Modification of cement-based materials with nanoparticles
Shiho Kawashima
a,
⇑
, Pengkun Hou
a,b
, David J. Corr
a
, Surendra P. Shah
a
a
Department of Civil and Environmental Engineering, Northwestern University, 2145 Sheridan Road, A236, Evanston, IL 60208, United States
b
College of Materials Science and Engineering, Chongqing University, Chongqing 400045, PR China
article info
Article history:
Received 1 February 2012
Received in revised form 30 April 2012
Accepted 7 June 2012
Available online 27 June 2012
Keywords:
Clay
NanoCaCO
3
NanoSiO
2
Fly ash
Rheology
Nanomodification
abstract
This is a summary paper on the work being done at the Center for Advanced Cement-Based Materials at
Northwestern University on the modification of cement-based materials with nanoparticles, specifically
nanoclays, calcium carbonate nanoparticles, and nanosilica. The rheological properties of clay-modified
cement-based materials are investigated to understand the influence of nanoclays on thixotropy. The
influence of the method of dispersion of calcium carbonate nanoparticles on rate of hydration, setting,
and compressive strength are evaluated. And an in-depth study on the mechanisms underlying the
influence of nanosilica on the compressive strength gain of fly ash–cement systems is discussed. The
motivation behind these studies is that with proper processing techniques and fundamental under-
standing of the mechanisms underlying the effect of the nanoparticles, they can be used to enhance
the fresh-state and hardened properties of cement-based materials for various applications. Nanoclays
can increase the green strength of self-consolidating concrete for reduced formwork pressure and slip-
form paving. Calcium carbonate nanoparticles and nanosilica can offset the negative effects of fly ash
on early-age properties to facilitate the development of a more environmentally friendly, high-volume
fly ash concrete.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
Due to the complexity of concrete, which is heterogeneous at
all length scales, and the recent innovations in nanotechnology,
nanomodification of cement-based materials has generated much
research interest. Nanomodification is the manipulation of the
structure at the nanoscale (less than 100 nm) to develop cement
composites that exhibit enhanced or novel properties and func-
tions. Carbon nanotubes (CNTs) dispersed by ultrasonication can
significantly improve the flexural strength of cement composites
by controlling cracks at the nanoscale [1]. Through other mecha-
nisms, often attributed to filler and/or seeding effects, nanoparti-
cles can accelerate rate of hydration and improve early-age
mechanical properties of cementitious materials, including those
with cement replacement by fly ash. Among the types of nanopar-
ticles investigated are titanium dioxide (TiO
2
) nanoparticles [2],
zinc dioxide (ZnO
2
) nanoparticles [3], calcium carbonate (CaCO
3
)
nanoparticles [4,5], and nanoclays [6], although the majority of
studies thus far have focused on nanosilica [7–12]. However, due
to the novelty of the technology, more investigation needs to be
done to further understand the mechanisms underlying the effect
of the nanoparticles, to improve processing, and to evaluate their
influence at later-ages.
At the Center for Advanced Cement-Based Materials at North-
western University (ACBM-NU), work is being done on modifying
the fresh-state and hardened properties of cement-based sys-
tems (including those containing fly ash) with nanoclays, nanoC-
aCO
3
and nanoSiO
2
. This paper is a summary of the current
studies. The rheological properties of nanoclay-modified ce-
ment-based materials are investigated to further understand
the influence of nanoclays on fresh-state stiffening and form-
work pressure. The influence of the method of dispersion of
nanoCaCO
3
powder on early-age properties is evaluated. And
an in-depth study on the mechanisms underlying the influence
of nanoSiO
2
on the compressive strength gain of fly ash–cement
systems is discussed.
