See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/240749417

Physicochemical changes in nano-silica and silica fume modified cement

pastes in response to leaching

Article  in  International Journal of Materials and Structural Integrity · January 2009

DOI: 10.1504/IJMSI.2009.028608

CITATIONS READS

6 240

2 authors:

Jitendra Jain Narayanan Neithalath

Solidia Technologies, Inc Arizona State University
22 PUBLICATIONS   436 CITATIONS    168 PUBLICATIONS   4,453 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ultra High Performance Concrete View project

New iron-based carbon-negative building material View project

All content following this page was uploaded by Narayanan Neithalath on 16 December 2014.

The user has requested enhancement of the downloaded file.

114 Int. J. Materials and Structural Integrity, Vol. 3, Nos. 2/3, 2009

Physico-chemical changes in nano-silica and silica

fume modified cement pastes in response to leaching

Jitendra Jain and Narayanan Neithalath*

Department of Civil and Environmental Engineering,
Clarkson University,
Potsdam, NY 13699-5712, USA
E-mail: [email protected]
E-mail: [email protected]
*Corresponding author

Abstract: Leaching of calcium ions from cement pastes modified with silica
fume and nano-silica, and the physico-chemical changes that this induces in the
system are the focus of this paper. The pastes were subjected to slow (in
deionised water) and accelerated (in 6M ammonium nitrate) leaching. Mass
loss, increase in porosity and thermal analysis were used to quantify the
physical and chemical changes in the system due to leaching. The modified
cement pastes are found to exhibit better leaching resistance under both
exposure conditions. It is shown that leaching in deionised water and cementing
materials reaction are essentially coupled. When leached in ammonium nitrate,
it is observed that the entire CH is leached from all the pastes, and pastes with
very low initial CH contents suffer increased decalcification from C-S-H gel. It
is shown that, when subjected to very aggressive leaching agents, a certain
amount of CH acts as a buffer delaying C-S-H dissolution. The beneficial effect
of nano-silica in controlling calcium ion leaching especially when subjected to
accelerated leaching is brought out.

Keywords: nano-silica; silica fume; leaching; CH and C-S-H; deionised water;

ammonium nitrate.

Reference to this paper should be made as follows: Jain, J. and Neithalath, N.

(2009) ‘Physico-chemical changes in nano-silica and silica fume modified
cement pastes in response to leaching’, Int. J. Materials and Structural
Integrity, Vol. 3, Nos. 2/3, pp.114–133.

Biographical notes: Jitendra Jain is a PhD candidate at the Department of

Civil and Environmental Engineering, Clarkson University, Potsdam, NY
13699 USA.

Narayanan Neithalath is Assistant Professor at the Department of Civil and

Environmental Engineering, Clarkson University, Potsdam, NY 13699 USA.

1 Introduction

Leaching of calcium ions from cementitious matrices is a combined diffusion-dissolution

process that results in long term durability issues, and can be critical for underground and
under-water structures such as nuclear waste repositories. Pure water is one of the strong

Copyright © 2009 Inderscience Enterprises Ltd.

 Physico-chemical changes in nano-silica and silica fume modified cement 115

decalcifying agents for cement based materials (Adenot and Buil, 1992; Faucon et al.,
1996). The concentration gradients between the pore solution in concrete and the
surrounding medium results in diffusion of ions from the pore solution to the external
leaching medium. In order to correct the imbalance caused by the diffusion of calcium
ions from the pore solution into the surrounding leaching medium, calcium ions diffuse
into the pore solution from the calcium hydroxide (CH) and calcium silicate hydrate (C-
S-H) phases. This dissociation of the solid phases of the cement paste results in increased
porosity, and consequently loss of mechanical properties and increased permeability
(Kamali et al., 2003; Marchand et al., 2000; Carde et al., 1996; Carde and Francois,
1997). However, the leaching kinetics is slow in pure water, and hence laboratory
methods have also focused on accelerating the leaching kinetics of cement pastes,
especially by using chemical methods where ammonium nitrate is used as the leaching
medium (Chen et al., 2006; Carde et al., 1997; Agostini et al., 2007). Ammonium nitrate
reacts with the cement paste to form calcium nitrate as well as calcium nitro-aluminate.
Calcium from these compounds is easily transported to the surface, causing leaching. The
use of ammonium nitrate as a leaching medium just accelerates the process; it does not
alter the dissolution-diffusion nature of the degradation process.
Leaching kinetics also is significantly dependent on the initial porosity of the system
(Haga et al., 2005). Hence studies have focused on the use of supplementary cementing
materials such as fly ash and silica fume that can reduce the porosity due to pozzolanic
reaction, on the leaching resistance of cementitious systems (van Eijk and Brouwers,
1998; Jain and Neithalath, 2009). Pozzolanic reaction also results in a reduction in the
Ca-Si molar ratio of the C-S-H gel, which has also been reported to increase the leaching
resistance (Chatterji, 1995; Gaitero et al., 2008). In addition to the leaching environment
and porosity, leaching kinetics is also dependent on leachant volume (especially in the
case of pure water as the leaching agent) (Haga et al., 2002).
The focus of this study is on the influence of nano-silica and silica fume as cement
replacement materials, on the leaching resistance of cement pastes exposed to either
deionised water or ammonium nitrate solution. The cement replacement levels of silica
fume used in this study are about 10 times higher than that of nano-silica, which is in line
with the difference in their particle sizes which is an order of magnitude. Physical and
chemical changes occurring in the system as a result of decalcification are monitored
using volumetric and gravimetric methods, and thermal analysis.

