak in the isothermal calorimetry data (“peak
III”), and each set of diffractograms contains one sample taken before the time of this peak, and another after the peak, with times as marked on the plots.
crystallinity of the products between the systems activated with
12 M NaOH compared to 10 M NaOH implies that zeolite phases
preferred to form in systems with a slightly lower alkali content
and higher reaction temperature. This may again be a solubility
effect, or it may be because very high NaOH concentrations favor
the formation of hydroxysodalite rather than zeolite A [34], and so
the crystallization processes are slowed close to the phase bound-
ary between these species. Although peak III is observed later in
o
12M-35C-36 h
12 M
10 M
8 M
o
12M-35C-72 h
o
12M-40C-18 h
o
12M-40C-54 h
6 M
o
Metakaolin
10M-40C-54 h
o
10M-40C-108 h
4000 3600 3200 2800 2400 2000 1600 1200 800 400
4000 3600 3200 2800 2400 2000 1600 1200
800
400
Wavenumber( cm-1)
Wavenumber( cm-1)
Fig. 5. FTIR spectra of selected geopolymerization products of MK activated with
Fig. 4. FTIR spectra of MK and the geopolymerization products after activation with
NaOH solution at 35 ◦C and 40 ◦C for different periods.
NaOH solution at 30 ◦C for 72 h.

28
Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.6
a
d
Original
Curve fit
10 M
Original
Curve fit
Metakaolin
1210
0.5
0.4
0.3
0.2
0.1
0.0
1080
996
1078
950
818
1136
1020
858
1212
1200
888
1300
1200
1100
1000
900
800
1300
1100
1000
900
800
-1
-1
Wavenumber (cm )
Wavenumber (cm )
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0.5
0.4
0.3
0.2
0.1
0.0
b
e
Original
Curve fit
Original
Curve fit
852
12 M
6 M
1081
990
1082
1005
944
960
858
1204
1200
1216
852
1300
1100
1000
-1
900
800
1300
1200
1100
1000
900
800
-1
Wavenumber (cm )
Wavenumber (cm )
0.6
c
Original
Curve fit
8 M
0.5
0.4
0.3
0.2
0.1
0.0
1080
996
951
1210
1200
855
1300
1100
1000
900
800
-1
Wavenumber (cm )
Fig. 6. FTIR spectral deconvolution of MK (a) and the geopolymerization products after activation with NaOH solution at 30 ◦C for 72 h: (b) 6 mol/L, (c) 8 mol/L, (d) 10 mol/L
and (e) 12 mol/L.
the system with 10 M NaOH at 40 ◦C than in either of the 12 M
NaOH systems shown in Fig. 3, comparison of the diffractograms
obtained after 54 h in Fig. 3b and c shows a higher crystallinity
at the same reaction time. There is much which remains to be
understood regarding the kinetic and structural aspects of zeolite
crystallization within geopolymers, as both kinetic and thermody-
namic parameters influence phase relationships in these systems
in complex ways, and there are almost always disordered and/or
metastable phases formed.
3.3. FTIR analysis
The FTIR spectra of selected geopolymers are presented in
Figs. 4 and 5. The most evident spectral differences are found

Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
29
in the low-wavenumber region (400–800 cm−1) and the middle-
1010
1000
990
980
970
960
950
940
wavenumber region (800–1250 cm−1).
Principal new band:Si(Al)-O stretching in Si-O-T
In the low-wavenumber region, for the spectrum of MK (Fig. 4),
the bands around 465 cm−1 are assigned to T O (T = tetrahedral Al
or Si) bending modes in TO4 tetrahedra, and the band at 800 cm−1
is assigned to the Al O bending mode of AlO6 octahedra [17,35].
After geopolymerization, the intensities of these two bands both
decrease, particularly at ∼800 cm−1, indicating the breakdown of
the original octahedral structure of the residual kaolinite particles.
The band at 710 cm−1 (Fig. 4) in the geopolymer spectra indicates
the formation of AlIV as the main Al environment in the polymer.
The small band centered at ∼600 cm−1 (Fig. 4) is probably caused
by T O vibrations in double rings in zeolite precursors [35]. A
more prominent band at 557 cm−1 (Fig. 5) appears only in the sys-
tems 10 M-40 ◦C-54 h and 10 M-40 ◦C-108 h. This band is typically
assigned to the external linkage vibrations of the TO4 tetrahedra in
the double rings of zeolite A [36], which agrees well with the XRD
results.
