Study of carbonation rate of synthetic

C-S-H by XRD, NMR and FTIR

Bei Wu 1, Guang Ye 1, 2
1 Microlab, Delft University of Technology, the Netherlands
2 Magnel Laboratory for Concrete Research, Ghent University, Belgium

Supplementary cementitious materials (SCMs) like fly ash (FA) and blast furnace slag (BFS)
are broadly used in concrete to replace part of the Ordinary Portland Cement (OPC) because
of both economic and environmental issues. In concrete blended with SCMs, C-S-H with
different C/S ratios, formed from the hydration and pozzolanic reactions of blended cement,
is the major calcium-bearing phases which reacts with CO2 during carbonation. Therefore, it
is important to study the carbonation rate of different C-S-H phases. In this paper, the C-S-H
phases (C/S ratio: 0.66 to 2.0) were synthesized and used for accelerated carbonation testing.
Synthetic C-S-H phases with different C/S ratios were identified by X-ray diffraction and 29Si
nuclear magnetic resonance (NMR). Carbonation rate and products of different C-S-H phases
are also determined. The results show that C-S-H (I) phases with different target C/S ratio
(lower than 1.40) were synthesized in the mix solution of lime and fume silica. The
portlandite appears in the products when the designed C/S ratio is higher than 1.40 under this
synthetic condition. C-S-H with lower C/S ratio is decomposed faster than that with a higher
C/S ratio. After exposition to the accelerated carbonation condition for three days, in this
research, the C-S-H phases with different C/S ratio were all fully decomposed to CaCO3 and
silica gel.

Keywords: C-S-H, C/S ratio, structure, carbonation rate

1 Introduction

Carbonation of portlandite (CH) and calcium silicate hydrates (C-S-H) are the main
carbonation reactions happened inside the concrete. The consequence is the reduction of
alkalinity, which increases the corrosion risk of the reinforcement. Carbonation of CH
leads to a reduction of the porosity ascribed to the positive difference of molar volume
between CH and CaCO3 (Ngala and Page 1997, Delmi, Aït-Mokhtar et al. 2006), which will
slow down the carbonation development in the Ordinary Portland Cement (OPC) concrete.
However, the situation will be potentially different in the blended cement concrete.

HERON Vol. 64 (2019) No. 1/2 21

Supplementary cementitious materials (SCMs) like fly ash (FA) and blast furnace slag
(BFS) are broadly used in the concrete to replace parts of OPC because of both economic
and environmental issues. Introducing SCMs in the concrete will usually decrease the CH
amount in blended cement concrete, caused by the consumption in pozzolanic reactions
and dilution effects. Therefore, the major carbonation reaction in blended cement concrete
will happen between the C-S-H and CO2.

Carbonation of C-S-H has been studied by many authors (Šauman 1972, Black, Breen et al.
2007, Morandeau, Thiery et al. 2014). It is a complex decalcification-polymerization process
of the C-S-H and the formation of amorphous silica gel, see Eq. (1).

C x S H y + xC → xC C + S H t + ( y − t ) H (1)

In which, the cement notations are explained as follows: C = CaO, H = H2O, S = SiO2 ,

A = A2O3 , C = CO2 ; x, y and t are the molecular numbers. From the above-mentioned
reaction equation, the molar volume change during the carbonation of C-S-H is depended
on the properties of C-S-H (like C/S ratio (C/S), water content) and the water remained in
silica gel. Unlike the carbonation reaction between CH and CO2, the volume changes
during the carbonation of C-S-H varies among C-S-H phases with different C/S. Therefore,
the effect of the C-S-H carbonation on the porosity development is still controversial.

C-S-H phase has a layer structure can be inferred from the refined structure of 11Å
tobermorite (Ca4.5Si6O16(OH)· 5H2O) (Merlino, Bonaccorsi et al. 2001). The layer structure
of 11Å tobermorite consists of three parts: CaO2 sheet, ‘dreierkette’ form SiO chain and
interlayer. In the CaO2 sheet layer, the 7-fold coordinated Ca2+ share all the oxygen atoms
with Si4+ in SiO chains of SiO4 tetrahedra on both side of the CaO2 layer. In the SiO chain
layer, SiO4 tetrahedra coordinate themselves to Ca ions by linking in a dreierketten
arrangement, to repeat a kinked pattern after three tetrahedra (Megaw and Kelsey 1956).
Two of the tetrahedra are linked to CaO polyhedra by sharing O-O edges with the central
Ca-O part of the layer, called paired tetrahedra (PT); the third tetrahedra shares an oxygen
atom at the pyramidal apex of a Ca polyhedron, called bridging tetrahedra (BT). There are
Ca2+ cations and water molecules in the interlayer.