2 Experimental program

2.1 Materials and mixtures

Type I ordinary Portland cement conforming to ASTM C 150 was used in all the cement
pastes described in this study. A commercial dry densified silica fume (SF) conforming to
ASTM C 1240, or a commercial nano-silica (NS), obtained from Degussa Materials,
were used as partial cement replacement materials. The chemical composition and
physical characteristics of these materials are given in Table 1. All the cement pastes
were prepared with a water-to-cementing materials ratio (w/cm) of 0.40. Cement
replacement by silica fume was carried out at dosages of 6% and 9% by mass, whereas
nano-silica was used at replacement levels of 0.5% or 1.5% by mass. The chosen dosages
of silica fume are those commonly used in construction applications. For nano-silica
116 J. Jain and N. Neithalath

modified pastes, the chosen dosages were dictated by the ability to disperse nano-silica in
a cement paste system. 1.5% by mass of nano-silica as cement replacement was found to
be an upper threshold with respect to mixing ability of the pastes and dispersability.
Nano-silica was dispersed in a mixture of the requisite mixing water and 1% of
superplasticiser by mass of cement using a magnetic stirrer before adding it to the other
constituents of the paste. The pastes were mixed in a laboratory blender and cast in
special acrylic moulds (Figure 1). Each specimen had a size of 50 mm × 50 mm × 3 mm.
The specimens were removed from the moulds after 24 hours and cured in saturated
limewater for a period of 8 weeks.
Table 1 Chemical composition and physical characteristics of cement, silica fume and
nano-silica

Composition (% by mass) / property Cement Silica fume Nano-silica

Silica (SiO2) 20.2 93.4 > 99.8
Alumina (Al2O3) 4.7 0.42 < 0.05
Iron oxide (Fe2O3) 3 0.52 < 0.003
Calcium oxide (CaO) 61.9 1.91 0
Magnesium oxide (MgO) 2.6 0 0
Sodium oxide (Na2O) 0.19 0.25 0
Potassium oxide (K2O) 0.82 0.79 0
Sulfur trioxide (SO3) 3.9 0.34 0
Loss on ignition 1.9 2.3 <1
Fineness, % passing 45 μm sieve 97.4 - > 99.95
2
Specific surface area (m /kg) 388 ~20000 ~130000
3
Density (kg/m ) 3150 - 50
Average particle size 30 microns 100 nm 15 nm

Figure 1 Acrylic mould used for preparing specimens for this study (see online version for
colours)
 Physico-chemical changes in nano-silica and silica fume modified cement 117

2.2 Test methods

2.2.1 Non-evaporable water contents to determine degrees of hydration
Non-evaporable water contents (wn) can be used as indicators of degrees of hydration in
the cement pastes. After eight weeks of curing in saturated limewater, wn values were
determined for all the pastes using thermogravimetric analysis (TGA). The specimens
were heated in a TG analyser to 1,050°C at a rate of 10°C per minute. The mass loss of
the specimen between 105°C and 1,050°C, corrected for the loss of ignition of the
unhydrated cement (or of the unhydrated cement and the replacement material multiplied
by their respective mass fractions) and the carbonate loss was recorded as wn.

2.2.2 Leaching procedure

The specimens cured in saturated limewater for eight weeks were subjected to two
different leaching procedures:
1 slow leaching in deionised water
2 accelerated leaching in a 6M ammonium nitrate solution.
For leaching in deionised water, the specimens of size 50 mm × 50 mm × 3 mm were cut
into 20 mm × 20 mm × 3 mm sizes and individually placed in plastic containers filled
with deionised water from which carbon dioxide was removed. The volume of leachant to
the volume of the paste was kept at 1,200. Eight specimens from each mixture were
subjected to leaching. These specimens were stored in the leachant for four weeks
(28 days). Deionised water was not changed during the leaching period. Three replicate
specimens were used for mass loss and porosity determination while another three
specimens were used for phase identification and quantification. The pH of the leachant
was measured at the end of the leaching duration. For leaching in 6M ammonium nitrate
solution, four specimens of 20 mm × 20 mm × 3 mm size were kept in 50 ml of
ammonium nitrate solution in separate containers. Two specimens each were used for
mass loss and porosity determination and the other two for phase quantification. Since
ammonium nitrate is a very strong decalcifying agent, the period of exposure was limited
to seven days. The specimens subjected to both these leaching procedures, as well as the
unleached specimens were subjected to tests to determine their porosities, and the CH and
C-S-H contents.

2.2.3 Porosity determination

The specimens saturated for 8 weeks, as well as those leached in deionised water for
28 days and 6M ammonium nitrate solution for seven days, were subjected to porosity
measurements as per the procedure described in RILEM CPC11.3 (1984). The 20 mm ×
20 mm × 3 mm size paste specimens were removed from the containers, surface-dried,
and kept in an oven at 105°C for 24 hours to remove all evaporable water. After cooling
to room temperature, the initial masses (m1) of the specimens were determined. The
specimens were further subjected to vacuum for three hours, after which they were
saturated with water under vacuum for another hour, and left to soak for 18 hours. The
masses (m2) were then recorded. The difference between these masses expressed as a
percent of the initial volume of the specimen is reported as the porosity.
118 J. Jain and N. Neithalath

2.2.4 Phase identification and quantification

The phases present in the unleached and leached cement pastes were determined using
X-ray diffraction (XRD) patterns obtained by subjecting the powdered samples to a
BrukerTM X-ray diffractometer between a 2θ range of 5° to 40° at a time step of ten
seconds.
The unleached and leached pastes were powdered and subjected to TGA to determine
the CH and C-S-H contents. TGA was carried out at the end of eight weeks of curing in
saturated limewater, after 28 days of exposure to deionised water, and after seven days of
exposure to 6M ammonium nitrate solution. The powdered sample was heated from room
temperature up to 1,050°C at a rate of 10°C per minute in a TG analyser. The mass drop
in the thermogravimetric curve at temperatures between 400°C and 450°C indicates loss
of water from CH. Representative TGA curves for the unleached plain paste as well as
those leached in deionised water and ammonium nitrate solution are shown in Figure 2.
The amount of CH in the specimen (as a percentage of mass of the sample at 105°C) is
calculated directly from the thermogravimetric curves using the following equation:
MWCH
CH (%) = WLCH (%) × (1)
MWH

where WLCH corresponds to the mass loss in percentage attributable to CH dehydration,

and MWCH and MWH are the molecular weights of CH (74.01 g/mol) and water
(18 g/mol) respectively.