Associated new band:Si(Al)-O stretching in Si-O-T & Si-O-Na
6
8
10
12
NaOH concentration (mol/L)
In the mid-wavenumber region, a shift in the position of the
Fig. 7. The positions of the resolved principal new band and the associated new
band in the FTIR deconvolutions of samples as a function of NaOH concentration.
broad Si
O T asymmetric stretching peak (containing multiple
overlapping components which sum to give a broad peak) from
∼1080 cm−1 in MK to ∼1000 cm−1 in the geopolymer product is
observed, regardless of whether the product is entirely X-ray amor-
phous or contains a higher degree of crystalline material. This
band shift is usually considered to be due to the formation of
geopolymers [31,37], although it is also observed in the formation
of zeolites from disordered aluminosilicate precursors [35,36].
3.4. SEM analysis
Fig.
8 exhibits micrographs of the fracture surfaces of
geopolymerization products. Generally, after reaction with highly
concentrated NaOH solutions, MK particles are converted into a
porous, particulate gel, and are bound with each other on the sur-
faces or at the edge, but also maintain their plate-like morphology
to some extent. The observed porous gel morphologies in this work
are similar to previously published images of metakaolin geopoly-
mers with low Si/Al ratios [41], and also resemble the precursor
gels formed in some hydrothermal zeolite synthesis systems at very
early stages [42].
Fig. 9 shows a fracture surface of the geopolymer obtained from
system 10 M-40 ◦C. The distinct particles as marked by the arrows
and emphasized in the circles in Fig. 9 are identified as being zeolite
A crystallites, although these are still far from the perfect cubic
crystal habit of pure zeolite A [35].
The small band appearing at around 1440 cm−1 is related to
2−
the asymmetric stretching of the O
C
O bonds of CO3
due
to atmospheric carbonation on the surface of powdered prod-
ucts, which is only found within 12 M system for XRD pattern.
The absorption bands at 1630–1660 cm−1, combined with a very
weak absorption band at around 3750 cm−1, suggest the pres-
ence of isolated non-interacting surface silanol groups [38]. The
increased intensity of bands at 1630–1660 and ∼3400 cm−1 with
increasing NaOH concentration implies that the activated prod-
ucts form with more surface hydroxyl groups hydrogen-bonded
to adsorbed water ( Si–OH· · ·H2O and Al–OH· · ·H2O) in sys-
tems with higher alkali content. This is consistent with the
common understanding that high pH favors depolymerization of
aluminosilicates.
4. Modeling of geopolymerization kinetics
To obtain a better understanding of the causes of the band shift
in the mid-wavenumber region, the spectra of products obtained
at 30 ◦C in this range are enlarged and deconvoluted (see Sec-
tion 2.3 for discussion of the deconvolution procedure) in Fig. 6.
After alkali-activation of the metakaolin, the two new bands at
1005–990 cm−1 and 960–944 cm−1 are the most notable changes in
the spectra. The principal new band at 1005–990 cm−1 is assigned
to the asymmetric stretching vibration of Si–O–T links in geopoly-
mer framework; this band is known to be sensitive to connectivity
and Si/Al ratio [23], and in this case it is observed at a wavenum-
ber consistent with the presence of predominantly Si–O–Al bonds,
which agrees well with the stoichiometry of the systems studied.
The second new band, at 960–944 cm−1, is assigned to asymmet-
ric stretching vibration of non-bridging oxygen sites, in particular
Si–O–Na type structures [39,40]. Fig. 7 indicates the influence of
the NaOH concentration on the positions of these two bands. With
increasing NaOH concentration, both of these new bands shift to
lower wavenumber. The shift in the Si–O–T component peak to
lower wavenumber indicates a higher degree of conversion of Al
from non-tetrahedral environments in metakaolin to tetrahedral
environments bonded to Si in the geopolymer framework, which
agrees well with the trends in extent of reaction of MK, and is also
supported by findings in the literature for alkali-activated fly ash
systems [40].
4.1. Geopolymerization process of NaOH-activated MK
Based on the calculated reaction enthalpy and the XRD and FTIR
analysis presented in this study, a reasonable diagram relating the
three exothermic peaks with three stages in reaction can be con-
structed as shown in Fig. 10, which is derived from the conceptual
arguments which have previously been developed into a reaction
kinetic model for geopolymerization [25,26,43].
