Cement and Concrete Research 70 (2015) 39–49

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: http://ees.elsevier.com/CEMCON/default.asp

Co-existence of aluminosilicate and calcium silicate gel characterized

through selective dissolution and FTIR spectral subtraction
Sravanthi Puligilla, Paramita Mondal ⁎
Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801, United States

a r t i c l e i n f o a b s t r a c t

Article history: Alkali activation of a fly ash and slag mixture supposedly forms a combination of aluminosilicate and calcium
Received 18 July 2014 silicate hydrate gel as reaction products. Fourier transform infrared spectroscopy (FTIR) is often used to charac-
Accepted 6 January 2015 terize the complex nature of the reaction products. However, interpretation of FTIR results is difficult as the Si–O–
Available online 21 January 2015
T (T = Al, Si) vibrations from the initial ash, geopolymer product, unreacted slag, and calcium silicate hydrate
yield an overlapping spectrum. This paper reports how the issue can be resolved by selectively dissolving calcium
Keywords:
Alkali activated cement (D)
silicate hydrate and geopolymer product, and subtracting FTIR spectra of the insoluble residue from the original
Fly Ash (D) material to determine IR spectrum of the material dissolved.
Granulated Blast-Furnace Slag (D) © 2015 Elsevier Ltd. All rights reserved.
Calcium-Silicate-Hydrate (C-S-H) (B)
Spectroscopy (B)

1. Introduction in the microstructure is not found in SEM-EDS studies [8,9]. Instead,

Geopolymers are defined as amorphous to semi-crystalline three
dimensional aluminosilicate structures with Si/Al ratios ranging from
1 to 3 [1]. When geopolymers are synthesized and cured at room
temperature, they are mostly amorphous unlike when they are hydro-
thermally treated [1]. Classical characterization techniques such as X-
ray diffraction (XRD) produce a broad halo rather than sharp diffraction
peaks for such materials. Therefore, their structure cannot be investigat-
ed using XRD alone. Advanced techniques such as Fourier transform
infrared spectroscopy (FTIR), and nuclear magnetic resonance spectros-
copy (MAS-NMR) in conjunction with XRD, provides greater insight
into the molecular framework.
Phase characterization becomes more complicated for geopolymers
made from precursors (aluminosilicate sources) that contain some
amount of calcium. Various phases form as reaction products, and
their response to various characterization techniques is difficult to iso-
late. According to several researchers, C–A–S–H, (C,(N,K))–A–S–H, and
(N,K)–A–S–H are known to coexist as reaction products in fly ash/
metakaolin geopolymers in the presence of any soluble form of calcium
[2–7]. The co-existence of C–A–S–H and (N,K)–A–S–H in the micro-
structure has been studied by scanning electron microscopy, elemental
analysis through energy dispersive spectroscopy (SEM-EDS), and XRD
[4–6,8,9]. Presence of C–S–H has been suggested by the synchrotron
XRD results and by average elemental analysis on several points of a
geopolymer microstructure [4,9]. However, a segregated C–S–H phase
⁎ Corresponding author at: 205 N. Mathews Ave., Urbana IL 61801-2352.
E-mail addresses: puligil1@illinois.edu (S. Puligilla), pmondal@illinois.edu (P. Mondal).
findings show that it is always intermixed with geopolymer phases,
possibly owing to a large interaction volume of EDS [6,8,9]. Differentia-
tion between phases is difficult if they are smaller than the interaction
volume which is mostly the case in early age samples. Furthermore,
EDS study involves time-consuming sample preparation that can be
avoided using an infrared spectroscopy method.
Fourier transform infrared spectroscopy has been used extensively
to characterize synthesized geopolymer gel and to monitor the reaction
progress of alkali-activated fly ash, fly ash-slag, metakaolin, and
metakaolin-slag geopolymers [10–19]. However, characterization and
interpretation of the results is difficult as the Si–O–T (T = Al, Si) vibra-
tions from the initial ash, geopolymer product, unreacted slag, and cal-
cium silicate hydrate yield an overlapping spectrum [10]. Researchers
faced a similar issue characterizing portland cements and their reaction
products using an XRD technique [20,21]. The mineral phases in
portland cement and clinker have overlapping peaks in XRD that
makes the characterization difficult. So, cement and clinker samples
have been traditionally subjected to various selective dissolution tech-
niques in order to remove certain phases and to amplify the response
from others for a more accurate characterization. The knowledge gained
through these selective dissolution studies helped to eliminate any
ambiguity in the phase analysis of cement/clinker and to establish
characterization using other methods such as XRD-Rietveld. Salicylic
acid-methanol (SAM) extraction, initially developed by Takashima,
gained popularity for dissolving amorphous calcium silicate phases,
and it is widely adopted for cement and clinker research [20–22]. This
method was applied more extensively later to other materials like
cement-slag, cement-fly ash blends, alkali-activated slag systems, and
synthesized C–S–H gels in order to determine the reaction degree and

http://dx.doi.org/10.1016/j.cemconres.2015.01.006
0008-8846/© 2015 Elsevier Ltd. All rights reserved.
40 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49