Figure 2 TGA curves for the unleached and leached plain paste specimens (see online version for
colours)

100
Mass Remaining, %

90

Leached in
Unleached deionized water
80

Leached in ammonium nitrate

70
0 200 400 600 800 1000
Temperature, C o
 Physico-chemical changes in nano-silica and silica fume modified cement 119

To determine the water lost from C-S-H gel, previous studies have used mass losses
corresponding to temperature ranges of 180–300°C (Ruiz et al., 2005), 200–400°C
(Odelson et al., 2007), or from 105°C until the water loss corresponding to CH
dehydration (Taylor, 1997). In this study, the mass loss between 150°C and the
temperature at which CH loss begins (~400°C) is considered to indicate the dehydration
from C-S-H gel. The amount of C-S-H is calculated as:
MWCSH
C − S − H (%) = WLCSH (%) × a (2)
2.1* MWH

where WLCSH corresponds to the mass loss in percentage that occurs during C-S-H
dehydration, and MWCSH and MWH are the molecular weights of C-S-H (193.2 g/mol)
and water respectively. Even though the typical chemical formula of C-S-H is C1.7SH4,
some part of the 4 moles of water will be removed at temperatures below 150°C. The
equilibrium composition of C-S-H is thus taken as C1.7SH2.1 (Jain and Neithalath, 2009;
Taylor, 1997), and hence the value of 2.1 in the denominator of equation (2).
For the plain and modified cement pastes leached in 6M ammonium nitrate solution,
the thermogravimetric curves did not show any indication of the presence of CH. Hence a
temperature range between 150°C and 400°C is not used to quantify the amount of C-S-H
present in pastes exposed to ammonium nitrate. It is also possible that some of the mass
loss in this range is due to the complexes formed between the leaching calcium ions and
the penetrating ammonium nitrate. Therefore the calculation of the C-S-H contents of
pastes exposed to 6M ammonium nitrate solution in this paper is based on the loss of
mass of the pastes when subjected to ammonium nitrate. This is explained in detail in a
later section.

3 Results, analysis and discussions

3.1 Non-evaporable water contents

Figure 3 shows the non-evaporable water contents (wn) determined from TGA for all the
five pastes used in this study after 56 days of curing in saturated limewater (before
subjected to any of the leaching tests). It can be observed that the silica fume modified
pastes show comparable wn values as that of the plain paste whereas the nano-silica
modified pastes show slightly higher wn values. Even though it is known that the
replacement materials typically bind less water than the cement (Weng et al., 1997;
Schwarz and Neithalath, 2008), the higher wn values shown by the nano-silica modified
pastes in particular can be attributed to the enhancement in cement hydration facilitated
by the increased amount of surfaces provided by these ultrafine particles in the mixture.
If it can be assumed that the high silica replacement materials such as silica fume does
not bind any additional water due to its pozzolanic reaction (Papadakis, 1999), then the
entire increase in wn can be attributed to enhancement in cement hydration. In such a
case, the total degrees of hydration of the modified pastes can be calculated by dividing
the wn values with the wn of plain cement paste at complete hydration (0.23) for this
cement based on chemical composition and the ultimate wn values of the compound
phases). The degrees of hydration will be used in a later section where the calcium ion
concentrations in the hydration products will be estimated.
120 J. Jain and N. Neithalath

Figure 3 Non-evaporable water contents (wn) for the unleached specimens

24

20

16
Wn, %

12

Plain SF 6 SF 9 NS0.5 NS1.5

3.2 Porosity changes and mass loss due to leaching

As stated earlier, leaching is a diffusion-dissolution process, and it creates porosity within
the paste matrix. The porosities of the unleached and leached pastes were determined
using the procedure described earlier. Figure 4(a) shows the increase in paste porosity, Δφ
(the difference between the porosity after leaching and the porosity of the unleached
paste) as a result of leaching in deionised water for 28 days or in 6M ammonium nitrate
solution for seven days. The porosity increase when the pastes are exposed to ammonium
nitrate for seven days is about 3–4 times higher than when leached in deionised water for
28 days. This shows the accelerated removal of calcium ions from the cementitious
matrix by ammonium nitrate. CH is completely dissociated in less than seven days
resulting in higher porosities. This is also confirmed from the XRD spectra shown later in
this paper. Decalcification of C-S-H gel also occurs in this case. The plain paste shows
the highest values of Δφ for leaching under both the environments. The presence of
higher amounts of CH in the plain paste (as will be shown in the following section) that
results in an increased rate of leaching could be attributed to this observation. With
increase in silica fume or nano-silica dosage, Δφ values decrease, indicating the
beneficial effects of these replacement materials in increasing the leaching resistance of
cement pastes. It is well known that the highly reactive cement replacement materials
used in this study results in a reduction in CH contents as well as densifies the
microstructure through the formation of secondary C-S-H gel, both of which contribute to
increased leaching resistance.
The mass loss due to leaching (Δm) in both deionised water and ammonium nitrate,
as a percentage of the initial mass for all the pastes are shown in Figure 4(b). The
aggressiveness of the ammonium nitrate solution is evident in this case also. Similar
trends as shown in Figure 4(a) for Δφ are observed for Δm also, for similar reasons. The
response of the pastes to both types of leaching environments is similar, the only
difference being in the magnitude of either the mass loss or porosity increase.
 Physico-chemical changes in nano-silica and silica fume modified cement 121

Figure 4 (a) Change in porosity (Δφ) (b) mass loss (Δm) of specimens leached in deionised water
and ammonium nitrate (see online version for colours)
30
30 Deionized water Deionized water
6M Ammonium nitrate 6M Ammonium nitrate
Increase in Porosity (φL), %

20

Mass loss, %
20

10 10

0 0
Plain SF 6 SF 9 NS 0.5 NS 1.5 Plain SF 6 SF 9 NS 0.5 NS 1.5

(a) (b)
3.3 CH contents in the unleached and leached pastes

The CH contents in the unleached and leached pastes were determined from TGA using
equation (1). Figure 5 shows the CH contents remaining in cement pastes stored in
saturated lime water for eight weeks and those leached in deionised water for four weeks
after this curing duration. In the unleached state, the CH content is the highest in the plain
paste because of its higher cement content as compared to the modified mixtures. With
increase in silica fume or nano-silica dosage in the paste, the CH content decreases. This
is because of a combination of the reduced cement content in the modified pastes and the
consumption of CH by the high silica replacement material. The silica fume modified
pastes show lower CH contents than the nano-silica modified pastes, possibly because of
the low cement replacement levels in nano-silica modified pastes. It is also conceivable
that nano-silica facilitates increased hydration of cement by providing additional surfaces
as can be noticed from their higher wn values in Figure 3, thereby producing additional
CH. The 9% silica fume modified mixture is found to have the lowest CH content.
After leaching in deionised water for 28 days, the plain paste shows the largest loss in
CH. This is because of its higher initial CH content, which results in an increased rate of
leaching (Carde et al., 1995). With increase in the dosage of either silica fume or
nano-silica, the amount of CH remaining in the pastes after leaching is found to be
higher, indicating better leaching resistance. For higher cement replacement levels by
silica fume and nano-silica (9% and 1.5% respectively in this paper), the differences in
the CH contents between the unleached and leached specimens are very low. This can be
attributed to the lower porosities in these pastes that limit the influx of the leachant,
which reduces the dissolution of CH. Also, leaching of some CH from the external
surfaces and nearby zones provides pathways for the leachant (deionised water) into the
specimen, facilitating further cement hydration that produces more CH. While CH
production due to cement hydration compensates for some CH loss in low porosity pastes
(such as the ones with 9% silica fume or 1.5% nano-silica), for the higher porosity pastes
(such as the plain paste), even though there might be CH production from cement
hydration, the higher porosity results in faster CH loss also. Thus, calcium leaching and
microstructure development through hydration is seen to be essentially coupled.
122 J. Jain and N. Neithalath