In Stage I of the reaction process, the geopolymerization begins
from the dissolution of metakaolin (MK) into silicate monomers
(abbreviated S in the reaction scheme) and aluminate monomers
(abbreviated A). In Stage II, the polymerization of S and A results in
aluminosilicate oligomers (O), which immediately polymerize into
small geopolymeric fragments (P) or ‘proto-zeolitic nuclei’ (N). P
and N are thermodynamically metastable and incompletely cross-
linked, and will assemble into aluminosilicate inorganic polymer
gels (G) (approaching percolation of the network) and crystallized
phases (Z), which are larger in molecular weight, higher in connec-
tivity and more ordered in microstructure [25]. These two basic
steps, O + S + A → P and/or N, and subsequently P and/or N → G
and/or Z, overlap within the second broad exothermic peak in
the ICC data, as the heat release involved in these processes has
been calculated to be relatively low compared to the initial dis-
solution and oligomerization reactions [41]. From XRD and FTIR

30
Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
Fig. 9. SEM micrograph of the surface of the geopolymerization products of system
10 M-40 ◦C after 54 h of reaction. Circles and arrows mark products which are similar
in morphology to zeolite A crystallites.
cross-linked and thermodynamically stable states. If the former is
true, there should be some zeolites detectable in the systems 12 M-
35 ◦C-72 h and 12 M-40 ◦C-54 h which were not present in these
systems prior to this heat release peak. However, neither XRD nor
FTIR indicates that any notable quantity of zeolite is formed dur-
ing this heat evolution peak. The reorganization of G seems more
reasonable, possibly including reactions between the disordered
polymers to form proto-zeolitic nuclei [44] which are too small to
be detected by XRD. The transformation of a geopolymer from an
initial amorphous state to a final amorphous state without notable
development of crystallinity has been observed a number of times
in the literature [1,24,28], and so this is likely to be the explanation
for the occurrence of the third exothermic peak.
4.2. Quantification of geopolymerization kinetics
4.2.1. Thermodynamics of G and Z
An isothermal conduction calorimetry measurement supplies
information regarding the reaction enthalpy during the geopoly-
merization process as described in Section 4.1. To use the ICC data
to quantify the extent of geopolymerization, the determination of
the thermodynamic parameters of G and Z is critical. However, due
to the absence of existing high-quality thermodynamic data, and
the difficulty in measuring the thermodynamic parameters through
experiments from system to system, the thermodynamic parame-
ters of G and Z can only be estimated at present. The estimation
procedure used here provides significant advances over previous
work aimed at estimating these parameters [43], because the use
of pair distribution function has shown that metakaolin-derived
˚
geopolymer binders show very similar local (<6 A) structures to
Dissolution
Polymerization
Transformation
MK→S+A
S+A → P and/or N → G and/or Z
G / Z → more ordered
Fig. 8. SEM micrographs of the fracture surfaces of geopolymerization products
of MK activated with NaOH solution: (a) 6 mol/L, (b) 8 mol/L, (c) 10 mol/L and (d)
12 mol/L.
I
II
III
analysis, phases consistent with both G (the disordered gels) and
Z (crystalline zeolite) detectable in the reaction products obtained
at 35 or 40 ◦C, at the end of Stage II of the reaction (Fig. 3). The
final exothermic peak, in Stage III of the reaction, is thus related
to two possible transformations: the crystallization of G into Z;
and the further reorganization of G and/or Z into other more
Reaction time
Fig. 10. Schematic of the kinetics of geopolymer synthesis by NaOH activation of
MK, as determined by isothermal calorimetry.

Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
31
zeolites of the analcime family—leucite for K-geopolymers [45,46],
Na:Al
0.98
1.10
0.74
1.47
and pollucite for Cs-geopolymers [47], meaning that this well-
understood mineral structure is now able to be used as a model
for thermochemical calculations.
0.5
0.4
0.3
0.2
0.1
0.0
4.2.2. An assumption is thus raised for the estimation as follows
G and Z are the geopolymerization products with an approx-
imate composition of M+[AlO2·nSiO2]−·wH2O (with n close to 1
for the systems studied here, but variable depending on the start-
ing materials), in which the AlO4 and SiO4 tetrahedra link in Q4
analcime-like structures when viewed at short range. The stan-
dard formation enthalpies of G and Z are assumed to be similar,
as it has been shown that the energetics of crystalline and non-
crystalline alkali aluminosilicates are very similar [48]. Thus, these
values can be calculated using the standard enthalpies of formation
of analcime ([NaAlSi2O6]·H2O) and amorphous glass (SiO2), i.e.:
6
8
10
12
NaOH concentration (mol/L)
ꢁHf0,298(G
= ꢁHf0,298(analcime)
or Z)
Fig. 11. The reaction extent of geopolymerization systems at 25 ◦C.