the nature of the reaction products [23–26]. However, when SAM ex-
traction was performed on a mixture of synthesized N–A–S–H and C–
S–H, FTIR characterization failed to differentiate between them, which
led to inconclusive results [11].
Several researchers also used 1:20 HCl to separate geopolymers and
zeolites from unreacted fly ash to determine the reaction degree of fly
ash, and 1:9 HCl solution followed by Na2CO3 to determine the reaction
degree of metakaolin geopolymer [10–12,27]. However, the presence of
calcium complicates the response of alkali activated systems (due to a
complex combination of products) to HCl extraction. Very few studies
have used both SAM and HCl extractions on a complex geopolymer sys-
tem prepared from precursors with a soluble form of calcium to charac-
terize reaction products [12]. As the development of geopolymers
suitable for practical applications becomes important, there is an in-
creasing need for rigorous characterization of binders often made
from blends of fly ash-slag and metakaolin-slag. This paper elucidates Fig. 1. XRD pattern of raw materials (fly ash and slag) where Q = Quartz, M = Mullite and
the importance of using SAM and HCl extractions together to character- H = Hematite. Both raw materials show broad amorphous hump. The main crystalline
ize geopolymers formed in a fly ash-slag bended system through infra- phases observed in fly ash are quartz, mullite and hematite.
red spectroscopic studies. The method used can be extended to other
blended systems where geopolymer forms in the presence of a soluble
calcium source. (Fig. 2). Spectra are analyzed for an asymmetric stretching vibration of
Early and late age geopolymer samples are studied using infrared a Si–O–T bond, which lies between 1100 and 900 cm−1. Hereafter, the
spectroscopy after selectively dissolving C–S–H or C–A–S–H through Si–O–T asymmetric stretching vibration band will be referred to as the
SAM extraction and (N,K)–A–S–H through HCl extraction. FTIR analysis main band.
is conducted before any chemical extraction and also on solid residues The position of the main band in fly ash appears at 1037 cm−1. The
after SAM and HCl extractions. As the FTIR spectra collected after main band of slag is broad and appears between 850 and 930 cm−1. The
selective dissolution can still be difficult to interpret, spectral subtrac- difference in the position of the main band in both materials is
tion is performed for complete analysis. To further corroborate the accounted for by the difference in the nature of aluminosilicate glass.
results, isothermal calorimetry, scanning electron microscopy (SEM), The vibration spectra of vitreous or amorphous materials, unlike crystal-
and mechanical strength testing have been performed on different line materials, are very broad and imprecisely defined as evident in
geopolymer mixes. Fig. 2. In addition to the main band, fly ash spectrum also shows peaks
at 795 cm−1 and 777 cm− 1 which are attributed to quartz [10]. Slag
2. Materials and methods spectrum has another peak at 1487 cm− 1 which corresponds to an
asymmetric CO stretch. The position of the main band in LS spectra is
The materials used in this study were class-F fly ash (supplied by at 1030 cm−1, and the positions of the remaining peaks are similar to
Boral Material Technologies), slag (Lafarge Corporation), potassium hy- the fly ash spectrum. The position of main band in the HS spectrum is
droxide (Fisher Scientific), and potassium silicate (produced by Pfaltz at 1020 cm−1 with remaining peaks very similar to that of the fly ash
and Bauer, and supplied by Fisher Scientific). Keeping the concentration spectrum. Although slag spectrum shows carbonation, when fly ash
of the activating solution constant (2.3 M), fly ash was partially replaced and slag are added proportionally to obtain LS and HS spectra, carbon-
by slag (25% and 15% by weight). Geopolymer was synthesized by ation peaks are insignificant. This is believed to be due to the small
mixing fly ash and slag with the activator solution made from potassium intensity of carbonation peaks relative to the main band.
hydroxide and potassium silicate of SiO2/K2O = 1.25 (water/solids = Salicylic acid/methanol (SAM) extraction was performed on samples
0.32). Samples at various ages were collected and powdered with ace- following the procedure described by Stutzman [20]. The attack dis-
tone to stop the reaction and were stored in a vacuum desiccator until solves calcium silicate hydrate but is not supposed to dissolve unreacted
testing. Selective dissolution extractions were followed by characteriza- fly ash, slag or geopolymers [11,28]. For SAM extraction, 1 g of pow-
tion using FTIR, XRD, and SEM on dried-powdered samples within 24 h dered geopolymer sample was added to a solution containing 4 g of
after being treated with acetone.
X-ray diffractograms were recorded on a Siemens-Bruker D 5000
using CuKα radiation at a voltage of 40 kV and a current of 30 mA.
Specimens were scanned from 5 to 70°2θ at 0.02°2θ steps and stepped
at a rate of 1 min− 1. The X-ray diffraction patterns of raw materials
are presented in Fig. 1. The fly ash spectrum shows the main peaks of
quartz, mullite, and hematite. The broad amorphous hump in the XRD
pattern, which is a characteristic of fly ash, is observed between 20°
and 40°2θ. Slag is an amorphous glass, as evidenced by its XRD pattern
with a hump centered at 30°2θ.
A Thermo Nicolet Nexus 670 Fourier transform infrared spectropho-
tometer was used in absorption mode to acquire an infrared spectrum
of samples at various ages. A golden gate attenuated total reflectance
accessory (ATR) with diamond crystal was used. The frequency range
was 2000 to 700 cm−1 at a resolution of 2 cm−1. Each spectrum is an
average of 64 scans. FTIR spectra for raw materials are given in Fig. 2.
Hereafter, the mix with 25% slag–75% fly ash will be referred to as
high slag sample (HS), and the mix with 15% slag–85% fly ash will be re-
ferred to as low slag sample (LS). IR spectra of these blends are obtained Fig. 2. FTIR spectra of raw materials. FA: Fly ash, slag, LS: 85% FA + 15% slag, and HS: 75%
by proportionally adding an individual spectrum of raw materials FA + 25% slag. The number in parentheses denotes the peak position in cm−1.
 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
Fig. 3. IR spectrum of the raw material and the insoluble residue after SAM treatment on
slag. The number in parentheses denotes the peak position in cm−1.
salicylic acid mixed in 60 ml of methanol [20]. The mixture was stirred
for 2 h and the suspension was vacuum filtered using a Buchner funnel
and a Whatman filter (0.2 μm pore size). Insoluble residue was washed
with methanol and stored in vacuum desiccator until analysis. IR spectra
of these specimens were collected to observe the changes following
extraction.
For another extraction method, finely ground powder samples were
treated with 1:20 HCl (by volume). Acid attack provokes the dissolution
of chief reaction products of alkali activated fly ash (aluminosilicate gel
and zeolites) and leaving behind the unreacted ash as insoluble residue
[10,27]. However, acid extraction also decomposes calcium silicate
hydrate by removing Ca2 + and leaving silica gel behind [21]. As
described elsewhere, the experimental procedure consists of adding
1 g of activated fly ash (fly ash and slag in this case) to a beaker contain-
ing 250 ml of HCl (1:20) [10]. The mixture was stirred for 3 h and
followed by filtration. The insoluble residue was washed with de-
ionized water several times to a neutral pH, dried at 100 °C for 24 h
and then stored in a vacuum desiccator until the analysis was finished.
FTIR spectra before and after chemical extractions were analyzed and
spectral manipulations including addition and subtraction were
performed using OMNIC software. Selective dissolutions were repeated
3 times on all of the samples.
Reaction progress in HS and LS samples was studied using Thermo-
metric 3114/3236 TAM Air Isothermal Conduction Calorimeter (ICC) at
22 °C. All of the raw materials and the activator solution were stored at
22 °C before mixing. Solids were mixed with the activator solution out-
side of the instrument and the paste was transferred to the instrument
within 3-4 min after the mixing. Due to the heat generated from friction
Fig. 4. IR spectrum of the raw material and the insoluble residue of fly ash after the SAM
treatment, HCl treatment, and HCl treatment on insoluble residue after SAM treatment.
The number in parentheses denotes the peak position in cm−1.
41
Table 1
Analysis of selective dissolution treatments presented as dissolved % by the initial mass.
Dissolved Dissolved % insoluble Dissolved in HCl
in SAM in HCl residue after SAM treatment
FA 0.97 3.14 96.86 2.70
Slag 1.85 100.00 0.00 98.15
IR: insoluble residue.
while placing the samples inside the instrument, the first 45 min of data
was ignored.
Microstructural development of HS and LS samples was monitored
by using a JEOL JSM-6060 LV Scanning Electron Microscope (SEM),
equipped with an energy dispersive X-ray spectroscopy (EDS) at an ac-
celerating voltage of 15 kV. The fractured surfaces of an HS sample at
1 h, 24 h, and 7 days and an LS sample at 7 days were studied in second-
ary electron mode (SE). The specimen surface was sputter coated with
gold-palladium prior to imaging.
3. Results and discussion
3.1. Verification of the effects of SAM and HCl extractions on raw materials
The effect of SAM and HCl extractions on raw materials (fly ash and
slag) has been verified before studying the effect of extraction on the
activated fly ash-slag samples. FTIR characterization of 100% slag and
fly ash before and after selective dissolutions is presented in Figs. 3
and 4, respectively.
3.1.1. Slag
The effect of SAM on slag is debatable [24,25]. This topic has been
studied widely in the past. SAM is reported to partly dissolve slag and
its reaction products such as hydrotalcite and calcium silicate hydrate
[24,25]. In this study, a weight loss of approximately 1.85% was
observed when slag was treated with SAM. The IR spectra before
(peak position: 850–930 cm− 1) and after (peak position: 850–
926 cm−1) treatment are very similar to each other as evident in
Fig. 3. Considering the small weight loss and the negligible shift in
peak position, the treatment is considered to be suitable for selectively
dissolving certain phases and enriching the remaining phases for char-
acterization. The work conducted by other researchers, using similar
treatments like EDTA, indicated that such chemical treatments enrich
the response of undissolved phases for characterization, using methods
like NMR [25,26].
Slag dissolves completely when treated with HCl (1:20) solution.
The dissolution of slag is considered congruent in this study as evident
Fig. 5. XRD patterns of HS sample at various ages (1 h, 24 h, 7 days). Inset panel shows a
close up of XRD patterns between 27 and 35°2θ angle. Q: Quartz, M: Mullite, H: Hematite.
42 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49