The C/S of C-S-H in the concrete is normally higher than 0.83. The increase of the C/S in
C-S-H based on the tobermorite structure can be caused by: omission of the bridging

22
tetrahedra and incorporation of additional calcium in the interlayer. If all the bridge
tetrahedra are removed, the C/S can increase to 1.25 (Taylor and Howison 1956,
Richardson 2004). Further incorporation of extra Ca2+ in the interlayer can form the C-S-H
with a much higher C/S. The theoretical value is 1.50 when all the bridging sites are
removed and taken up by extra Ca2+ (Richardson 2004). The extra Ca2+ is normally
balanced by the omission of H+ or the incorporation of OH- or both. If the amount of extra
Ca continues increasing and achieves a C/S even higher than 1.5, the structure of C-S-H is
more close to the C-S-H/CH ‘solid solution’ (T/CH model) (Richardson 2004).

From the above discussion, three types of Ca exist in the C-S-H structure (classification by
the position of Ca): Ca in the CaO2 sheet layer, Ca in the bridging site and Ca in the
interlayer, shown in Figure 1. The proportions of them vary among C-S-H with different
C/S ratio. The removal of Ca from the three chemical sites may need different energy,
which means the decalcification rate are different among C-S-H phases with different C/S
ratio.

The aim of this work is to verify the above-mentioned assumptions, by studying the
carbonation rate and products of synthesized C-S-H phases with different C/S and further
to explain the phase transformation and microstructure development of blended cement
paste during the carbonation, discovered in the previous research (Wu and Ye 2015).

SiO chain

CaO sheet

SiO chain

Interlayer Ca 2+ and H2 O

Figure 1: Possible positions of Ca2+ in the C-S-H layer structure

23
2 Experiments and test methods

2.1 Raw Materials

Raw materials used in the synthesis are CaO and fumed silica. CaO is freshly prepared by
the calcination of CaCO3 under 1000 °C for at least 4 hours before synthesis. Fumed silica is
from Sigma-Aldrich, with the surface area of 175-225 m2/g.

2.2 Synthesis of C-S-H with different C/S

C-S-H gels were prepared by using stoichiometric amounts of CaO and fumed silica, to
give approximate C/S ranging between 0.66 and 2.0. The mix design of different C-S-H is
described in Table 1. Solid agents were mixed together with CO2-free water with the
water/solid ratio of 50 : 1. The solution was stirred by magnetic stirrer at 20 °C. The whole
synthesis procedure was under N2 protection to prevent carbonation. After 2 or 4 weeks of
reaction, samples of solid and liquid were extracted as a slurry. Solid products were
obtained by filtering the slurries through a Balston No. 45 paper. Then they were quickly
moved into the vacuum drying chamber and dried under 35 °C for 24 h. After drying, the
samples were stored in the desiccator with the relative humidity of 30%, regulated by the
standard saturated CaCl2· 6H2O solution. The set-up of the synthesis device is illustrated in
Figure 2.

Figure 2: Schematic diagram of the synthesis set-up

2.3 Accelerated carbonation of C-S-H

Well-dried C-S-H samples synthesized for 4 weeks were ground into powders and moved
into a carbonation chamber. The CO2 concentration is maintained at 3% ± 0.2 automatically
by the solenoid valve connected with a CO2 sensor. The temperature is regulated at 20 °C

24
and the relative humidity is controlled at around 75% by using the saturated NaCl
solution. The exposure time varies from 0.5 h to 7 days.

Table 1: Mass fraction of raw materials for preparing different C-S-H

Sample No. C/S ratio CaO (g) SiO2(g) H2O (g)
1 0.66 3.0 4.9 400
2 0.86 3.6 4.4 400
3 1.18 4.2 3.8 400
4 1.40 4.5 3.5 400
5 1.70 4.9 3.1 400
6 2.00 5.2 2.8 400

2.4 Test methods

Test methods used for the identification of different types of C-S-H and the carbonation
products were X-ray diffraction (XRD), 29Si nuclear magnetic resonance (NMR) and Fourier
transform infrared spectroscopy (FTIR). Before the above-mentioned test, well-dried
samples were ground into powders with an average particle size less than 75 μm.