Figure 5 CH contents of unleached pastes and the pastes leached in deionised water (see online
version for colours)
20
Unleached
Leached
16

CH content, % 12

0
Plain SF 6 SF 9 NS0.5 NS1.5

XRD spectra for unleached cement pastes, and those leached in deionised water and 6M
ammonium nitrate solution are shown in Figure 6. A comparison of Figures 6(a) and (b)
shows that the CH peaks (observed at 2θ of 18.4° and 34°) typically become smaller with
leaching in deionised water. Figure 6(c) shows the XRD spectra of pastes exposed to 6M
ammonium nitrate solution. CH peaks are entirely absent in those spectra, showing that
exposure to this aggressive solution leaches out the entire calcium ions from CH. This is
in conformance with previously reported studies (Agostini et al., 2007; Carde et al., 1997;
Ulm et al., 1999) that suggest complete dissolution of CH after about seven days of
accelerated calcium leaching. The only distinguishable peaks observed in Figure 6(c) are
those corresponding to weakly crystalline C-S-H with low Ca-Si molar ratios at 2θ of
around 30°.

Figure 6 XRD patterns for (a) unleached pastes (b) pastes leached in deionised water (c) pastes
leached in ammonium nitrate (see online version for colours)

(a) 2θ(°) (b) 2θ(°) (c) 2θ(°)

 Physico-chemical changes in nano-silica and silica fume modified cement 123

This is similar to the XRD patterns for low Ca-Si molar ratio C-S-H gel reported in
(Sugiyama, 2008). This peak is not conspicuous for unleached pastes as well as pastes
leached in deionised water because the molar ratio of Ca to Si is higher in those pastes as
will be shown later in this paper while discussing equilibrium curves for calcium.
The pH of the leachant solutions were also measured at the end of the leaching
duration for the pastes in deionised water. The lower initial pH of the leachant as
compared to the pore solution drives the movement of ions from pore solution into the
leachant, thus increasing the pH of the leachant. The leachant solution in which the plain
paste was stored showed the highest pH (8.35), and the one in which 1.5% nano-silica
modified paste was stored showed the lowest pH (8.1), which also corroborates the
observations made earlier of increased leaching resistance with increase in nano-silica
dosage.

3.3.1 Depths of CH dissolution fronts

The depths of the dissolution front of CH (dCH) in plain and modified pastes exposed to
deionised water were determined in this study using an indirect method reported in
(Jain and Neithalath, 2009). The CH lost as a percentage of mass of the paste due to
leaching can be approximated as:
(CH )lost = (CH )unleached − (CH )leached (3)

where (CH)unleached and (CH)leached are the CH contents in the unleached and the leached
specimens obtained from thermogravimetric data. (CH)lost was converted into a volume
percentage, along with the known volume of the sample can be used to estimate the
volume of CH leached. Dividing the volume by the combined area of the exposed
surfaces gives the leaching depth.
Figure 7 shows dCH values for the plain and modified pastes after 28 days of leaching
in deionised water. A dCH of 0.40 mm is obtained for the plain paste after 28 days of
leaching. The dissolution depths for nano-silica and silica fume modified pastes are lower
(~0.3 mm) than that of the plain paste, indicating better leaching resistance.

Figure 7 CH dissolution depth for pastes leached in deionised water

0.5
Dissolution depth(dCH), mm

0.4

0.3

0.2

0.1

0
Plain SF 6 SF 9 NS0.5 NS1.5
124 J. Jain and N. Neithalath

It can be noticed from this figure that the all the modified pastes show similar dissolution
depths. This is not in direct agreement with the results in Figure 5 where the difference in
CH contents between the unleached and leached states (that is, the amount of CH
leached) for the pastes modified with higher dosages of replacement materials was seen
to be the lowest. The reason for this observed disparity lies in the approximation used in
calculating the dissolution depth. It was assumed that the difference between the CH
contents of the unleached and the leached specimens amount to the CH lost due to
leaching. It was shown in the previous section that calcium leaching and microstructure
development are essentially coupled. In other words, while CH is lost from the specimens
because of leaching, additional cement hydration might replenish some CH, especially in
low porosity systems. Hence the difference between the CH contents of unleached and
leached pastes cannot be directly taken as the amount of CH lost due to leaching for an
accurate determination of dissolution depth. If the calculated values are corrected based
on the above explanation, the dissolution depths for the modified pastes would increase.
This is not counter-intuitive because in modified pastes, to dissolve same amount of CH,
the leachant has to penetrate deeper into the specimen because of the lower CH contents.
However, this approximate calculation provides a distinction between the plain and
modified pastes, and brings out the effect of these replacement materials in ensuring
better leaching resistance of pastes.

3.4 C-S-H contents in the unleached and leached pastes

Quantification of the C-S-H contents is important in ascertaining the leaching resistance
of cement pastes. One reason for this being the microstructural densification imparted by
the use of cement replacement materials such as silica fume and nano-silica is a result of
formation of increased amounts of C-S-H gel. This in turn reduces the porosity and
provides a variety of beneficial properties to the material. A higher amount of C-S-H is
beneficial towards better leaching resistance than CH from the viewpoint of chemical
equilibrium also because C-S-H is stable even at lower concentrations of calcium ions in
the solution.
Figure 8 shows the C-S-H contents as a percentage of the mass of plain and modified
pastes subjected to both the leaching environments employed in this study along with that
in the unleached state. Equation (2) was used to determine the C-S-H contents of the
unleached pastes as well as those exposed to deionised water. For pastes leached in
ammonium nitrate solution, the loss of mass is due to the leaching of calcium ions from
both CH and C-S-H phases. The difference between total mass lost and the mass loss
attributed to CH (all CH is lost when exposed to ammonium nitrate solution as seen from
the XRD spectra and TGA data) gives the mass loss due to C-S-H dissolution. The mass
lost due to C-S-H dissolution was subtracted from the original mass of C-S-H in the
pastes to determine the C-S-H contents remaining after leaching.
The C-S-H contents in the unleached pastes increase with increase in the dosage of
the replacement material, attributable to the pozzolanic reaction of the high silica
replacement materials with the CH formed from cement hydration. The C-S-H content is
higher for the nano-silica modified pastes as compared to the silica fume modified pastes
possibly because of the increased reactivity of the nano-silica particles. When the C-S-H
contents of the unleached pastes and those leached in deionised water are compared, it
can be seen that the C-S-H content of the plain paste is reduced. This indicates that, in
addition to leaching from CH, a small amount of calcium ions are also removed from
 Physico-chemical changes in nano-silica and silica fume modified cement 125