in cement blends and alkali activated slag systems in other studies [29,
30]. Therefore, unreacted slag at various ages in fly ash-slag geopolymer
responds to SAM and HCl extractions in the same way as 100% slag
would respond.

3.1.2. Fly ash

The effect of HCl and SAM on fly ash was studied by researchers and
is not as debatable as slag [31,32]. In this study, fly ash was subjected to
SAM and HCl treatments and the weights dissolved in each of these
treatments are presented in Table 1. The amount of fly ash dissolved
in SAM treatment is less than 1%, which matches with existing literature
[32,33]. The amount of fly ash dissolved after HCl treatment is also small
(~ 3%). Insoluble residue of fly ash after SAM treatment was further
treated with HCl (dissolved in HCl after SAM treatment). The amount
of fly ash dissolved in both of these treatments combined is also small
(~3%).
The FTIR spectrum of fly ash before and after treatments (SAM, HCl) Fig. 7. Heat released from HS and LS samples at 22 °C.
is given in Fig. 4. The position of the main band in fly ash remains at
1037 cm−1 after SAM treatment, implying that the effect of this treat-
ment is minimal. The fly ash residue obtained after SAM extraction Thus, the response of both raw materials is assumed to be congruent to
was further treated with HCl (HCl residue after SAM), and it was SAM/HCl attack at all ages.
observed that the effect was minimal (small weight loss and no change
in peak position). However, the position of the main band (1037 cm−1) 3.2. Evolution of reaction products with time
changes to a higher wavenumber (1055 cm−1) after HCl treatment. This
implies dissolution of certain phases from fly ash which matches with The compressive strengths of LS and HS sample at 7 days are 13.9 ±
the observations of Sarbak et al. [31]. Sarbak et al. have concluded that 1.0 MPa and 28.8 ± 3.3 MPa, respectively (average of three measure-
the treatment of fly ash with 2.5 M HCl solution changed its surface ments on 25.4 mm cubes). The significant difference in the strength
area and affected the microstructure significantly due to dissolution of between these samples may be due to differences in the extent of
aluminum-rich components [31]. Owing to the small amount of weight geopolymer product formation or the chemical makeup of the reaction
loss, the effect of HCl on fly ash is considered to be minimal. products formed. Different chemical and microstructural characteriza-
The dissolution of fly ash is considered congruent in geopolymers tion methods such as calorimetry, SEM, XRD, selective dissolution treat-
after the early release of Al-rich species [34]. In fly ash-slag ments, and FTIR were performed on HS samples at various ages to
geopolymers, the dissolved Al species will probably be consumed monitor the evolution of reaction products, and on an LS sample at
during the precipitation of C–A–S–H. This prevents Al species in the 7 days to explain the reasons behind the differences in strength.
solution from retarding further ash dissolution and accelerates the
onset of congruent ash dissolution [34]. Also, at early ages, when the 3.2.1. XRD
solution has a high pH (above 13) and the solution is undersaturated, Fig. 5 shows an XRD pattern for the HS sample at 1 h, 24 h, and
the dissolution of synthetic (calcium)aluminosilicate glasses (including 7 days. XRD patterns at various ages resemble that of the fly ash
compositions similar to fly ash and slag) is known to be congruent [30]. and are similar to each other. However, close and careful observation

Fig. 6. Secondary electron images of a: HS 1 h, b: HS 24 h, c: HS 7 days, d: LS 7 days.

 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49 43

presence of calcite, and the hump could indicate the existence of C–S–
H (the amount of which is increasing with time). However, the presence
of geopolymer could not be confirmed from the XRD patterns. Since the
XRD results were not conclusive on the HS samples, the method was not
used to characterize the LS samples.

3.2.2. SEM
The SEM micrographs of the HS sample at 1 h shows fly ash spheres
covered with small amounts of products and minimal connectivity
(Fig. 6). The nature of the microstructure could be described similarly
at 24 h (Fig. 6). As the reaction proceeds, at 7 days, a greater amount
of product was seen that covered most of the fly ash particles complete-
ly. This implies greater connectivity between the fly ash spheres and
greater load carrying capacity (compressive strength). The microstruc-
ture seen here resembles the fly ash geopolymers made in the presence
of Ca as observed by other researchers [6,35]. The microstructure of the
LS sample at 7 days is similar to that of the HS sample at 7 days. A slight
difference can be observed, if the microstructure of the LS sample is
studied carefully. The reaction product in many places is nearly trans-
parent potentially due to the relatively thin layer of product formed,
and the fly ash spheres underneath are visible. The difference between
the two micrographs is qualitative, and more substantial evidence that
can explain the difference in the 7 day compressive strength was
pursued.

Fig. 8. Evolution of product with time using FTIR for a: HS, b: LS.
reveals that there are new peaks in the 27–35°2θ region highlighted in
Fig. 5 (inset panel). These XRD patterns were collected at a rate of
0.15 min−1 in order to increase the signal to noise ratio.
The peak at 29.3°2θ becomes sharper and more defined with time.
In addition to the sharp peak, a hump was also observed in the 7 days
sample at the same 2θ angle. The sharp peak might correspond to the
3.2.3. Calorimetry
Fig. 7 presents the heat evolution of the HS and LS samples at 22 °C,
which compares well with the published literature on alkali-activated
slag systems, metakaolin-slag/CH blends, and fly ash-slag blends [7,
36–38]. Calorimetric curves of HS and LS samples have similar patterns.
An initial peak was followed by an induction period and by a prominent
second peak. Apart from the heat generated from friction during the
placement of the ampoules into the instrument, the initial peak also cor-
responds to the particle wetting, initiation of slag dissolution, and the
complexation of silicate units with calcium, sodium, and aluminum [7,
36–38]. From a previous study on a similar system, the authors noted
that this peak also corresponds to the precipitation of the initial reaction
products, which caused rapid hardening [6]. It has been hypothesized
that the precipitation of C–S–H causes initiation and an acceleration of
hardening during the first 60–80 min [3,6,8]. This hypothesis will be
verified in the present paper through an FTIR study. The second main

Table 2
Position of IR spectrum bands in HS samples in the original spectrum and after selective dissolution treatments. Bands in the corresponding subtraction spectrum are also given.