XRD X-ray diffraction (XRD) was used for the identification of different types of C-S-H
and the carbonation products. XRD was performed by a Philips PW 1830 diffractometer
using CuKa radiation (λ = 0.154056 nm, 60 mA, 40 kV). Two types of scanning were
performed on different samples. The first one is the normal scanning with a scan step size
of 0.03°, 3 s per step, which was used for analyzing the synthetic products and carbonated
products. The second one is the slow scanning with a scan step size of 0.02°, 8 s per step,
which was performed on samples used for studying the carbonation rate. Both types of
scanning are covering the range from 5° to 70°.

Samples used for studying carbonation rate, were blended with a certain amount of the
internal standard corundum (Al2O3) (from 12 - 13%wt) and kept inside the sample holder
during the carbonation. After each carbonation stage, the sample was taken out with the
holder and tested by the XRD slow scanning.

FTIR The FTIR spectra were collected over the wavelength range of 4000 to 400 cm-1 by the
TM 100 Optical ATR-FTIR spectrometer. The resolution was 4 cm-1.

25
 NMR Solid state 29Si single pulse magic angle spinning (MAS) NMR spectra were acquired
using a BrukerMSL-400 spectrometer (magnetic field of 9.8 T; operating frequencies of 79.5
MHz). C-S-H powders were packed into the zirconia rotor sealed at either end with Teflon
end plugs and spun at 6 kHz in a Varian 7 mm wide-body probe. The spectra were
acquired using a pulse recycle delay of 5 s, a pulse width of 4.97 μs, and an acquisition
time of 20 ms; 2002 scans were collected for each sample. 29Si chemical shifts are given
relative to tetrakis(trimethylsilyl)silane (TTMSS) at -9.8 ppm, with kaolinite as an external
standard at - 91.2 ppm.

3 Results and discussion

3.1 Synthetic products identification

3.1.1 XRD
Normal scanning XRD test results of synthetic products are described in Figure 3. The
identical peaks with the d-spacing value of 3.04, 2.79, 1.82, 1.66 Å in the pattern indicate
that the C-S-H phase synthesized from CaO and fumed silica in the solution is similar to C-
S-H (I), which is one of the C-S-H phases found in the Portland cement concrete(Chen,
Thomas et al. 2004).

Due to the high reactivity of fumed silica, the C-S-H phases can be formed after 2 weeks’
reaction, see Figure 3a. When the designed C/S is less than 1.40, there is only C-S-H phase
existing in the product and no trace of unreacted raw materials or carbonation product.
However, the portlandite appears when the designed C/S is higher than 1.40 in this
research, see Figure 3b. This is consistent with other researchers’ results(Garbev, Beuchle et
al. 2008, Garbev, Bornefeld et al. 2008, Renaudin, Russias et al. 2009), in which the C-S-H
phases are synthesized by using the same method.

3.1.2 NMR
The NMR test results of synthetic products are described in Figure 4a and Figure 4b. The
peaks at around -78 ppm and -85 ppm indicates the Q1 and Q2 type of tetrahedra
respectively (Richardson and Groves 1997) (Cong and Kirkpatrick 1996). Apparently, the
intensity of the peak indicating Q1 increases with increasing of the C/S while the peak
intensity of Q2 decreases.

26
 d=3.04
C:C-S-H(I)
C: C S H(I)
P: portlandite d=2.79

d=3.04
25 26 27 28 29 30 31 32 33 34 35
C

d=2.79

d=1.82

d=1.66

d=1.40
C
C Ca/Si
C C
1.40
1.18

a 0.86
0.66
(a) 10 20 30 40 50 60 70
2- Theta
C: C-S-H(I), P: portlandite

d=3.04
P

d=2.79
P
d=3.04

P P 25 26 27 28 29 30 31 32 33 34 35
d=2.79

C
d=1.82

d=1.40
Ca/Si
d=1.66

C P P 2.00
C C PC 1.70
1.40

b
1.18
0.86
0.66

(b) 10 20 30 40 50 60 70
2- Theta

Figure 3: XRD test results of C-S-H with different C/S (0.66-2.0). C-S-H d-spacings are noted
above their respective peaks in the units of Ångstrom (Å). a-synthesis for 2 weeks; b-synthesis for 4
weeks

The decrease of peak intensity of Q2 means that the portion of tetrahedra located in the
middle of silicate chain decreases. It indicates the breaking of silicate chain and will be
followed by the formation of more Q1 tetrahedra, which can be observed in 29Si NMR
spectra as well. Therefore, the mean length of silicate chain will decrease with increasing
C/S. Considering the relationship between the mean silicate chain length (MCL) and C/S
of C-S-H, the MCL can be used to identify or distinguish C-S-H with different C/S.