C-S-H. Previous studies (Jain and Neithalath, 2009; Haga et al., 2002) have also reported
such observations. This is because, at higher leachant to paste volume (or mass) ratios
similar to the one used in this study (1,200), even with some CH dissolution the leachant
is under-saturated with CH. This results in removal of some Ca ions from C-S-H, thereby
changing the structure of C-S-H.

Figure 8 C-S-H contents of unleached and leached pastes (see online version for colours)

Unleached
Deionized Water
60
6M ammonium nitrate
C-S-H content, %

40

20

0
Plain SF 6 SF 9 NS0.5 NS1.5

Most of the modified pastes are seen to have a higher C-S-H content after leaching in
deionised water than the unleached specimens. This is another indication of the coupling
of leaching and hydration as described earlier, where some of the unhydrated cement
particles hydrate in the presence of the leachant diffusing into the paste, or unreacted
particles of nano-silica or silica fume react with CH in the presence of deionised water
that is diffusing into the microstructure. However, this is not the case for 0.5%
nano-silica modified paste because of the very small quantity of nano-silica in the matrix,
which would have mostly reacted by the time leaching test was started. Further cement
hydration is only seen to compensate for the loss of calcium ions from C-S-H gel.
For the pastes leached in ammonium nitrate solution, it is observed from Figure 8 that
the 9% silica fume modified paste has a lower C-S-H content after leaching compared to
the other modified pastes. The reason for this observation lies in the CH contents of the
unleached pastes shown in Figure 5. This paste was seen to have the lowest CH content
in the unleached state for reasons described earlier. When subjected to a highly
aggressive leachant such as ammonium nitrate, the smaller amount of CH present is
dissociated very quickly, exposing the C-S-H gel to leaching. Thus, by the end of the
leaching period (7 days), more Ca ions have been lost from C-S-H. This has implications
when the total Ca ions lost from the solid phases is determined in a later section. The
increase in porosity shown in Figure 3(a) does not really mimic this trend because
porosity determination is more sensitive to the larger pores created as a result of CH
dissociation rather than the smaller pores created by C-S-H decalcification. The
significant observation from these results is that, while a very efficient pozzolanic
reaction that consumes all the CH is advantageous from the viewpoint of leaching
resistance to deionised water, when highly aggressive leachants are used, some amount of
CH in the matrix provides a buffer to the C-S-H gel and delays its decalcification. This of
126 J. Jain and N. Neithalath

course is also dependent on the paste porosity – a lower porosity helping to delay the
progress of leaching.
Leaching of calcium ions from the C-S-H gel of the modified pastes will also be
slower than that from plain pastes because of the lower Ca-Si molar ratio in modified
pastes. Secondary C-S-H gel typically has a Ca-Si ratio of 1.1 as opposed to 1.7 for the
primary hydration product. Lower Ca-Si molar ratio has also been shown to result in
increased silicate anion chain lengths. This can happen either because of progress of
dissolution (Haga et al., 2002) or because of the incorporation of nano-silica particles in
the matrix (Gaitero et al., 2008).

3.5 Calcium equilibrium curves

The previous sections in this paper has described the influence of silica fume and
nano-silica on the leaching resistance of cement pastes through experiments that focused
on measuring the masses, porosities, and the phase contents. In this section, solid-liquid
equilibrium curves for calcium will be developed for unleached and leached pastes.
Equilibrium curves depict the relationship between the calcium ion concentrations in the
solid phases (CH and C-S-H gel) and in the pore solution. These curves are helpful in:
1 quantifying the calcium ion concentrations in CH and C-S-H gel for pastes subjected
to various exposure conditions
2 bringing out the influence of the leachant type and the replacement material type in
leaching resistance
3 providing a quantitative idea of the influence of Ca-Si molar ratio (and thus the
calcium ion concentration in the liquid) on the leaching resistance.

Figure 9 A typical solid-liquid equilibrium curve for calcium ions (see online version for colours)

Ctotal
Ca ion in solids, mmol/m3

CCH
CC-S-H

2 mol/m3

0 5 10 15 20 25
Ca ion in liquid, mol/m3

Figure 9 shows a typical solid-liquid equilibrium curve for calcium ions. CCSH is the
calcium ion concentration in C-S-H gel, and CCH is the calcium ion concentration in the
 Physico-chemical changes in nano-silica and silica fume modified cement 127

CH phase. CH dissolution occurs when the Ca ion concentration in the pore solution
drops below 22 mol/m3. C-S-H gels are stable between liquid Ca ion concentrations of
22 mol/m3 and 2 mol/m3, depending on the Ca-Si molar ratio of the gel (higher the Ca-Si
molar ratio, higher the equilibrium liquid ion concentration). At Ca ion concentrations in
liquid of less than 2 mol/m3, only the silicon rich gel is stable.
Development of calcium equilibrium curves requires determination of CCSH and
CCH. The total calcium and silicon ion concentrations in the C-S-H and CH phases for
both plain and modified pastes were determined stoichiometrically from the
standard hydration equations of C3S and C2S as well as the reaction of CH with S to
produce C1.1SH3.9. The Ca ion concentrations determined were adjusted for the actual
degrees of hydration of pastes determined from the non-evaporable water content, as
mentioned in Section 3.1. Detailed account of the above procedure including
the equations used is provided in Jain and Neithalath (2009). Based on the concentrations
of Ca and Si ions calculated, the Ca-Si molar ratio of the C-S-H gel in the paste was
computed. The value of Ca-Si molar ratio is 1.7 for the C-S-H gel in the plain
paste whereas it reduces for modified pastes because of the presence of secondary C-S-H
gel with a lower Ca-Si molar ratio. Table 2 shows Ca/Si for the plain and modified
pastes.
For the pastes leached in deionised water or 6M ammonium nitrate, the CH and
C-S-H contents remaining in the specimen were shown in Figures 5 and 8 respectively.
The difference between the CH or the C-S-H contents of the unleached paste and the
leached paste was used to calculate the net amount of Ca ions lost from CH and C-S-H.
This in turn was converted into moles of Ca ions per m3 and subtracted from the Ca ion
concentration of CH and C-S-H in the unleached paste to provide the Ca ion
concentration of the leached pastes. The Ca-Si molar ratios are lower for the leached
pastes as seen from Table 2 because it is only the Ca ions that are leached, and Si remains
in the solid phase.