Band (cm−1) HS sample

Original SAMresidue HCl residue HCl residue Dissolved in SAM Dissolved in HCl Dissolved in HCl
after SAM after SAM

1h
Si-O-T asymmetric 999 1024, 910(s) 1050 1054 978, 860(s), 1107(s) 980, 916–815(bs), 868–1011(b), 1100(s)
stretching 1110(s)
Carbonate 1395, 1467 Absent Absent Absent 1395, 1467, 866 1395, 1467, 867 Absent

24 h
Si–O–T asymmetric 970 1026, 910(s) 1040 1050 952, 834(s), 1114 950, 970(s), 834(bs), 1000, 860–910(bs), 1100(s)
stretching 1110
Carbonate 1400, 1465, 866 Absent Absent Absent 1400, 1465, 866 1400, 1465, 866 Absent

7 day
Si–O–T asymmetric 969, 1020(s) 1030, 910(s) 1050, 922(s) 1060 960, 1105(s) 960, 1130(s) 992, 1020(s), 860–940(bs),
stretching 1145
Carbonate 1415, 1465, 872 Absent Absent Absent 1415, 1465, 872 1415, 1465, 872 Absent

LS 7 day
Si–O–T asymmetric 965, 910(s), 1024(s) 1040, 910(s) 1040, 920(s) 960, 1110(s) 1066–1040(b), 1111(s)
stretching
Carbonate 1387, 1485 Absent Absent 1387, 1485 Absent

s: shoulder.
b: broad.
44 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49

peak (acceleration and deceleration) corresponds to the precipitation of geopolymers (N/K–A–S–H) do not dissolve in SAM [11,12,23,28,42]. In
the reaction products, which was hypothesized to be both geopolymer this paper, geopolymer that is formed in the presence of calcium ion is
and C–S–H. A less prominent third peak is evident at a later age. These also considered to be insoluble in SAM because no evidence of its solu-
observations are common for both HS and LS systems. The difference bility in SAM is noted in the literature [11]. Therefore, it is concluded
between LS and HS samples is the time at which the second peak and that the shift observed in Fig. 9 is due to the dissolution of calcium
third peak appear, and the area under the curves. The time at which silicate hydrate [11,23,28,32].
the second peak appears depends heavily on the amount of slag in the As shown in Fig. 9, HCl treatment on samples of all ages shifts the
system [6,36]. The early appearance of the second peak might indicate main band to 1050 cm− 1, which matches with the peak position in
a faster reaction in the HS sample compared to the LS sample. The the fly ash spectrum after HCl treatment. When powdered samples
area under the curve is greater in HS than LS sample, which indicates are treated with 1:20 HCl solution, products like geopolymers and
a greater reaction degree in the HS sample. This could explain the differ- zeolites dissolve along with any unreacted slag. A negligible portion of
ence in the compressive strength at 7 days. unreacted fly ash dissolves during HCl treatment as well (~ 3%). HCl
treatment also dissolves calcium carbonate, and it affects calcium
3.2.4. Infrared spectroscopy
Infrared spectroscopy was used to monitor the evolution of reaction
products with time in the HS and LS samples and is presented in Fig. 8.
The position of the main band at 0 h (at 1020 cm−1 and 1030 cm−1 in
HS and LS, respectively) associated with Si–O–T asymmetric stretching
vibration in the raw material moves to a lower frequency (to
969 cm− 1 and 965 cm− 1, respectively) in 7 days. The band shift to
lower frequencies is due to the precipitation of the geopolymer with
an increased substitution of Si with Al [10,27,39]. The Si–O–T angle
reduces when Al substitutes for Si and when the bond force constant
of Si–O–Al is smaller than that of the Si–O–Si bond, which shifts the
main band to a lower frequency [10,39]. The exact position of this
band depends on the Al/Si ratio of the product [10]. The peaks
(794 cm−1 and 774 cm−1) belonging to quartz are present in the re-
maining unreacted fly ash. The IR spectra of samples at various ages
show bands corresponding to the asymmetric stretching of CO23 −
around 1380–1470 cm−1 as an evidence of carbonation (also observed
from XRD as shown in Fig. 5). The scissoring mode of water at
1638 cm−1 is evident in all of the IR spectra [40].
As mentioned above, the position of the main bands in both the HS
and LS samples is comparable at 7 days. This does not explain the differ-
ences between the two types of samples as seen from the calorimetric
curve and the compressive strength results. Furthermore, it is evident
from the SEM images presented earlier and elsewhere that the reaction
of fly ash and slag is not complete after 7 days [6]. The main band of the
sample at all ages is broad and encompasses the response of the
unreacted fly ash, slag, aluminosilicate gel and C–S–H. Differentiating
among them is not straightforward.