With proper processing techniques and fundamental under-
standing of the mechanisms underlying the effect of the nano-
particles, they can be used to enhance the fresh-state and
hardened properties of cement-based materials for various appli-
cations. Nanoclays can increase the green strength of self-consol-
idating concrete (SCC) for reduced formwork pressure and
slipform paving. NanoCaCO
3
and nanoSiO
2
can offset the nega-
tive effects of fly ash on early-age properties, namely slowed
rate of hydration and compressive strength gain, to facilitate
the development of a more environmentally friendly, high-
volume fly ash concrete.
0958-9465/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cemconcomp.2012.06.012
⇑
Corresponding author. Tel.: +1 847 491 7161(O).
E-mail address: s-kawashima@northwestern.edu (S. Kawashima).
Cement & Concrete Composites 36 (2013) 8–15
Contents lists available at SciVerse ScienceDirect
Cement & Concrete Composites
journal homepage: www.elsevier.com/locate/cemconcomp
2. Experimental methods and materials
2.1. Material properties
Tap water and ordinary Portland cement (OPC) were used in all
mixes, along with a type F fly ash in select mixes. The chemical
properties are given in Table 1. The fly ash particles are round with
a size range of 2–20 microns, as shown in the scanning electron
microscopy (SEM) image, Fig. 1a. The spherical morphology of fly
ash helps increase the flowability of cementitious materials,
including those with nanoparticles, which increase stiffness due
to their high specific surface area.
The three types of nanoparticles used in the current studies are
nanoclay, nanoCaCO
3
, and nanoSiO
2
. The nanoclay was a purified
magnesium aluminosilicate, or palygorskite, with a rod-like shape
(1.75
l
m in length, 3 nm in diameter). The nanoclay has been
chemically exfoliated to preserve its uniform shape and size while
removing all impurities (such as quartz and swelling clays). As re-
ceived, they are highly-agglomerated. To disperse the nanoclay,
they are blended with water in a household blender prior to mixing
with the other dry ingredients. The nanoCaCO
3
came in dry powder
form, with a particle size range of 15–40 nm. Similarly to the nano-
clay, they are agglomerated to the micron scale in the as-received
state. Two types of colloidal nanoSiO
2
(CNS) with an average parti-
cle size of 20 nm (CNS-20) and 10 nm (CNS-10) were used. The
transmission electron microscopy (TEM) images of CNS-10
(Fig. 1b) and CNS-20 (Fig. 1c) indicate that most of the nanoparti-
cles are well-dispersed, although some agglomeration may occur.
2.2. Testing procedures
Shear rheological tests were performed in a temperature-con-
trolled rheometer with a coaxial cylinder geometry set at room
temperature. The rate of hydration of pastes were measured in a
semi-adiabatic calorimeter, where a sample was placed in an insu-
lated drum and its temperature change was recorded for 24 h.
ASTM C191 was followed to measure initial and final setting time
of pastes with a manual Vicat needle apparatus [13]. ASTM C230
was followed to measure the slump flow of mortars using a flow
table [14].
ASTM C109 was followed to measure the compressive strength
of pastes or mortars with a 1000 kip (4448 kN) MTS hydraulic test-
ing machine [15] – the loading rate of the test was 0.008 mm/s. For
each mix at each age, three samples were tested and the average
value was taken to be the representative strength.
Thermogravimetric analysis (TGA, TGA/sDTA 851) was carried
out to measure the calcium hydroxide (CH) content of samples.
Samples of about 20 mg were heated at atmospheric pressure at
a rate of 15 °C/min. The weight loss between 440 °C and 510 °C
was considered to be the decomposition of CH. Before measuring,
samples were oven – dried at 105 °C for 4 h.
Hitachi S-4800 FE-SEM equipped with energy dispersive spec-
troscope (EDS) was used to analyze the morphology and elemental
compositions of the cement paste. A small fractured sample was
Table 1
Chemical properties of cement and fly ash.