Table 2 Ca-Si molar ratios and liquid ion concentrations in unleached and leached pastes

Ca/Si molar ratio Liquid ion concentration (mmol/l)

Unleached DI water NH4NO3 Unleached DI water NH4NO3
Plain 1.7 1.64 1.04 22 20.43 4.56
SF6 1.64 1.65 1.04 20.43 20.69 4.56
SF9 1.63 1.64 0.8 20.17 20.43 3.5
NS05 1.65 1.63 1.31 20.69 20.17 11.7
NS1.5 1.63 1.63 1.61 20.17 20.17 19.64

3.5.1 Equilibrium curves for unleached and leached pastes and interpretations
Using the CCH and CCSH values calculated as described above for the unleached and
leached pastes, calcium equilibrium curves can be constructed using a model proposed in
(Nakarai et al., 2006). In this model, the Ca ion concentration in the solid phase (Csolid) is
expressed as a function of the Ca ion concentration in the liquid phase (Cliquid) as shown
below:
128 J. Jain and N. Neithalath

⎡ 1⎤
⎢ ⎛ Cliquid ⎞ 3 ⎥
CSolid = A ⎢CCSH ⎜ ⎟ +B (4)
⎝ Csat ⎠ ⎥
⎢ ⎥
⎣ ⎦
In this equation, Csat is the saturated liquid phase Ca ion concentration (mol/m3 or
mmol/l), which was taken as 22 mmol/l for the plain pastes. This value changes with
changes in Ca-Si molar ratio. A thermodynamic dissolution model for C-S-H (Sugiyama
and Fujita, 2006) is used in this paper to determine the saturated liquid phase Ca ion
concentrations based on the Ca-Si molar ratio of C-S-H gel. The values of Csat are also
provided in Table 2 along with the Ca-Si molar ratios.
The constants A and B in equation (6) are defined as follows: x1 = 2 mol/m3, which is
the Ca ion concentration in the liquid below which no C-S-H gel exists, and x2 is the Ca
ion concentration at which CH has completely dissolved. For computational ease, x2 is
taken as (Csat – 3) mol/m3 (Nakarai et al.2006).

⎧ 2 3 3 2
⎪− 3 C liquid + 2 Cliquid for 0 ≤ Cliquid < x1
A = ⎨ x1 x1 (5)
⎪0 for x2 < Cliquid
⎩

⎧0 for 0 ≤ Cliquid < x1

⎪
B=⎨ CCH (6)
( )
3
⎪ . Cliquid − x2 for x2 < Cliquid
⎩ ( Csat − x2 )
3

From the CCSH and CCH values determined as described earlier, solid-liquid equilibrium
curves can be constructed for unleached and leached specimens using equations (4)–(6)
and are shown in Figure 10. Figure 10(a) shows the solid-liquid equilibrium curves for
the unleached pastes. The plain paste shows the highest amount of Ca ions in CH as
expected. The paste with 1.5% nano-silica shows the highest amount of Ca ions in C-S-H
gel. All the modified pastes show lower amount of Ca ions in CH and higher amount of
Ca ions in C-S-H, because of the pozzolanic reaction. It can also be noticed from this
figure that the equilibrium calcium ion concentration in the liquid moves towards lower
values with increase in dosages of nano-silica or silica fume, with the 1.5% nano-silica
modified paste showing the lowest value. This is an indication of the increased stability
of C-S-H gel in modified pastes.
Figure 10(b) shows the solid-liquid equilibrium curves for the pastes leached for four
weeks in deionised water. All the pastes have lost Ca ions from CH as evident from the
lengths of the CH arms in these equilibrium curves as compared to those in the unleached
pastes. The plain paste is seen to have lost the maximum amount of Ca ions from CH, the
reason for which has been elucidated earlier. The nano-silica modified pastes show
higher concentration of Ca ions in C-S-H gel than the silica fume modified pastes. None
of the modified pastes show appreciable loss of Ca ions from C-S-H gel after being
subjected to 28 days in deionised water. Figure 10(c) shows the equilibrium curves for
pastes leached in 6M ammonium nitrate solution for 7 days. CH from all the pastes is
completely dissociated. Due to the very aggressive nature of the leachant, Ca ions from
C-S-H are also rapidly lost, especially for the plain pastes. The silica fume modified
 Physico-chemical changes in nano-silica and silica fume modified cement 129

pastes also have lost a majority of Ca ions from C-S-H during this exposure period.
However, the nano-silica modified pastes are found to suffer relatively lower loss of Ca
ions from C-S-H gel.

Figure 10 Calcium equilibrium curves for (a) unleached pastes (b) pastes leached in deionised
water (c) pastes leached in ammonium nitrate (see online version for colours)
12 12
Ca ion in solids, (x106) mmol/m3

Ca ion in solids, (x106) mmol/m3

NS 1.5
NS 1.5
SF6
8 8
NS 0.5
NS 0.5 SF9
SF 6

Plain SF 9
4 4 Plain

0 0
0 5 10 15 20 25 0 5 10 15 20 25
Ca ion in liquid, mmol/l Ca ion in liquid, mmol/l

(a) (b)
12
Ca ion in solids, (x106)mmol/m3

NS 0.5
NS 1.5
4 SF6 Plain
SF9

0
0 5 10 15 20 25
Ca ion in liquid, mmol/l

(c)