+ (n − 2) · ꢁHf0,298(amorphous SiO2)
(1)
ꢁHf0,298(analcime) = −3303.2 kJ/mol
[49],
and
the incompletely dissolved MK comprises 70% of the system with
6 mol/L NaOH at 30 ◦C, and 60% at 12 mol/L NaOH and 30 ◦C. Thus,
the ICC results agree very well with the FTIR deconvolution analysis.
Granizo and Blanco [18] reported a reaction extent of around 50%
when MK was activated with NaOH solution (12 mol/L) at 35–45 ◦C;
the higher reaction extent of their metakaolin is likely to be due to
the higher temperature and the higher specific surface area of their
MK particles than those used here.
Fig. 13 shows the effects of reaction temperature on the reaction
extent, which is calculated from the ICC data based on the assump-
tion that the reaction enthalpy of Eq. (2) is constant at −358.2 kJ/mol
within the temperature range of interest here. Since stage II of the
reaction has not been completed within 72 h at 20 ◦C, only sys-
tems at temperatures of 25–40 ◦C are compared. When Na/Al = 0.98,
raising the temperature from 25 to 30 ◦C gives an increase in the
fractional reaction extent from 0.26 to 0.33. However, further rais-
ing the temperature to 40 ◦C only achieves another 0.03 increase in
reaction extent. For systems with Na/Al = 1.10, the reaction extent
only increases from 0.36 to 0.40 when changing the temperature
from 25 to 40 ◦C. This means that for a metakaolin with a given spe-
cific surface area, when activated with NaOH at Na/Al > 1, varying
reaction temperature can change the reaction rate significantly but
has only a very slight influence on its reaction extent.
where
ꢁHf0,298(amorphous SiO2) = −904.2 kJ/mol [50].
Although the zeolite crystallites observed in this study are in
fact zeolite Na–A rather than analcime, such zeolites are commonly
observed in many other studies of geopolymer synthesis [30,51],
and as mentioned above, the local structure of several geopolymers
has been found to resemble zeolites of this family. Additionally, the
thermochemical data available for analcime are considered reliable,
whereas data for zeolite A are scarce. The content of structural H2O
in a zeolite also obviously has an effect on its standard formation
enthalpy [52]. The content of structural (chemically bound) H2O
in geopolymeric gels has been found to be between 4 and 10 wt.%
[53–56], depending on the raw materials and the reaction condi-
tions. However, in this study and with the purpose of simplifying
the calculations, it is assumed to be fixed with a molar ratio of H2O
to AlO4 set equal to 1.
Accordingly, the whole process of NaOH activating MK to form
(via a multi-step process) G and Z, can be simplified into a single
overall equation:
Al2O3·2SiO2 + 2OH− + 2Na+ + H2O → 2NaAlSiO4·H2O
(2)
Here the dissolution of Si and Al from MK is assumed
to be stoichiometric. With the standard enthalpies of MK
(Al2O3·2SiO2) (−3213.4 kJ/mol) [57], OH− (−230.0 kJ/mol) [50], Na+
(−240.3 kJ/mol) [50] and H2O (−285.8 kJ/mol) [50], the reaction
enthalpy of Eq. (2) at 298 K is estimated to be −358.2 kJ/mol
(−1612.4 J/g) MK. If the ‘general’ reaction extent ˛ is then defined
as the ratio of the experimental heat release at any point in time to
this theoretical maximum heat release at full reaction, the extent of
reaction during the geopolymerization of NaOH-activated MK can
now for the first time be quantified by calorimetry.
The above results and analysis have demonstrated that ICC is a
powerful technique in studying the process of geopolymerization
as it gives real-time heat exchange information from metakaolin
100
-1
Wavenumber(cm
1204-1215
1078-1086
990-1005
)
80
60
40
20
0
944-960
852-858
4.3. Effects of NaOH concentration and temperature on the
reaction extent
Fig. 11 shows the effect of NaOH concentration on the reac-
tion extent ˛. When Na/Al is <1, increasing NaOH concentration
is effective in increasing the reaction extent. Although using ICC
to determine the reaction extent may be considered to some
degree an approximate method, the reaction extent ˛ obtained
is reasonable, according to the FTIR deconvolution analysis in
Fig. 6. If it is assumed that the resolved bands at 1204–1215
and 1078–1086 cm−1 are assigned to the Si–O–T vibration in the
incompletely dissolved MK present within the geopolymer, and
their relative areas are proportional to their quantities (Fig. 12),
6
8
10
12
NaOH concentration (mol/L)
Fig. 12. Relative area of the deconvoluted component peaks in the main Si–O–T
vibration band region (800–1300 cm−1) in geopolymer samples.