The MCL was calculated from the NMR data using Eq. (2) (Richardson 1999) (Richardson
and Groves 1992).

27
 2(Q 1 +Q 2 )
MCL = (2)
Q1

In which, the Q1 and Q2 are the fractions of Si present in Q1 and Q2 tetrahedra respectively.
Quantitative information of the Si fraction was obtained by deconvolution of the single
pulse NMR spectra. The spectra were fitted to Gaussian/Lorentzian mixed function (with
the ratio of 1.0) using the dmfit2015 software package. An example of the fitting result is
illustrated in Figure 5.

a b
Ca/Si = 2.00

Ca/Si = 1.70
-78 ppm
Ca/Si = 1.40
-78 ppm
Ca/Si = 1.18

Ca/Si = 0.86
-85 ppm -85 ppm
Ca/Si = 0.66

(a) (b)
5
0
5
0
5
0
-1 5
-1 0
-1 5
10

5
0
5
0
5
0
-1 5
-1 0
-1 5
10
-6
-7
-7
-8
-8
-9
-9
0
0

-6
-7
-7
-8
-8
-9
-9
0
0

Chemical shift (ppm)

Figure 4: 29Si NMR test results of C-S-H with different C/S (0.66-2.0), a- synthesis for 2 weeks; b-
synthesis for 4 weeks

Experimental data
Fitting result
Peak-Q2
Peak-Q1
Residue

1 2
Q Q

-65 -70 -75 -80 -85 -90 -95 -100 -105 -110
Chemical shift (ppm)
Figure 5: Fitting results of 29Si NMR spectrum (C/S = 1.4, synthesis for 4 weeks)

28
The deconvolution process is performed on the 29Si NMR test results of synthetic C-S-H
with C/S ranging from 0.66 to 2.0. MCL is calculated and plot against the C/S, comparing
with the results of C-S-H phases synthesized by the same method (Macphee, Lachowski et
al. 1988, Cong and Kirkpatrick 1996, Chen, Thomas et al. 2004, Beaudoin, Raki et al. 2009,
He, Lu et al. 2013, L'Hôpital 2014), see Figure 6.

Synthesized for 2 Weeks

40
40
Synthesized for 4 Weeks
Chen et al, 2004
35
35
Cong et al, 1996 - SEWCS series
SCFUM series
30
30
CSHFS series
Macphee et al, 1988
25
25 Beaudoin et al, 2009
MCL He et al, 2013
MCL

20
20 Hopital, 2014
Rodriguez, 2015
15
15
10
10
55

0
0.4
0.4 0.6
0.6 0.8
0.8 1.0
1.0 1.2
1.2 1.4 1.6
1.4 1.6 1.8
1.8 2.0
2.0
Ca/Si
Ca / Si
Figure 6: Calculated mean silicate chain length (M CL) of C-S-H with different nominal C/S

The MCL of the C-S-H decreases with the increasing C/S when the C/S is less than 1.4.
Then the MCL stays at around 2~3 tetrahedron units long when the C/S increases from 1.4
to 2.0. This can verify the evolution model of C-S-H with higher C/S based on the
tobermorite (C/S = 0.83) (Taylor and Howison 1956, Richardson 2004). By omission the
bridge tedrahedra, the C/S will increase; meanwhile, the long silicate chain will be broken
into shorter polymeric species of SiO4 tetrahedra. Therefore, the MCL of C-S-H decreases
with the increase of the C/S. The C/S increases to 1.25 by the omission of all the bridge
tetrahedra in the silicate chain. The further incorporation of Ca in the interlayer can
achieve the C/S of 1.4, which will not affect the MCL. This is reason that the MCL of C-S-H
keeps unchanged when the C/S increases from 1.4 to 2.0.