3.6 Total Ca ions lost from the pastes: influence of leaching environment and
material microstructure
A comparison of the total loss of calcium ions (from both CH and C-S-H) as a result of
both the leaching methods for all the pastes is shown in Figure 11. The loss of Ca ions as
a percentage of the initial amount of Ca ions is shown here. The data for these
calculations were extracted from the equilibrium curves. When exposed to deionised
water, the plain paste suffers the maximum loss of Ca ions. Also, the higher leachant to
paste volume ratio results in simultaneous C-S-H decalcification. The 9% silica fume
modified paste and the 1.5% nano-silica modified paste show an effective gain of Ca ions
when leached in deionised water for 28 days. This goes back to the discussions in
Sections 3.3 and 3.4 where it was shown that microstructural alterations due to leaching
in deionised water are produced by the combined effect of loss of Ca ions from the matrix
130 J. Jain and N. Neithalath

and production of more hydration products facilitated by the ingress of water. These
pastes were shown to have higher C-S-H contents after leaching (Figure 8), and a lower
difference in CH contents between the unleached and the leached states (Figure 5), both
of which can be attributed to increased cement hydration and/or pozzolanic reaction. For
the 0.5% nano-silica modified paste, the C-S-H content after leaching was lower than that
in the unleached state, and the amount of CH leached higher, resulting in a net loss of Ca
ions. Even though the 6% silica fume modified mixture showed a higher C-S-H content
after leaching in deionised water, the CH leached from this mixture was large enough to
nullify any increase in total Ca ion concentration.

Figure 11 Total calcium ions lost from the pastes as a function of initial Ca content for pastes
leached under both the conditions (see online version for colours)
80
Deionized water
70 6M ammonium nitrate
Loss of Calcium ions, %

60

50

40

30

20

10

-10 Plain SF 6 SF 9 NS0.5 NS1.5

The mass loss for pastes leached in 6M ammonium nitrate presents a somewhat different
picture. In this case, the 9% silica fume modified paste is seen to suffer the maximum
loss of Ca ions. It was observed from the C-S-H contents in Figure 8 that this paste had
the lowest C-S-H content among the modified pastes after leaching in ammonium nitrate.
The lower initial CH content in this paste because of the very effective pozzolanic
reaction and a lower cement content results in the C-S-H gel being subjected to
decalcification sooner than it would have been in the case of a paste with a slightly higher
CH content. While a reduction in CH and increase in C-S-H contents as a consequence of
pozzolanic reaction is mostly a favourable outcome, when such pastes are subjected to
aggressive leaching agents, this presents a risk of exposing the C-S-H gel to
decalcification early on. However, a reduction in porosity of the modified paste because
of the pozzolanic reaction can be expected to offset this effect to a certain extent. Based
on the observations in this study, when ammonium nitrate is used as the leaching agent,
presence of some amount of CH in the system delays the C-S-H gel decalcification. A
paste with the lowest initial CH content (because of the pozzolanic reaction) suffers
maximum loss of Ca ions from C-S-H gel.
6% silica fume modified paste as well as both the nano-silica modified pastes
leached in ammonium nitrate are found to exhibit better leaching resistance, with the
1.5% nano-silica paste showing the lowest loss of Ca ions, and thus the maximum
leaching resistance among all the pastes studied. This is in line with the results obtained
from mass loss and porosity studies. Lower initial porosity as a result the pozzolanic
reaction that converts some of the CH into C-S-H gel that has a higher leaching resistance
 Physico-chemical changes in nano-silica and silica fume modified cement 131

because of its lower Ca-Si ratio, and possibly the increase in silicate chain length
provided by the incorporation of nano-silica as stated in (Gaitero et al., 2008) can be
attributed as the reasons for this behaviour.

4 Conclusions

This paper has dealt with the physico-chemical changes that occur in pastes with cement
replacements of 6% or 9% by silica fume and 0.5% or 1.5% by nano-silica by mass, and
exposed to deionised water or 6M ammonium nitrate as the leaching agents. The physical
changes in the system evaluated in this study were mass loss and increase in porosity as a
result of leaching. The chemical changes primarily result from the loss of Ca ions from
the CH and C-S-H phases. The conclusions presented below are applicable for the
dosages and types of the cement replacement materials used, the specimen sizes, and the
leaching environments and durations adopted in this study. This paper focuses more on
quantifying the overall influence of the replacement materials on leaching resistance with
respect to the chemistry of the hydration products, and a statistical analysis is not
attempted.
• The modified pastes were found to have lower mass losses and increased porosities
as compared to the plain paste under exposure to both the leaching agents. The paste
with the higher nano-silica dosage was observed to have the least values for both
these parameters, indicating the effectiveness of nano-silica in providing leaching
resistance to the pastes. Lower initial porosity and the pozzolanic reaction of
nano-silica (that lowers the initial porosity as well as provides further C-S-H during
leaching) are believed to be responsible for this effect. The lower Ca-Si molar ratios
of the modified pastes also contributed to their increased leaching resistance.
• All the modified pastes were found to have lower CH contents and higher C-S-H
contents than those of the plain paste in the unleached and leached states. When the
pastes were leached in deionised water at a high leachant to paste volume ratio, some
C-S-H decalcification also occurred in addition to CH dissolution because the
leachant is under-saturated with CH. For the pastes exposed to ammonium nitrate,
dissolution of entire CH occurred in less than seven days, along with some
decalcification of C-S-H. Hence, in the paste that had the lowest amount of CH,
C-S-H decalcification begins earlier, and more Ca ions are lost from C-S-H. A highly
efficient pozzolanic reaction can result in reduced porosities and thus better leaching
resistance, but when exposed to very aggressive leachants, the lower CH content
might result in faster exposure of the C-S-H gel to the leachant and consequently
higher decalcification of the C-S-H gel.
• From the difference in the CH and C-S-H contents between the unleached and
leached states, it was shown that calcium ion leaching in deionised water is coupled
with cement hydration and/or pozzolanic reaction of the replacement materials. The
net loss of Ca ions from the paste, which indicates the leaching resistance (or the
lack of it) of cement pastes depends on which of these mechanisms is dominant.
However, it needs to be noted that for longer durations of leaching in deionised
water, there will always be a net loss of Ca ions from the paste.
132 J. Jain and N. Neithalath

• Solid-liquid equilibrium curves for calcium were generated for the unleached and
leached pastes. These curves serve as a useful aid to visualise the change in the Ca
ion concentrations in CH and C-S-H phases of plain and modified pastes subjected to
leaching. While no significant difference was noticed between the responses of silica
fume and nano-silica pastes to leaching in deionised water, the equilibrium curves
for pastes leached in ammonium nitrate clearly showed the beneficial nature of
nano-silica in reducing Ca ion leaching.