Materials Type I cement Type F fly ash
SiO
2
20.2 46
Al
2
O
3
4.7 17.8
Fe
2
O
3
3.3 18.2
SO
3
3.3 2.59
CaO 62.9 8.4
MgO 2.7 0.95
Na
2
O – 0.59
K
2
O – 2.16
LOI 1.1 1.49
Total 98.2 98.2
Fig. 1. Morphology of (a) fly ash (SEM), (b) CNS-10 nm (TEM) and (c) CNS-20 nm
(TEM).
S. Kawashima et al. / Cement & Concrete Composites 36 (2013) 8–15
9
soaked in acetone to stop hydration and dried at 80 °C for 2 h.
Then the sample was coated with 20 nm of gold to make it conduc-
tive. The accelerating voltage and current were 15–20 kV and
10–20
l
A, respectively. The upper detector was used to collect
the secondary electrons. Hitachi H-8100 was used to analyze the
morphology of the nanoSiO
2
. Samples were diluted in alcohol be-
fore testing and the accelerating voltage and current were 200 kV
and 10
l
A, respectively.
3. Nanoclays
Previous work at ACBM has demonstrated that proper propor-
tioning of fly ash, superplasticizer, and a small addition of clays
(<1% by mass of binder) can significantly increase the green
strength of SCC mixes immediately after casting with little com-
promise to initial flowability [16–19]. Such properties can effec-
tively reduce SCC formwork pressure. Fig. 2 shows the formwork
pressure response of SCC mixes with and without a 0.33% nanoclay
addition by mass of binder – NC0.33 and NC0, respectively. These
are the results from a previous study [20], where a lab-scale form-
work pressure device was used to simulate casting by applying a
vertical pressure and measure the lateral pressure over time. It is
apparent that the clays significantly reduced lateral pressure. This
behavior has previously been tied to flocculation behavior, where
studies have shown that clays increase flocculation strength [21]
and floc size [22]. In the study discussed here, the contribution of
water adsorption by the clays on stiffening was examined through
a shear rheological approach [23]. If the water content is kept con-
stant, the incorporation of any nanoparticle will lead to an increase
in stiffness due to increase in water demand. The nanoclays have a
high water adsorption of 200% by mass, making it a possible gov-
erning factor.
3.1. Constant applied shear rate (CR) protocol
The water content of cement pastes with and without a 1%
nanoclay addition by mass of cement were adjusted for clay water
adsorption (assuming 200% water adsorption by mass). The paste
mix compositions are given in Table 2. Each mix was subjected
to a constant applied shear rate of 300 s
1
for 60 min, during which
the tangent viscosity was calculated from the measured shear
stress as follows:
l
¼
s
_
c
ð1Þ
where
l
is viscosity (Pa s), T is shear stress (Pa) and
_
c is applied
shear rate (s
1
).
The results are shown in Fig. 3. Fig. 3a shows the viscosity evo-
lution up to 5 min, during which each mix reached equilibrium
(steady-state). At this point the flocs cannot be broken down any
further under the given shear condition. If the loss of free water
by clay water adsorption is a governing stiffening mechanism,
mixes NC0 and NC1H and mixes NC1 and NC0H should exhibit
similar viscosity evolutions. However, it is apparent that the vis-
cosities of these mixes did not coincide and adjusting water con-
tent did not offset the stiffening effect of the clays. Also, both
mixes with clays exhibited a significant decrease in viscosity upon
the introduction of shear due to deflocculation. This shows that the
clays had an immediate stiffening effect through flocculation,
resulting in a highly thixotropic material.
After equilibrium, each paste goes onto experience a similar in-
crease in viscosity, as shown in Fig. 3b. Table 3 shows that the
change in viscosity of all the mixes from equilibrium to final
(60 min) were very close. This indicates that although the clays
Fig. 2. Formwork pressure response of SCC with and without a 0.33% nanoclay
addition [20].
Table 2
Mix composition of pastes for CR protocol, where NC0 is the control for NC1H and
NC1 is the control for NC0H.