As the MCL of C-S-H phases has a special relation with the C/S of the C-S-H, this feature
is used here to distinguish the C-S-H phases with different C/S. If comparing the data
points of this research with the reference data in Figure 6, the C/S of the synthetic C-S-H is

29
in accord with the initial designed value. Moreover, the XRD test results show only the C-
S-H (I) phase. Therefore, the C-S-H phases with different C/S (lower than 1.4) are
successfully synthesized in this research. The experiment set-up used in this study can
prevent the carbonation effectively.

3.2 Carbonation of C-S-H with different C/S

As per the XRD test results, the portlandite appears in the synthetic product and mixes
with C-S-H phases when the designed C/S is higher than 1.4. To avoid the effects from the
carbonation of portlandite, only the C-S-H with the C/S of 0.66 to 1.40 are used in the
accelerated carbonation study. The results are shown and discussed as follows.

3.2.1 NMR
Development of silicate chain in C-S-H with different C/S was studied by 29Si NMR test,
shown in Figure 7. These samples were carbonated up to 7 days. In the 29Si NMR spectra,
characteristic peaks at around -101 and -111 ppm indicate the Q3 and Q4 tetrahedra,
respectively. During carbonation, the fraction of Q3 and Q4 tetrahedra increases
dramatically while the fraction of Q1 and Q2 tetrahedra decreases at the same time. This
indicate the polymerization process of the silicate chain and the removal of Ca in the CaO2
sheet layer. The consequence is the linkage of the adjacent silicate chains and to form the
silicate network in two and three dimensions. After 7 days’ accelerated carbonation, only
peaks indicating Q3 and Q4 tetrahedra can be observed in the NMR spectra of C-S-H
studied. Meanwhile, peaks of Q1 and Q2 disappeared. This proves that the final
carbonation products of C-S-H is silica gel, not the mixture or ‘solid solution’ of CaCO3 and
C-S-H with lower C/S (Papadakis, Vayenas et al. 1989).

3.2.2 XRD
Slow scanning XRD tests were performed on the C-S-H with the C/S of 0.66, 0.86 and 1.40
to compare the carbonation rate difference. The exposure time varied from 0.5 h to 24 h.
The test results are described in Figure 10 to Figure 10.

In the pattern, the major peaks are the characteristic peaks of corundum and C-S-H. And
there are nearly no changes of these peaks in the pattern of C-S-H with the C/S of 1.40,
during the carbonation. However, in the pattern of C-S-H with the C/S of 0.66 or 0.86,
three peaks (2θ = 24.9, 27.0 and 32.8) appear after 1 hour’s carbonation, which are the
characteristic peaks of μ-CaCO3 (vaterite). Vaterite is one of the three polymorphs of

30
 -85 -111
Q2 Q4
-101
3
-78 Q Ca
1 =0.66
Q Si

-60 -70 -80 -90 -100 -110 -120 -130

3 days -85 Q2 -111
7 days
-85 Q4
Q2 -101
3
Q
Ca
=0.86
Si
1
-78 Q

-60 -70 -80 -90 -100 -110 -120 -130

4
7 days Q1 -85 Q
-111
-78
Q2 Q4
Q1
-101
3
Q
Ca
=1.18
Si

-60 -70 -80 -90 -100 -110 -120 -130

3 days
-111
7 days -78
-78 -85
1 Q
2 Q4
QQ1
-101
3
Q
Ca
=1.40
Si

Chemical shift [ppm]

-60 -70 -80 -90 -100 -110 -120 -130
Figure 7: NMR test results of different types of C-S-H, carbonated for different periods

31
 Ca/Si = 0.66; 4 weeks
Vaterite ♦: Corundum(Αl2O3)
♦ ♠: C-S-H
♦♠ ♦
♦
♠ ♦ ♠♦ ♦♦
♠
24 h
12 h
6.0 h
3.0 h Figure 8: XRD test results of
1.0 h
C-S-H with C/S of 0.66, carbonated
0.5 h
0h for different periods
5 10 15 20 25 30 35 40 45 50 55 60 65 70
2- Theta (°)

Ca/Si = 0.86; 4 weeks

Vaterite
♠ ♦: Corundum(Αl2O3)
♦ ♦ ♠: C-S-H
♦ ♦
♠ ♦ ♠♦
♠ ♦♦
24 h
12 h
6.0 h
3.0 h
Figure 9: XRD test results of
1.0 h
0.5 h C-S-H with C/S of 0.86, carbonated
0h for different periods
5 10 15 20 25 30 35 40 45 50 55 60 65 70
2- Theta (°)

♠ Ca/Si = 1.40; 4 weeks

♦: Corundum(Αl2O3)
♦ ♦ ♠: C-S-H
♦ ♠ ♦ ♦
♠♦ ♦♦
♠
24 h
12 h
6.0 h
3.0 h
Figure 10: XRD test results of
1.0 h
0.5 h C-S-H with C/S of 1.40, carbonated
0h for different periods
5 10 15 20 25 30 35 40 45 50 55 60 65 70
2- Theta (°)

32
calcium carbonate, the appearance of which indicates the carbonation of C-S-H. Therefore,
the carbonation is developed much faster in the C-S-H with lower C/S ratio of 0.66 or 0.86.