32
Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
0.5
0.4
0.3
0.2
0.1
0.0
0.5
a
b
0.4
o
40C
o
40C
o
35C
0.3
o
35C
o
30C
o
30C
o
0.2
0.1
0.0
25C
o
25C
Na:Al=1.10
Na:Al=0.98
60
0
12
24
36
48
72
0
12
24
36
48
60
72
Reaction time (h)
Reaction time (h)
Fig. 13. Effects of reaction temperature on the reaction extent in geopolymer systems as represented by the cumulative heat of reaction: (a) Na:Al = 0.98; (b) Na:Al = 1.10.
particles contacting the activator. Based on the assumptions of the
thermodynamic parameters of the reaction products, ICC can fur-
ther quantify the reaction kinetics and thus determine the reaction
extent of metakaolin particles. The effects of some important fac-
tors, such as alkali concentration and reaction temperature, on the
kinetics can be qualified by this methodology. Further investiga-
tions into more complex systems will be performed in subsequent
publications, such as for sodium silicate activation of metakaolin.
goal which has remained largely elusive to date, and will be further
developed in the future to the analysis of more complex geopoly-
mer systems, including those with silicate activating solutions.
Acknowledgements
ZZ thanks Chen Y. and Yang T. (NJUT) for their technical
assistance in ICC measurement and XRD characterization. The par-
ticipation of JLP was funded by the Australian Research Council,
including partial funding through the Particulate Fluids Processing
Centre.
5. Conclusions
The geopolymerization process and early-age reaction prod-
ucts of NaOH-activated metakaolin, reacted at 20–40 ◦C, have been
studied by the use of isothermal conduction calorimetry combined
with structural and microscopic analysis. The calorimetric curves
show either two or three distinguishable exothermic peaks after
the metakaolin is mixed with the NaOH solution. The first sharp
peak corresponds to the dissolution of metakaolin, and the second
is associated with the polymerization of dissolved species together
with the transformation of freshly formed oligomeric and/or poly-
meric products into fully cross-linked frameworks and/or zeolitic
products. The degree of crystallinity is enhanced by increasing
temperature, but not always by increasing the NaOH concentra-
tion, due to the higher solubility of aluminosilicates at very high
alkalinity. In samples which are sufficiently reactive to partially
crystallize (either as moderately large crystallites or as partially
ordered pseudo-zeolitic structures), a third exothermic peak is
observed, which is attributed to the further reorganization of the
solid structures into a more stable (although not necessarily notably
more crystalline) state. Low reaction temperatures reduce the for-
mation of zeolite phases.
Based on this mechanistic insight, and making use of the fact
that the geopolymer binder structure is known to be similar to
that of analcime-group minerals on a nanoscale level, the reaction
extent of geopolymerization has been quantified from the isother-
mal conduction calorimetry data. Fractional reaction extents of up
to 0.4 are observed at 72 h after mixing, and the results of deconvo-
lution of infrared spectra are consistent with the values obtained
from the calorimetric data. Increasing the NaOH concentration from
Na/Al = 0.74 to 1.10 leads to an increase in the reaction extent. How-
ever, the addition of more NaOH beyond this point is not helpful
in improving the reaction extent or shortening the main reaction
period as indicated by the time at which the first two peaks are
observed. Raising the reaction temperature from 25 ◦C to 40 ◦C
accelerates the reaction, but shows little effect on the final reaction
extent, particularly when Na/Al > 1. The methodology developed
here shows good potential in providing an accurate measure of
the extent of reaction of metakaolin-based geopolymer systems, a
References
[1] P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van
Deventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42
(2007) 2917–2933.
[2] B.C. McLellan, R.P. Williams, J. Lay, A. van Riessen, G.D. Corder, Costs and carbon
emissions for geopolymer pastes in comparison to ordinary Portland cement,
J. Cleaner Prod. 19 (2011) 1080–1090.
[3] A.M. Rashad, S.R. Zeedan, The effect of activator concentration on the residual
strength of alkali-activated fly ash pastes subjected to thermal load, Constr.