3.3. IR spectra of insoluble residue after selective dissolution treatment

Selective dissolution was performed on the HS samples at three

different ages (1 h, 24 h, 7 days) to monitor the phase development
and was compared with the LS sample at 7 days. The position of the IR
bands in the HS samples at various ages, along with changes in the
positions after each treatment, is presented in Table 2.
Fig. 9 shows the spectrum of the original sample along with the
spectra of insoluble residues after SAM and HCl extractions at 1 h,
24 h, and 7 days. The insoluble residues after SAM and HCl treatments
will be referred hereafter as ‘SAM residue’ and ‘HCl residue’, respective-
ly. The insoluble residue after SAM extraction was also subjected to HCl
treatment and will be denoted as ‘HCl residue after SAM’.
Fig. 9a shows that HS at 1 h has a main band with a peak centered at
999 cm−1. After SAM extraction, the band moves to a higher frequency
1024 cm− 1 and a shoulder at 910 cm−1. The shoulder at 910 cm− 1
indicates the presence of Si–OH groups [41]. Similar observations are
made at 24 h and 7 days. To determine the cause of the shift in IR spec-
tra, the solubility of various phases in SAM was considered. SAM treat-
ment is known to dissolve calcium hydroxide, calcium oxides, and
calcium silicates, but it does not affect calcium carbonate [21]. From
Table 1, it is clear that the amount of fly ash and slag that dissolves in
SAM is less than 2%, and the spectral shift caused by the dissolution of Fig. 9. IR spectrum of HS sample before and after different treatments, SAM, HCl, HCl on
fly ash and slag in the SAM treatment is negligible. It is known that SAM at various ages a: 1 h, b: 24 h, and c: 7 day.
 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
silicate hydrate by removing calcium from the structure, leaving silica
gel behind [21]. Due to the dissolution of several products, it is conclud-
ed that the spectrum of insoluble residue after HCl treatment does not
present any significant information.
The insoluble residue after the SAM treatment was subjected to HCl
treatment to further simplify the issue of overlapping IR spectra and to
avoid the response from the silica gel in the HCl residue. This spectrum
is presented in Fig. 9a as ‘HCl residue after SAM’ treatment for 1 h sam-
ple. The position of the main band is now at 1054 cm−1 which is differ-
ent from the position of the main band after HCl treatment indicating
differences in the nature of dissolving products. To gain further informa-
tion on the products that have dissolved during various treatments,
spectral subtraction was performed and is presented in the later
sections.
Fig. 10 presents the weight % of the sample that was dissolved during
SAM treatment (C–A–S–H), HCl treatment, HCl after SAM treatment and
the amount (wt.%) of insoluble residue after HCl treatment (unreacted
fly ash + silica gel) of HS sample (at 1 h, 24 h, 7 days) and LS sample
(at 7 days). The amount of sample dissolved in SAM increases with
time, indicating an increase in the amount of C–A–S–H. At 7 days, the
HS sample showed a higher amount of weight loss during the SAM
treatment than the LS sample, indicating a greater amount of C–A–S–
H formed in the HS sample. This is a reasonably expected result as the
HS sample had a higher amount of calcium available from the higher
slag replacement. The weight loss due to the HCl treatment (dissolved
in HCl) increases with time; however, it cannot be directly correlated
with the amount of geopolymer formed because of the other phases
dissolving in HCl simultaneously. To eliminate the issue of partially
dissolving C–A–S–H during HCl treatment, the weight loss due to HCl
treatment after SAM was considered. Weight loss due to HCl treatment
after SAM decreased with time. This observation is counterintuitive at a
first glance since the amount of geopolymer should not decrease with
time; however, it is not impossible, as the amount of unreacted slag
(that also dissolves in HCl) decreases with time. The weight of the insol-
uble residue after HCl treatment provides a general indication of fly ash
reactivity as the insoluble residue is primarily unreacted fly ash and
silica gel (left behind from C–A–S–H dissolution). It is clear that
the amount of unreacted fly ash decreased with time. Comparing the
weight of insoluble residue after HCl treatment at 7 days between the
HS and the LS samples, it is clear that the slag addition increases fly
ash reactivity. This finding matches with similar conclusion made by
Fig. 10. Analysis of selective dissolution treatments presented as dissolved % by the initial mass of HS sample (1 h, 24 h, and 7 days) and LS sample.
45
other researchers and corroborates the microstructural and calorimetric
study presented earlier [6,43].
3.4. Spectral subtraction to determine IR spectrum of dissolved products in
selective dissolution treatments
Spectral subtraction was performed to understand the products
dissolving with each treatment. Subtraction was performed on spectra
collected at various ages: Fig. 11 (1 h), Fig. 12 (24 h), and Fig. 13 (7 days).
The spectra of components dissolved during the treatment were
obtained by subtracting the IR spectra obtained after attack from the
original spectra, using OMNIC software. The spectra collected before
treatment is noted as ‘original’, after a specific treatment is noted as
‘(treatment) residue’. The resultant spectrum obtained through subtrac-
tion is referred to as ‘dissolved in (treatment)’. Table 2 summarizes the
band positions in the subtraction spectra at various ages for the HS
sample.
3.4.1. Spectral subtraction after selective dissolution on 1 h sample
Fig. 11a presents the following IR spectra for a 1 h old sample: ‘orig-
inal’, ‘SAM residue’ and the subtraction spectrum obtained for phases
dissolved in SAM (dissolved in SAM spectrum). The spectrum after sub-
traction reveals a broad band at 978 cm−1 and a shoulder at 860 cm−1.
The peak at 978 cm−1 may be attributed to Si–O stretching vibrations
associated with Q2 tetrahedra of the C–S–H gel, whose position depends
on Ca/Si ratio [44]. EDS analysis showed the presence of C–A–S–H and
K–A–S–H in a 1 h old sample and a combination of C–A–S–H, K–A–S–
H, Ca(K)–A–S–H at later ages [3–6,8,9]. The spectroscopy and spectral
subtraction presented here also predicts the presence of C–S–H in the
early ages.
The 1107 cm−1 band is typical of Si–O asymmetric stretching vibra-
tions that are associated with Q3 sites [11,44,45]. The presence of this
peak in the subtracted spectrum indicates the dissolution of a product,
which is rich in silica, and may have taken up some amount of calcium
and alkali [11]. As the intensity of this band is smaller than at
978 cm−1, the amount of such silica rich gel with calcium was predicted
to be low. This peak is evident in all of the samples at various ages. In the