3.2.3 FTIR
The FTIR tests were also performed on the carbonated C-S-H samples tested by slow
scanning XRD. The test results are compared in Figure 11 to Figure 13 . In the FTIR spectra
of C-S-H, peaks at approximate 970 cm-1 and 1066 cm-1 indicate the Si-O stretching
vibrations of Q2 tetrahedra. The peak at around 875 and 1400~1500 cm-1 represents the
bending (v2) of CO32- and the stretching (v3) of CO32-, respectively. During carbonation, the
peak 1066 cm-1 will grow into an obvious and independent peak. The peak 970 cm-1 will
shift to the higher frequency and have a dramatic decrease in the intensity. These changes
reveal the progressive polymerization of silicate chains. (McMillan, Wolf et al. 1992, Sykes
and Kubicki 1993, Yu, Kirkpatrick et al. 1999). These changes reveal the progressive
polymerization of silicate chains. Meanwhile, both peak 875 cm-1 and 1400~1500 cm-1
have a dramatic increase in the intensity, indicating the destruction of silicate chain by the
carbonation.

Considering together the changes of above-mentioned peaks (~970 cm-1 and 1066 cm-1),
the ‘start time’ for the carbonation of C-S-H can be observed. This ‘critical’ time for the C-S-
H with the C/S of 0.66, 0.86 and 1.40 are 3h, 6 h and 12 h respectively. Obviously, the C-S-
H with a higher C/S is decomposed slower than the C-S-H with lower C/S under the same
condition and has a better resistance to the carbonation. This conclusion is in consistent
with the XRD test results.

3.2.4 Quantitative information of carbonation rate

Amount of calcium carbonate (vaterite) produced from the carbonation of C-S-H was
calculated using Rietveld refinement analysis of XRD spectra showing in Figure 8 and
Figure 9. The Rietveld refinement program used is BGMN, realized by the user interface
software Profex (Doebelin and Kleeberg 2015). Crystal structure information of vaterite
(CaCO3) and Corundum (Al2O3) are imported from the PDF card 04-011-5985 and 00-10-
0173. The weight percentage of vaterite was calculated and plotted against the exposure
time shown in Figure 14.

33
 1066
968
1.0
Ca/Si = 0.66

871
0.9 1400~1500
0.8
0.7 24 h
Absorbance (%)

0.6
18 h
0.5
12h
0.4
6.0 h
0.3
3.0 h
0.2
1.0 h Figure 11: FTIR test results of
0.1
0.5 h C-S-H with C/S of 0.66, carbonated
0.0
0h
for different periods
4000 3600 3200 2800 1600 1200 800
Wavenumbers (cm-1)

1066
871
968
1.0
0.9 Ca/Si = 0.86 1400~1500
0.8
0.7
24 h
Absorbance (%)

0.6
18 h
12h
0.5
6.0 h
0.4
3.0 h
0.3
1.0 h Figure 12: FTIR test results of
0.2
0.5 h
0.1 C-S-H with C/S of 0.86, carbonated
0h
0.0 for different periods
4000 3600 3200 2800 1600 1200 800
Wavenumbers (cm-1)
871
1066

1.0 1400~1500
968

Ca/Si = 1.40
0.9
0.8
0.7
24 h
Absorbance (%)

18 h
0.6
12h
0.5
6.0 h
0.4
3.0 h
0.3
1.0 h Figure 13: FTIR test results of
0.2
0.5 h C-S-H with C/S of 1.40, carbonated
0.1
0h
0.0 for different periods
4000 3600 3200 2800 1600 1200 800
Wavenumbers (cm-1)
34
 18
C/S=0.66 Slope = 0.7452, R2 = 0.9283

devided by the Initial Ca/Si of C-S-H (%)