Build. Mater. 25 (2011) 3098–3107.
[4] Z. Zhang, X. Yao, H. Zhu, Potential application of geopolymers as protection
coatings for marine concrete. I. Basic properties, Appl. Clay Sci. 49 (2010) 1–6.
[5] J.W. Phair, J.S.J. van Deventer, Effect of silicate activator pH on the leaching and
materials characteristics of waste-based inorganic polymers, Miner. Eng. 14
(2001) 289–304.
[6] U. Rattanasak, P. Chindaprasirt, Influence of NaOH solution on the synthesis of
fly ash geopolymer, Miner. Eng. 22 (2009) 1073–1078.
[7] E. Álvarez-Ayuso, X. Querol, F. Plana, A. Alastuey, N. Morenoa, M. Izquierdo, O.
Font, T. Morenoa, S. Diez, E. Vázquez, M. Barra, Environmental, physical and
structural characterisation of geopolymer matrixes synthesised from coal (co-
)combustion fly ashes, J. Hazard. Mater. 154 (2008) 175–183.
[8] J.L. Provis, C.A. Rees, Geopolymer synthesis kinetics, in: J.L. Provis, J.S.J. van
Deventer (Eds.), Geopolymers: Structure, Processing, Properties and Industrial
Applications, Woodhead Publishing, Abingdon, UK, 2009, pp. 118–136.
[9] T. Ejaz, A.G. Jones, P. Graham, Solubility of zeolite A and its amorphous precursor
under synthesis conditions, J. Chem. Eng. Data 44 (1999) 574–576.
[10] J.I. Bhatty,
A
review of the applications of thermal analysis to
cement–admixture systems, Thermochim. Acta 189 (1991) 313–350.
[11] A. Fernández-Jiménez, F. Puertas, A. Arteaga, Determination of kinetic equa-
tions of alkaline activation of blast furnace slag by means of calorimetric data,
J. Therm. Anal. 52 (1998) 945–955.
[12] S. Alonso, A. Palomo, Alkaline activation of metakaolin and calcium hydroxide
mixtures: influence of temperature, activator concentration and solids ratio,
Mater. Lett. 47 (2001) 55–62.
[13] M.S. Mun˜iz-Villarreal, J.L. Reyes-Araiza, S. Sampieri-Bulbarela, J.R. Gasca-
Tirado, A. Manzano-Ramírez, J.C. Rubio-Ávalos, J.J. Pérez-Bueno, L.M. Apatiga,
A.Z. Cadena, V. Amigó-Borrás, The effect of temperature on the geopolymer-
ization process of a metakaolin-based geopolymer, Mater. Lett. 65 (2011)
995–998.
[14] H. Rahier, B. van Mele, M. Biesemans, J. Wastiels, X. Wu, Low-temperature syn-
thesized aluminosilicate glasses. 1. Low-temperature reaction stoichiometry
and structure of a model compound, J. Mater. Sci. 31 (1996) 71–79.
[15] X. Yao, Z. Zhang, H. Zhu, Y. Chen, Geopolymerization process of alkali-
metakaolinite characterized by isothermal calorimetry, Thermochim. Acta 493
(2009) 49–54.

Z. Zhang et al. / Thermochimica Acta 539 (2012) 23–33
33
[16] A. Buchwald, R. Tatarin, D. Stephan, Reaction progress of alkaline-activated
metakaolin-ground granulated blast furnace slag blends, J. Mater. Sci. 44 (2009)
5609–5617.
[37] Y. Zhang, W. Sun, Z. Li, Infrared spectroscopy study of structural nature of
geopolymeric products, J. Wuhan Univ. Technol. 23 (2008) 522–527.
[38] I. Giannopoulou, D. Panias, Hydrolytic stability of sodium silicate gels in the
presence of aluminum, J. Mater. Sci. 45 (2010) 537–5377.
[39] S.A. Bernal, J.L. Provis, V. Rose, R. Mejía de Gutiérrez, Evolution of binder struc-
ture in sodium silicate-activated slag-metakaolin blends, Cem. Concr. Compos.
33 (2011) 46–54.
[40] W.K.W. Lee, J.S.J. van Deventer, Use of infrared spectroscopy to study geopoly-
merization of heterogeneous amorphous aluminosilicates, Langmuir 19 (2003)
8726–8734.