few extant studies on acidic zeolites, bands that present at approximate-
ly 1100 cm−1 were assigned to in-plane (plane was defined as the trian-
gle Si–O–Al) hydroxyl bending modes [46–49]. The subtraction
spectrum also shows carbonate peaks at 1395 cm−1, 1467 cm−1 and
46 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
Fig. 11. Subtraction spectrum a: HS 1 h after SAM extraction, b: HS 1 h after HCl extraction,
c: HS 1 h after HCl extraction on the IR after SAM treatment.
866 cm− 1, which are absent in the SAM residue spectrum. This may
indicate the presence of a soluble or low stability carbonate species in
geopolymer that are removed during SAM extraction.
Fig. 11b presents the IR subtraction spectrum after HCl treatment.
Close observation reveals that the ‘dissolved in HCl’ spectrum in
Fig. 11b and the ‘dissolved in SAM’ spectrum in Fig. 11a show similar
peaks despite differences in the main band positions of their respective
insoluble residues. Both treatments are not expected to dissolve same
phases as HCl should dissolve geopolymer but SAM should not. The
spectrum for the dissolved phases in HCl after SAM treatment is given
in Fig. 11c and is quite different from the other two. This indicates that
indeed SAM and HCl do not dissolve the same phases.
The spectrum for dissolved phases in HCl after SAM has a broad
band, ranging between 868 and 1011 cm− 1, which resembles slag
(whose main band is in the range of 870–940 cm−1 as seen in Fig. 2).
Fig. 12. Subtraction spectrum a: HS 24 h after SAM extraction, b: HS 24 h after HCl
extraction, c: HS 24 h after HCL treatment on IR after SAM extraction.
However, the subtraction spectrum is broader in the higher wave
number region compared to the slag spectrum. The main band position
of K–A–S–H is around 1200–950 cm− 1 depending on the aluminum
substitution [10,13,14,16,17,27,39]. Therefore, the broadness of the
spectrum is attributed to the dissolution of a small amount of K–A–S–
H geopolymer and the complete dissolution of unreacted slag.
This result implies that the spectral subtraction after selective
dissolution is sensitive enough to identify the dominant phase in the
microstructure at an early age (C–S–H/C–A–S–H) and the small amount
of K–A–S–H present as reported in a study through EDS observations
[6]. Moreover, it confirms that the initial peak in calorimetry also corre-
sponds to the initial precipitation of C–A–S–H and K–A–S–H besides the
wetting of raw materials. Furthermore, it matches with the hypothesis
of researchers intending to explain the accelerated hardening of
metakaolin/fly ash geopolymers in the presence of soluble calcium [3,
 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
Fig. 13. Subtraction spectrum a: HS 7 day after SAM extraction, b: HS 7 day after HCl
extraction, c: HS 7 day after HCl treatment on IR after SAM extraction.
8,35]. The broad band observed due to a small amount of K–A–S–H
geopolymer at 1 h got sharper with time and presented in the later
sections.
3.4.2. Spectral subtraction after selective dissolution on 24 h sample
Fig. 12 shows the spectral subtraction results after similar treat-
ments on a 24 h sample. The peak positions of the dissolved phases at
24 h were similar to the 1 h sample. The following discussion highlights
the differences observed. Peaks related to Si–O stretching vibrations and
associated with Q2 sites of C–S–H gel shifted to a lower wavenumber
compared to the 1 h old sample. This implies either an increase in the
aluminum substitution in the C–S–H structure or an increase in Ca/Si
ratio with time [44]. The calcium silicate hydrate forming at an early
47
age (1 h) may have less aluminum due to the higher dissolution rate
of calcium from slag than aluminum [50]. C–S–H is mostly formed
from calcium that is dissolved from slag and silicon already present in
the activator solution. As raw material dissolution continues to provide
more aluminum, the calcium silicate hydrate with high aluminum
content may be forming, which would explain the shift in the position
of the main band. The aluminum substitution is also enhanced with a
decrease in the pH of the system [21,51–54]. Also, the shift towards
lower frequency could be attributed to the depolymerization of the sil-
icate chains with an increase in the Ca/Si ratio over time. A low Ca/Si
ratio in C–S–H is often observed in the early ages of slag activation
due to the presence of small amount of Ca from the dissolution of slag
and significant amounts of Si present in the activator solution [54–57].
With an increase in the dissolution of slag, the Ca/Si ratio of C–S–H
could increase which might also account for the shift.
As with the 1 h sample, the 24 h sample has the IR spectra of phases
dissolved in SAM similar to the IR spectra of phases dissolved in HCl
except for the relative intensities of peaks. The presence of K–A–S–H
was identified from the subtraction spectrum of phases dissolved in
HCl after SAM. This spectrum has a peak at 1000 cm−1, which resembles
that of geopolymer, a shoulder at 1100 cm−1, and broad band of 860–
910 cm−1 (unreacted slag).
It is clear that the method is sensitive enough to show an increase in
the K–A–S–H content, a decrease in the unreacted slag content, and an
increase in Al content and/or Ca content in C–A–S–H with time.
3.4.3. Spectral subtraction after selective dissolution on 7 day sample
The main differences observed between the 7 days sample and early
age samples are discussed here. The peak positions, corresponding to
Si–O stretching vibrations associated with Q2 sites of C–A–S–H gel,
shifted towards a lower wavenumber compared to the 1 h old sample
and towards a higher wavenumber compared to the 24 h old sample.
This implies an increased aluminum substitution in C–A–S–H with
time. As with the case of 1 h and 24 h samples, the IR spectra of phases
dissolved in SAM and dissolved in HCl are similar except for the relative
intensities of the peaks. The IR spectra of phases dissolved in HCl after
SAM treatment is different at 7 days compared to earlier ages. The
sharper peaks in the 7 days sample (compared to early age samples) in-
dicate a greater amount of product, which is also evident from the %
weight dissolved. The peak corresponding to the main geopolymer
product shifted to a lower frequency compared to early age samples.
This indicates continuing geopolymerization and an increase in the
Al/Si ratio of the geopolymer with time [17,58]. Unlike the 1 h or 24 h
samples, a shoulder appears at 1020 cm−1, indicating Q4 (nAl) units
in a geopolymer with an Al content that is lower than the one repre-
sented by a main band at 992 cm−1.
3.4.4. Evolution of reaction products in LS sample
To check if the selective dissolution treatment followed by the FTIR
spectral subtraction may identify the differences between two
geopolymer systems with different slag contents, experiments were