16 C/S=0.86 Slope = 0.1181, R2 = 0.9923

Weight percentage of Vaterite Formed

14
12
10
8
6
4
2
0
0 5 10 15 20 25
Exposure Time (h)
Figure 14: Changes of the weight percentage of calcium carbonate with the exposure time for C-S-H
with C/S of 0.66 and 0.86

fG

L (μ m)
G

Time (min)

Figure 15: Expected sizes of individual particles in a batch growth cell vs time; the growth rate
distribution is shown in the inset, after(Srisanga, Flood et al. 2015)

According to the crystal growth rate dispersion (GRD) model (Srisanga, Flood et al. 2015),
which is more accurately describing the variation in the crystal growth rates within a
population of crystals, each individual crystal is growing with a unique constant rate. The
crystal growth rate dispersion is the assembling of all the growth rates in the population.
And the particle size distribution at any time is a linear expansion of the growth rate
distribution, shown in Figure 15. Thus, the overall crystal growth rate of the whole
population should be a constant as well.

35
 If checking the data points for the exposure time more than 3 hours, the weight percentage
of calcium carbonate produced has a linear relation with the exposure time. It means that
the growth rate of the carbonate is a constant, which is in consistent with the GRD model.
Therefore, the carbonation rate of C-S-H is a constant when considering the longer-time
carbonation. Clearly, the carbonation rate of C-S-H with a C/S of 0.66 is higher than that
with a C/S of 0.86, which is in consistent with the results found in FTIR test.

4 Conclusions

In the paper, the C-S-H with different C/S, identified by XRD and 29Si NMR, is successfully
synthesized from CaO and fumed silica in the solution protected by the N2 flow. The
carbonation mechanism of C-S-H including final products and rate are studied by NMR,
FTIR and XRD. The carbonation resistance among different types of C-S-H are discussed.
The main conclusions are as follow:

 The C-S-H(I) phases with the C/S ranging from 0.66 to 1.4 were synthesized
successfully from CaO and fumed silica in the solution, which is confirmed by
different test method. When the designed C/S ratio is higher than 1.40, the
portlandite will appear in the synthetic products.

 C-S-H phases with different C/S were all completely decomposed into silicate gel
after 3 days’ accelerated carbonation in this research. Carbonation rate of C-S-H with
a higher C/S is relatively lower than that of C-S-H with lower C/S, which has better
resistance to the carbonation.

Literature

Beaudoin, J. J., L. Raki and R. Alizadeh (2009). "A 29Si MAS NMR study of modified C–S–H
nanostructures." Cement and Concrete Composites 31(8): 585-590.
Black, L., C. Breen, J. Yarwood, K. Garbev, P. Stemmermann and B. Gasharova (2007).
"Structural features of C–S–H (I) and its carbonation in air—a Raman spectroscopic
study. Part II: carbonated phases." Journal of the American Ceramic Society 90(3): 908-917.
Chen, J. J., J. J. Thomas, H. F. Taylor and H. M. Jennings (2004). "Solubility and structure of
calcium silicate hydrate." Cement and Concrete Research 34(9): 1499-1519.