[41] P. Duxson, J.L. Provis, G.C. Lukey, S. Mallicoat, W.M. Kriven, J.S.J. van Deventer,
Understanding the relationship between geopolymer composition microstruc-
ture and mechanical properties, Colloids Surf., A 269 (2005) 47–58.
[42] S. Chandrasekhar, P.N. Pramada, Microwave assisted synthesis of zeolite A from
metakaolin, Microporous Mesoporous Mater. 108 (2008) 152–161.
[43] J.L. Provis, P. Duxson, J.S.J. van Deventer, G.C. Lukey, The role of mathematical
modelling and gel chemistry in advancing geopolymer technology, Chem. Eng.
Res. Des. 83 (A7) (2005) 853–860.
[44] S. Mintova, N.H. Olson, V. Valtchev, T. Bein, Mechanism of zeolite A nanocrystal
growth from colloids at room temperature, Science 283 (1999) 958–960.
[45] J.L. Bell, P. Sarin, P.E. Driemeyer, R.P. Haggerty, P.J. Chupas, W.M. Kriven, X-
ray pair distribution function analysis of a metakaolin-based, KAlSi2O6·5.5H2O
inorganic polymer (geopolymer), J. Mater. Chem. 18 (2008) 5974–5981.
[46] C.E. White, J.L. Provis, T. Proffen, J.S.J. van Deventer, The effects of tem-
perature on the local structure of metakaolin-based geopolymer binder: a
neutron pair distribution function investigation, J. Am. Ceram. Soc. 93 (2010)
3486–3492.
[17] H. Rahier, W. Simons, B. Van Mele, M. Biesemans, Low-temperature synthesized
aluminosilicate glasses. 3. Influence of the composition of the silicate solution
on production, structure and properties, J. Mater. Sci. 32 (1997) 2237–2247.
[18] M.L. Granizo, M.T. Blanco, Alkaline activation of metakaolin. An isothermal
conduction calorimetry study, J. Therm. Anal. 52 (1998) 957–965.
[19] M.L. Granizo, M.T. Blanco-Varela, A. Palomo, Influence of the starting kaolin on
alkali-activated materials based on metakaolin. Study of the reaction parame-
ters by isothermal conduction calorimetry, J. Mater. Sci. 35 (2000) 6309–6315.
[20] H. Rahier, J.F. Denayer, B. van Mele, Low-temperature synthesized aluminosil-
icate glasses. Part IV. Modulated DSC study on the effect of particle size of
metakaolinite on the production of inorganic polymer glasses, J. Mater. Sci. 38
(2003) 3131–3136.
[21] I. Pane, W. Hansen, Investigation of blended cement hydration by isothermal
calorimetry and thermal analysis, Cem. Concr. Res. 35 (2005) 1155–1164.
[22] C.A. Rees, J.L. Provis, G.C. Lukey, J.S.J. van Deventer, In situ ATR-FTIR study
of the early stages of fly ash geopolymer gel formation, Langmuir 23 (2007)
9076–9082.
[23] A. Hajimohammadi, J.L. Provis, J.S.J. van Deventer, Time-resolved and spatially-
resolved infrared spectroscopic observation of seeded nucleation controlling
geopolymer gel formation, J. Colloid Interface Sci. 357 (2011) 384–392.
[24] J.L. Provis, J.S.J. van Deventer, Geopolymerisation kinetics. 1. In situ energy-
dispersive X-ray diffractometry, Chem. Eng. Sci. 62 (2007) 2309–2317.
[25] J.L. Provis, J.S.J. van Deventer, Geopolymerisation kinetics. 2. Reaction kinetic
modeling, Chem. Eng. Sci. 62 (2007) 2318–2329.
[26] J.L. Provis, P.A. Walls, J.S.J. van Deventer, Geopolymerisation kinetics. 3. Effects
of Cs and Sr salts, Chem. Eng. Sci. 63 (2008) 4480–4489.
[47] J.L. Bell, P. Sarin, J.L. Provis, R.P. Haggerty, P.E. Driemeyer, P.J. Chupas, J.S.J. van
Deventer, W.M. Kriven, Atomic structure of a cesium aluminosilicate geopoly-
mer: a pair distribution function study, Chem. Mater. 20 (2008) 4768–4776.
[48] A. Navrotsky, Z.R. Tian, Sytematics in the enthalpies of formation of anhydrous
aluminosilicate zeolite, glass and dense phases, Chem. Eur. J. 7 (2001) 769–774.