References
Adenot, F. and Buil, M. (1992) ‘Modeling of the corrosion of the cement paste by deionized water’,
Cement and Concrete Research, Vol. 22, pp.489–496.
Agostini, F., Lafhaj, Z., Skoczylas, F. and Loodsveldt, H. (2007) ‘Experimental study of
accelerated leaching on hollow cylinders of mortar’, Cement and Concrete Research, Vol. 37,
pp.71–78.
Carde, C., Francois, R. and Ollivier, J.P. (1995) ‘Microstructural changes and mechanical effects
due to the leaching of calcium hydroxide from cement paste’, Proceedings of the Material
Research Society's Symposium on Mechanisms of Chemical Degradation of Cement-based
Systems, Vol. 27–30, pp.30–37, Boston, USA.
Carde, C. and François, R. (1997) ‘Effect of the leaching of calcium hydroxide from cement paste
on mechanical and physical properties’, Cement and Concrete Research, Vol. 27, No. 4,
pp.539–550.
Carde, C., Escadeillas, G. and François, R. (1997) ‘Use of ammonium nitrate solution to simulate
and accelerate the leaching of cement pastes to deionized waters’, Magazine of Concrete
Research, Vol. 49, No. 181, pp.295–301.
Carde, C., Franqois, R. and Torrenti, J.M. (1996) ‘Leaching of both calcium hydroxide and C-S-H
from cement paste: modeling the mechanical behavior’, Cement and Concrete Research,
Vol. 26, No. 8, pp.1257–1268.
Chatterji, S. (1995) ‘Concrete durability and CaO/SiO mole ratio of C-S-H’, Cement and Concrete
Research, Vol. 25, No. 5, pp.929–932.
Chen, J.J., Thomas, J.J. and Jennings, H.M. (2006) ‘Decalcification shrinkage of cement paste’,
Cement and Concrete Research, Vol. 36, No. 5, pp.801–809.
Faucon, P., Bescop, P.L., Adenot, F., Bonville, P., Jacquinot, J.F., Pineau, F. and Felix, B. (1996)
‘Leaching of cement paste: study of the surface layer’, Cement and Concrete Research,
Vol. 26, No. 11, pp.1707–1715.
Gaitero, J.J., Campillo, I. and Guerrero, A. (2008) ‘Reduction of calcium leaching ratio of cement
paste by addition of silica nanoparticles’, Cement and Concrete Research, Vol. 38,
pp.1112–1118.
Haga, K., Shibata, M., Hironaga, M., Tanaka, S. and Nagasaki, S. (2002) ‘Silicate anion structural
change in calcium silicate hydrate gel on dissolution of hydrated cement’, Journal of Nuclear
Science and Technology, Vol. 39, No. 5, pp.504–547.
Haga, K., Sutou, S., Hironaga, M., Tanaka, S. and Nagasaki, S. (2005) ‘Effects of porosity on
leaching of Ca from hardened ordinary Portland cement paste’, Cement and Concrete
Research, Vol. 35, pp.1764–1775.
Jain, J. and Neithalath, N. (2009) ‘Analysis of calcium leaching behavior of plain and modified
cement pastes in pure water’, Cement Concrete Composites, in press.
Kamali, S., Gerard, B. and Moranville, M. (2003) ‘Modeling the leaching kinetics of cement based
materials – influence of materials and environment’, Cement and Concrete Composites,
Vol. 25, pp.451–458.
 Physico-chemical changes in nano-silica and silica fume modified cement 133

Marchand, J., Bentz, D.P., Samson, E. and Maltais, Y. (2000) ‘Influence of calcium hydroxide
dissolution on the transport properties of hydrated cement systems’, Proccedings of Workshop
on the Role of Calcium Hydroxide in Concrete 2000, pp.113–129, Holmes Beach, Anna Maria
Island, FL.
Nakarai, K., Ishida, T. and Maekawa, K. (2006) ‘Modeling of calcium leaching from cement
hydrates couples with micro-pore solution formation, Journal of Advanced Concrete
Technology, Vol. 4, No. 3, pp.395–407.
Odelson, J.B., Kerr, E.A. and Vadakan, W.V. (2007) ‘Young’s modulus of cement paste at elevated
temperatures’, Cement and Concrete Research, Vol. 37, No. 2, pp.258–263.
Papadakis, V.G. (1999) ‘Experimental investigation and theoretical modeling of silica fume activity
in concrete’, Cement and Concrete Research, Vol. 29, No. 1, pp.79–86.
RILEM CPC 11.3. (1984) ‘Absorption of water by immersion under vacuum’, Materials and
Structures, Vol. 17, pp.391–394.
Ruiz, L.A., Platret, G., Massieu, E. and Ehrlacher, A. (2005) ‘The use of thermal analysis in
assessing the effect of temperature on a cement paste’, Cement and Concrete Research,
Vol. 35, No. 3, pp.609–613.
Schwarz, N. and Neithalath, N. (2008) ‘Influence of a fine glass powder on cement hydration:
Comparison to fly ash and modeling the degree of hydration’, Cement and Concrete Research,
Vol. 38, pp.429–436.
Sugiyama, D. and Fujita, T. (2006) ‘A thermodynamic model of dissolution and precipitation of
calcium silicate hydrates’, Cement and Concrete Research, Vol. 36, pp.227–237.
Sugiyama, D. (2008) ‘Chemical alteration of calcium silicate hydrate (C–S–H) in sodium chloride
solution’, Cement and Concrete Research, Vol. 38, No. 11, pp.1270–1275.
Taylor, H.F.W. (1997) Cement Chemistry, Thomas Telford, London.
Ulm, F.J., Torrenti, J.M. and Adenot, F. (1999) ‘Chemoporoplasticity of calcium leaching in
concrete’, ASCE Journal of Engineering Mechanics, Vol. 125, No. 10, pp.1200–1211.
van Eijk, R.J. and Brouwers, H.J.H. (1998) ‘Study of the relation between hydrated Portland
cement composition and leaching resistance. Cement and Concrete Research, Vol. 28, No. 6,
pp.815–828.
Weng, J.K., Langan, B.W. and Ward, M.A. (1997) ‘Pozzolanic reaction of portland cement, silica
fume, and fly ash mixtures’, Canadian Journal of Civil Engineering, Vol. 24, pp.754–760.

View publication stats