performed on an LS sample at 7 days. It was observed that the spectrum
of the LS and HS samples at 7 days has similar peak positions (Fig. 8),
which does not explain the differences in the compressive strength.
Therefore, spectral subtraction was performed after SAM and HCl on
SAM treatments. Direct HCl treatment was omitted because it proved
to be less informative.
Along with subtraction spectrum, Fig. 14 presents the FTIR spectra
of the sample before and after selective dissolution techniques. The
positions of the bands in the spectra are given in Table 2. Fig. 14a
shows the spectrum of phases dissolved in SAM with a main band at
960 cm−1 (Si–O stretching vibrations associated with Q2 tetrahedra of
C–S–H gel), and shoulder at 1110 cm−1, which is similar to the HS sam-
ple at 7 days. Fig. 14b shows the spectrum of phases dissolved in HCl
after SAM which has a broad peak (1066–1040 cm−1) and another
broad band corresponding to slag dissolved in HCl. The broad peak
48 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
Fig. 14. Subtraction spectrum of LS mix a: 7 day after SAM extraction, b: 7 day after HCl
treatment on IR after SAM extraction.
corresponds to a geopolymer of lower Al content compared to the
geopolymer in the HS sample at 7 days. The relative peak intensities
compared with the shoulder (1110 cm−1) indicates that the LS sample
has a lower amount of geopolymer compared to the HS sample. Also,
the amount of dissolved component in the SAM treatment is greater
in the case of HS compared to that of LS. The lesser extent of fly ash dis-
solution, the lower amount of reaction product, and the low Al content
of geopolymer explain the low strength observed in the LS sample
compared to the HS sample. Selective dissolution treatments with
FTIR spectral subtraction successfully explain the difference between
two different geopolymer mixes.
4. Conclusions
Often it is difficult to determine the nature of reaction products
when calcium containing aluminosilicate precursors is activated using
an alkali and/or alkali silicate combination. It was found that IR spectra
collected on insoluble residue of salicylic acid-methanol (SAM) and HCl
treatment performed separately remains inconclusive. In this paper, the
authors showed that the combined results from SAM treatment and HCl
treatment on SAM residue, followed by spectral subtraction, render a
greater amount of information about calcium containing complex
systems. FTIR characterization and spectral subtraction performed
after SAM treatment and after successive chemical treatments using
SAM and HCl demonstrated the co-existence of C–S–H and K–A–S–H
in two different alkali-activated fly ash-slag blends. The technique was
shown to be sensitive to the presence of small amounts of reaction
products. The method successfully predicted the evolution of reaction
products (C–A–S–H and K–A–S–H) with time. The greater amount of
product in the later age samples as seen from SEM micrographs was
also confirmed from selective dissolution and FTIR spectral subtraction.
SEM micrographs and calorimetry further validate selective dissolution
followed by a spectral subtraction method in determining the nature of
products in a complex mix composition such as fly ash-slag.
Acknowledgments
X-ray diffraction analysis and scanning electro microscopy analysis
were carried out in part in the Frederick Seitz Materials Research
Laboratory Central Facilities, University of Illinois. ATR-FTIR was carried
out in Dr. Timothy Strathmann's lab in the Department of Civil
and Environmental Engineering Department, University of Illinois at
Urbana-Champaign.
References
[1] J. Davidovits, Geopolymer Chemistry and Applications, Institut Geopolymere, Saint-
Quentin, France, 2008.
[2] I. García-Lodeiro, A. Palomo, A. Fernández-Jiménez, D.E. Macphee, Compatibility
studies between N–A–S–H and C–A–S–H gels. Study in the ternary diagram
Na2O–CaO–Al2O3–SiO2–H2O, Cem. Concr. Res. 41 (2) (2011) 923–931.
[3] C.K. Yip, G.C. Lukey, J.L. Provis, J.S.J. van Deventer, Effect of calcium silicate sources on
geopolymerisation, Cem. Concr. Res. 38 (4) (2008) 554–564.
[4] C.K. Yip, J.S.J. Van Deventer, Microanalysis of calcium silicate hydrate gel formed
within a geopolymeric binder, J. Mater. Sci. 38 (18) (2003) 3851–3860.
[5] C.K. Yip, G.C. Lukey, J.S.J. van Deventer, The coexistence of geopolymeric gel and cal-
cium silicate hydrate at the early stage of alkaline activation, Cem. Concr. Res. 35 (9)
(2005) 1688–1697.
[6] S. Puligilla, P. Mondal, Role of slag in microstructural development and hardening of
fly ash-slag geopolymer, Cem. Concr. Res. 43 (2013) 70–80.
[7] S. Puligilla, P. Mondal, Microstructural changes responsible for hardening of fly ash–
slag geopolymers studied through infrared spectroscopy, in: L. Struble, J.K. Hicks
(Eds.), Geopolymer Binder Systems, STP 1566, ASTM International, West
Conshohocken, PA, 2013, pp. 21–33. http://dx.doi.org/10.1520/ STP156620120084.
[8] R.R. Lloyd, J.L. Provis, J.S.J. van Deventer, Microscopy and microanalysis of inorganic
polymer cements. 2: the gel binder, J. Mater. Sci. 44 (2) (2009) 620–631.
[9] J.E. Oh, J. Moon, S.-G. Oh, S.M. Clark, P.J.M. Monteiro, Microstructural and composi-
tional change of NaOH-activated high calcium fly ash by incorporating Na-
aluminate and co-existence of geopolymeric gel and C–S–H (I), Cem. Concr. Res.
42 (5) (2012) 673–685.
[10] A. Fernández-Jiménez, A. Palomo, Mid-infrared spectroscopic studies of alkali-
activated fly ash structure, Microporous Mesoporous Mater. 86 (1) (2005) 207–214.
[11] I. Garca-Lodeiro, A. Fernández-Jiménez, M.T. Blanco, A. Palomo, FTIR study of the
sol–gel synthesis of cementitious gels: C–S–H and N–A–S–H, J. Sol-Gel Sci. Technol.
45 (1) (2008) 63–72.
[12] 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 (1) (2002) 225–231.
[13] W.K.W. Lee, J.S.J. van Deventer, Structural reorganisation of class F fly ash in alkaline
silicate solutions, Colloids Surf. A Physicochem. Eng. Asp. 211 (1) (2002) 49–66.
[14] W.K.W. Lee, J.S.J. van Deventer, Use of infrared spectroscopy to study
geopolymerization of heterogeneous amorphous aluminosilicates, Langmuir 19
(21) (2003) 8726–8734.
[15] M. Criado, A. Fernández-Jiménez, A. Palomo, Alkali activation of fly ashes. Effect of
SiO2/Na2O ratio part I: FTIR study, Microporous Mesoporous Mater. 106 (2007)
180–191.
[16] 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 (17) (2007)
9076–9082.
[17] C.A. Rees, J.L. Provis, G.C. Lukey, J.S.J. van Deventer, Attenuated total reflectance
Fourier transform infrared analysis of fly ash geopolymer gel aging, Langmuir 23
(15) (2007) 8170–8179.
[18] A. Fernández-Jimenez, A.G. De La Torre, A. Palomo, G. López-Olmo, M.M. Alonso,
M.A.G. Aranda, Quantitative determination of phases in the alkali activation of fly
ash. Part I. Potential ash reactivity, Fuel 85 (5) (2006) 625–634.
[19] F. Puertas, S. Martínez-Ramírez, S. Alonso, T. Vazquez, Alkali-activated fly ash/slag