[49] P.S. Neuhoff, G.L. Hovis, G. Balassone, J.F. Stebbins, Thermodynamic properties
of analcime solid solutions, Am. J. Sci. 304 (2004) 21–66.
[50] Common Thermodynamic Database Project, http://www.ctdp.org/#http
%3A//www.ctdp.org/ctd import/species list/species 3785/view [25.08.2011].
[51] J. Wang, C. Han, J. Wang, Y. Li, Y. Teng, X. Wu, Composition and microstructure
of ‘alkali-slag-coal fly ash-metakaolin’ hydroceramics, J. Chin. Ceram. Soc. 39
(2011) 512–517.
[27] C.E. White, J.L. Provis, T. Proffen, J.S.J. van Deventer, Molecular mechanisms
responsible for the structural changes occurring during geopolymerization:
multiscale simulation, AIChE J., http://dx.doi.org/10.1002/aic.12743, in press.
[28] C.E. White, J.L. Provis, A. Llobet, T. Proffen, J.S.J. van Deventer, Evolution of
local structure in geopolymer gels: an in situ neutron pair distribution function
analysis, J. Am. Ceram. Soc. 94 (2011) 3532–3539.
[29] P. Rovnaník, Effect of curing temperature on the development of hard structure
of metakaolin-based geopolymer, Constr. Build. Mater. 24 (2010) 1176–1183.
[30] D. Akolekar, A. Chaffee, F.H. Russell, The transformation of kaolin to low-silica
X-zeolite, Zeolites 19 (1997) 359–365.
[31] M.L. Granizo, S. Alonso, M.T. Blanco-Varela, A. Palomo, Alkaline activation of
metakaolin. Effect of calcium hydroxide in the products of reaction, J. Am.
Ceram. Soc. 85 (2002) 225–231.
[32] S. Kumar, R. Kumar, S.P. Mehrotra, Influence of granulated blast furnace slag
on the reaction, structure and properties of fly ash based geopolymer, J. Mater.
Sci. 45 (2010) 607–615.
[33] A. Hajimohammadi, J.L. Provis, J.S.J. van Deventer, The effect of silica availability
on the mechanism of geopolymerization, Cem. Concr. Res. 41 (2011) 210–216.
[34] R.M. Barrer, D.E. Mainwaring, Chemistry of soil minerals. Part XIII. Reactions
of metakaolinite with single and mixed bases, J. Chem. Soc. Dalton Trans. 22
(1972) 2534–2546.
[52] G.K. Johnson, H.E. Flotow, P.A.G. O’Hare, W.S. Wise, Thermodynamic studies of
zeolites: analcime and dehydrated analcime, Am. Miner. 67 (1982) 736–748.
[53] D. Dimas, I. Giannopoulou, D. Panias, Polymerization in sodium silicate solu-
tions: a fundamental process in geopolymerization technology, J. Mater. Sci. 44
(2009) 3719–3730.
[54] Z. Zhang, X. Yao, H. Zhu, Y. Chen, Role of water in the synthesis of calcined
kaolin-based geopolymer, Appl. Clay Sci. 43 (2009) 218–223.
[55] D.L.Y. Kong, J.G. Sanjayan, K. Sagoe-Crentsil, Factors affecting the performance
of metakaolin geopolymers exposed to elevated temperatures, J. Mater. Sci. 43
(2008) 824–831.
ˇ
[56] V. Zivica, S. Balkovic, M. Drabik, Properties of metakaolin geopolymer hardened
[35] C.A. Ríos, C.D. Williams, M.A. Fullen, Nucleation and growth history of zeolite
LTA synthesized from kaolinite by two different methods, Appl. Clay Sci. 42
(2009) 446–454.
[36] M. Alkan, C. Hopa, Z. Yilmaz, H. Guler, The effect of alkali concentration and
solid/liquid ratio on the hydrothermal synthesis of zeolite NaA from natural
kaolinite, Microporous Mesoporous Mater. 86 (2005) 176–184.
paste prepared by high-pressure compaction, Constr. Build. Mater. 25 (2011)
2206–2213.
[57] N.C. Schieltz, M.R. Soliman, Thermodynamics of the various high temper-
ature transformations of kaolinite, in: Proceedings of the 13th National
Conference on Clays and Clay Minerals, Pergamon Press, New York, 1966,
pp. 419–428.