36
Cong, X. D. and R. J. Kirkpatrick (1996). "Si-29 MAS NMR study of the structure of calcium
silicate hydrate." Advanced Cement Based Materials 3(3-4): 144-156.
Delmi, M. M. Y., A. Aït-Mokhtar and O. Amiri (2006). "Modelling the coupled evolution of
hydration and porosity of cement-based materials." Construction and Building Materials
20(7): 504-514.
Doebelin, N. and R. Kleeberg (2015). "Profex: a graphical user interface for the Rietveld
refinement program BGMN." Journal of applied crystallography 48(5): 1573-1580.
Garbev, K., G. Beuchle, M. Bornefeld, L. Black and P. Stemmermann (2008). "Cell
Dimensions and Composition of Nanocrystalline Calcium Silicate Hydrate Solid
Solutions. Part 1: Synchrotron-Based X-Ray Diffraction." Journal of the American Ceramic
Society 91(9): 3005-3014.
Garbev, K., M. Bornefeld, G. Beuchle and P. Stemmermann (2008). "Cell Dimensions and
Composition of Nanocrystalline Calcium Silicate Hydrate Solid Solutions. Part 2: X-Ray
and Thermogravimetry Study." Journal of the American Ceramic Society 91(9): 3015-3023.
He, Y., L. Lu, L. J. Struble, J. L. Rapp, P. Mondal and S. Hu (2013). "Effect of calcium–silicon
ratio on microstructure and nanostructure of calcium silicate hydrate synthesized by
reaction of fumed silica and calcium oxide at room temperature." Materials and
Structures: 1-12.
L'Hôpital, É. M. (2014). Aluminium and alkali uptake in calcium silicate hydrates (C-S-H) PhD
thesis, EPFL.
Macphee, D., E. Lachowski and F. P. Glasser (1988). "Polymerization effects in CSH:
implications for Portland cement hydration." Advances in Cement Research 1(3): 131-137.
McMillan, P. F., G. H. Wolf and B. T. Poe (1992). "Vibrational spectroscopy of silicate
liquids and glasses." Chemical Geology 96(3): 351-366.
Megaw, H. D. and C. H. Kelsey (1956). "Crystal Structure of Tobermorite." Nature
177(4504): 390-391.
Merlino, S., E. Bonaccorsi and T. Armbruster (2001). "The real structure of tobermorite 11Å
normal and anomalous forms, OD character and polytypic modifications." European
Journal of Mineralogy 13(3): 577-590.
Morandeau, A., M. Thiery and P. Dangla (2014). "Investigation of the carbonation
mechanism of CH and CSH in terms of kinetics, microstructure changes and moisture
properties." Cement and Concrete Research 56: 153-170.
Ngala, V. T. and C. L. Page (1997). "Effects of carbonation on pore structure and diffusional
properties of hydrated cement pastes." Cement and Concrete Research 27(7): 995-1007.

37
Papadakis, V. G., C. G. Vayenas and M. Fardis (1989). "A reaction engineering approach to
the problem of concrete carbonation." AIChE Journal 35(10): 1639-1650.
Renaudin, G., J. Russias, F. Leroux, F. Frizon and C. Cau-dit-Coumes (2009). "Structural
characterization of C–S–H and C–A–S–H samples—Part I: Long-range order
investigated by Rietveld analyses." Journal of Solid State Chemistry 182(12): 3312-3319.
Richardson, I. (1999). "The nature of CSH in hardened cements." Cement and Concrete
Research 29(8): 1131-1147.
Richardson, I. (2004). "Tobermorite/jennite-and tobermorite/calcium hydroxide-based
models for the structure of CSH: applicability to hardened pastes of tricalcium silicate,
β-dicalcium silicate, Portland cement, and blends of Portland cement with blast-
furnace slag, metakaolin, or silica fume." Cement and Concrete Research 34(9): 1733-1777.
Richardson, I. and G. Groves (1992). "Models for the composition and structure of calcium
silicate hydrate (C−S-H) gel in hardened tricalcium silicate pastes." Cement and Concrete
Research 22(6): 1001-1010.
Richardson, I. and G. Groves (1997). "The structure of the calcium silicate hydrate phases
present in hardened pastes of white Portland cement/blast-furnace slag blends."
Journal of Materials Science 32(18): 4793-4802.
Šauman, Z. k. (1972). "Long-term carbonization of the phases 3CaO. Al2O3. 6H2O and
3CaO. Al2O3. SiO2. 4H2." Cement and Concrete Research 2(4): 435-446.
Srisanga, S., A. E. Flood, S. C. Galbraith, S. Rugmai, S. Soontaranon and J. Ulrich (2015).
"Crystal Growth Rate Dispersion versus Size-Dependent Crystal Growth: Appropriate
Modeling for Crystallization Processes." Crystal Growth & Design 15(5): 2330-2336.
Sykes, D. and J. D. Kubicki (1993). "A model for H2O solubility mechanisms in albite melts
from infrared spectroscopy and molecular orbital calculations." Geochimica et
Cosmochimica Acta 57(5): 1039-1052.
Taylor, H. and J. Howison (1956). "Relationships between calcium silicates and clay
minerals." Clay Minerals Bull 3(16): 98-111.
Wu, B. and G. Ye (2015). Development of porosity of cement paste blended with
supplementary cementitious materials after carbonation. 14th International Congress
on the Chemistry of Cement (ICCC2015).
Yu, P., R. J. Kirkpatrick, B. Poe, P. F. McMillan and X. Cong (1999). "Structure of Calcium
Silicate Hydrate (C‐S‐H): Near‐, Mid‐, and Far‐Infrared Spectroscopy." Journal of the
American Ceramic Society 82(3): 742-748.

38