cements: strength behaviour and hydration products, Cem. Concr. Res. 30 (10)
(2000) 1625–1632.
[20] P.E. Stutzman, Building, F. R. L. (US), Guide for X-ray Powder Diffraction Analysis of
Portland Cement and Clinker, US Department of Commerce, Technology Adminis-
tration, National Institute of Standards and Technology, Office of Applied Economics,
Building and Fire Research Laboratory, 1996.
[21] H.F.W. Taylor, Cement Chemistry, Thomas Telford, 1997.
[22] L.J. Struble, The effect of water on maleic acid and salicylic acid extractions, Cem.
Concr. Res. 15 (4) (1985) 631–636.
[23] A. Palomo, A. Fernández-Jiménez, G. Kovalchuk, L.M. Ordoñez, M.C. Naranjo, OPC-fly
ash cementitious systems: study of gel binders produced during alkaline hydration,
J. Mater. Sci. 42 (9) (2007) 2958–2966.
[24] K. Luke, F.P. Glasser, Selective dissolution of hydrated blast furnace slag cements,
Cem. Concr. Res. 17 (2) (1987) 273–282.
[25] V. Kocaba, E. Gallucci, K.L. Scrivener, Methods for determination of degree of reac-
tion of slag in blended cement pastes, Cem. Concr. Res. 42 (3) (2012) 511–525.
 S. Puligilla, P. Mondal / Cement and Concrete Research 70 (2015) 39–49
[26] H.M. Dyson, I.G. Richardson, A.R. Brough, A combined 29Si MAS NMR and selective
dissolution technique for the quantitative evaluation of hydrated blast furnace
slag cement blends, J. Am. Ceram. Soc. 90 (2) (2007) 598–602.
[27] A. Fernández-Jiménez, A.G. De La Torre, A. Palomo, G. López-Olmo, M.M. Alonso,
M.A.G. Aranda, Quantitative determination of phases in the alkaline activation of
fly ash. Part II: degree of reaction, Fuel 85 (14) (2006) 1960–1969.
[28] K. Somna, W. Bumrongjaroen, Effect of external and internal calcium in fly ash on
geopolymer formation, Developments in Strategic Materials and Computational
Design II: Ceramic Engineering and Science Proceedings, 322011. 1–16.
[29] G. Le Saoût, M. Ben Haha, F. Winnefeld, B. Lothenbach, Hydration degree of alkali-
activated slags: a 29Si NMR study, J. Am. Ceram. Soc. 94 (12) (2011) 4541–4547.
[30] R. Snellings, Solution-controlled dissolution of supplementary cementitious materi-
al glasses at pH 13: the effect of solution composition on glass dissolution rates, J.
Am. Ceram. Soc. 96 (8) (2013) 2467–2475.
[31] Z. Sarbak, M. Kramer-Wachowiak, Porous structure of waste fly ashes and their
chemical modifications, Powder Technol. 123 (1) (2002) 53–58.
[32] M.B. Haha, K. De Weerdt, B. Lothenbach, Quantification of the degree of reaction of
fly ash, Cem. Concr. Res. 40 (11) (2010) 1620–1629.
[33] B.A. Suprenant, G. Papadopoulos, Selective dissolution of Portland-fly-ash cements,
J. Mater. Civ. Eng. 3 (1) (1991) 48–59.
[34] C.A. Rees, J.L. Provis, G.C. Lukey, J.S. van Deventer, The mechanism of geopolymer gel
formation investigated through seeded nucleation, Colloids Surf. A Physicochem.
Eng. Asp. 318 (1) (2008) 97–105.
[35] J. Temuujin, A. Van Riessen, R. Williams, Influence of calcium compounds on the
mechanical properties of fly ash geopolymer pastes, J. Hazard. Mater. 167 (1–3)
(2009) 82–88.
[36] 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 (3)
(2010) 607–615.
[37] S.A. Bernal, J.L. Provis, V. Rose, R. Meja de Gutierrez, Evolution of binder structure in
sodium silicate-activated slag-metakaolin blends, Cem. Concr. Compos. 33 (1)
(2011) 46–54.
[38] S. Alonso, A. Palomo, Calorimetric study of alkaline activation of calcium hydroxide–
metakaolin solid mixtures, Cem. Concr. Res. 31 (1) (2001) 25–30.
[39] M. Handke, W. Mozgawa, Vibrational spectroscopy of the amorphous silicates, Vib.
Spectrosc. 5 (1) (1993) 75–84.
[40] N.Y. Mostafa, A.A. Shaltout, H. Omar, S.A. Abo-El-Enein, Hydrothermal synthesis and
characterization of aluminium and sulfate substituted 1.1 nm tobermorites, J. Alloys
Compd. 467 (1) (2009) 332–337.
[41] C.H. Rüscher, E.M. Mielcarek, J. Wongpa, C. Jaturapitakkul, F. Jirasit, L. Lohaus,
Silicate-, aluminosilicate and calciumsilicate gels for building materials: chemical
and mechanical properties during ageing, Eur. J. Mineral. 23 (1) (2011) 111–124.
49
[42] A. Mendes, W.P. Gates, J.G. Sanjayan, F. Collins, NMR, XRD, IR and synchrotron
NEXAFS spectroscopic studies of OPC and OPC/slag cement paste hydrates, Mater.
Struct. 44 (10) (2011) 1773–1791.
[43] P.J. Williams, J.J. Biernacki, L.R. Walker, H.M. Meyer, C.J. Rawn, J. Bai, Microanalysis of
alkali-activated fly ash–CH pastes, Cem. Concr. Res. 32 (6) (2002) 963–972.
[44] P. Yu, R.J. Kirkpatrick, B. Poe, P.F. McMillan, X. Cong, Structure of calcium silicate
hydrate (C–S–H): near-, mid-, and far-infrared spectroscopy, J. Am. Ceram. Soc. 82
(3) (1999) 742–748.
[45] I. Garca Lodeiro, D.E. Macphee, A. Palomo, A. Fernández-Jiménez, Effect of alkalis on
fresh C–S–H gels. FTIR analysis, Cem. Concr. Res. 39 (3) (2009) 147–153.
[46] W.P.J.H. Jacobs, J.H.M.C. Van Wolput, R.A. Van Santen, H. Jobic, A vibrational study of
the OH and OD bending modes of the Brönsted acid sites in zeolites, Zeolites 14 (2)
(1994) 117–125.
[47] P.J.H. Wim, J.H.M.C. ávan Wolput, R.A. ávan Santen, Fourier-transform infrared study
of the protonation of the zeolitic lattice. Influence of silicon: aluminium ratio and
structure, J. Chem. Soc. Faraday Trans. 89 (8) (1993) 1271–1276.
[48] W.P.J.H. Jacobs, H. Jobic, J.H.M.C. Van Wolput, R.A. Van Santen, Fourier transform
infrared and inelastic neutron scattering study of HY zeolites, Zeolites 12 (3)
(1992) 315–319.
[49] R.A. Van Santen, G.J. Kramer, Reactivity theory of zeolitic Brønsted acidic sites,
Chem. Rev. 95 (3) (1995) 637–660.
[50] C. Shi, P.V. Krivenko, D.M. Roy, Alkali-activated Cements and Concrete, CRC PressI
Llc, 2006.
[51] I. Garca Lodeiro, A. Fernández-Jimenez, A. Palomo, D.E. Macphee, Effect on fresh CSH
gels of the simultaneous addition of alkali and aluminium, Cem. Concr. Res. 40 (1)
(2010) 27–32.
[52] W. Mozgawa, M. Sitarz, M. Rokita, Spectroscopic studies of different aluminosilicate
structures, J. Mol. Struct. 511 (1999) 251–257.
[53] I.G. Richardson, The nature of CSH in hardened cements, Cem. Concr. Res. 29 (8)
(1999) 1131–1147.
[54] S.-D. Wang, K.L. Scrivener, 29Si and 27Al NMR study of alkali-activated slag, Cem.
Concr. Res. 33 (5) (2003) 769–774.
[55] B. Lothenbach, A. Gruskovnjak, Hydration of alkali-activated slag: thermodynamic
modelling, Adv. Cem. Res. 19 (2) (2007) 81.
[56] A. Gruskovnjak, B. Lothenbach, L. Holzer, R. Figi, F. Winnefeld, Hydration of alkali-
activated slag: comparison with ordinary Portland cement, Adv. Cem. Res. 18 (3)
(2006) 119–128.
[57] F. Puertas, A. Fernández-Jiménez, M.T. Blanco-Varela, Pore solution in alkali-
activated slag cement pastes. Relation to the composition and structure of calcium
silicate hydrate, Cem. Concr. Res. 34 (1) (2004) 139–148.
[58] B.N. Roy, Spectroscopic analysis of the structure of silicate glasses along the joint
xMAlO2-(1 − x) SiO2 (M = Li, Na, K, Rb, Cs), J. Am. Ceram. Soc. 70 (3) (1987) 183–192.