Mix Cement (g) Water (g) Clay (g)
NC0 500 215 0
NC1 500 217.6 5
NC1H 500 227.6 5
NC0H 500 205 0
Fig. 3. Viscosity evolution of pastes (a) before and (b) after equilibrium [23].
10 S. Kawashima et al. / Cement & Concrete Composites 36 (2013) 8–15
have an immediate stiffening effect, they have little or no influence
over time under a given constant shear condition. This will be tied
into previously obtained formwork pressure results and discussed
in the following section.
3.2. Formwork pressure
In the formwork pressure test introduced earlier, the vertical
pressure was applied as a step-wise function, Fig. 2. As shown in
Fig. 4, this allows for an instantaneous response, b, and a delayed
response,
a
, to be measured. This two function model was intro-
duced by Kwon et al. and Kim et al., the details of which can be
found elsewhere [24,25]. While the concrete is plastic,
a
and b
can be represented in terms of delayed coefficient, a, and instanta-
neous coefficient, b, both of which are material parameters:
a
ðt; t
0
Þ¼1 a
2
t
0
ðt t
0
Þð2Þ
bðt
0
Þ¼1 bt
0
ð3Þ
t is current time and t
0
is time of loading. The delayed and instanta-
neous responses of NC0 and NC0.33 over a 2 h casting period are
plotted in Fig. 5. The clays have little influence over the delayed re-
sponse – the rate at which the lateral pressure decreases as the ap-
plied vertical pressure is held constant is the same for both mixes,
Fig. 5a. However, the clays significantly reduce the instantaneous
response – the lateral pressure is significantly lower upon each
incremental increase in pressure, Fig. 5b. These results are similar
to those of the CR test, where the clays exhibit an immediate effect
upon a change in applied stress but little influence over time under
a constant condition.
Kim et al. used the two-function model to characterize a variety
of different SCC mixes [24]. They found that variation in coefficient
a had a negligible effect on the maximum formwork pressure,
r
max
,
while variation in b had a dominant one during the first few hours
of casting. For the design of formwork,
r
max
is a critical value. It fol-
lows that since the nanoclays significantly affect b, they have the
capacity to greatly reduce
r
max
.
4. Calcium carbonate nanoparticles
Although limestone powder is typically considered to be inert,
conventionally used as a filler to improve rheological properties,
recent studies have found that they can accelerate rate of
hydration when introduced as nanoparticles [4,5]. However, few
in-depth studies have been done thus far. This study focuses on
processing, where dry nanoCaCO
3
powder is dispersed through
ultrasonication. This method has been successfully implemented
to disperse carbon nanotubes and significantly improve the
mechanical properties of cement composites [1]. Through effective
dispersion, the motivation is to enhance the effect of the nanopar-
ticles and to decrease the addition level necessary.
In the present study, the effect of sonicated versus blended
nanoCaCO
3
on rate of hydration, setting time, and early-age com-
pressive strength gain were compared in cement and fly ash–ce-
ment pastes. NanoCaCO
3
was sonicated for 30 min in an aqueous
solution at 15% concentration. And 0.06% polycarboxylate superp-
lasticizer (by weight of water) was added to the suspension to help
with stabilization. Suspensions prepared by blending were pre-
pared in a household blender for 3 min.
4.1. Rate of hydration
The rate of hydration of OPC pastes was measured through
semi-adiabatic calorimetry. All samples had a water-to-cement
(w/c) ratio of 0.43, with or without a 5% nanoCaCO
3
addition by
mass of cement. Pastes with nanoCaCO
3
were prepared with a sus-
pension that was sonicated or blended. The results are shown
in Fig. 6. It is apparent that for both samples with a nanoCaCO
3
addition, there is an acceleration in rate of hydration. However,
Table 3
Change in viscosity from equilibrium to final (60 min).
Mix
D
Viscosity (Pa s)
NC0 0.192
NC1 0.175
NC1H 0.18
NC0H 0.176
Fig. 4. Two function model for formwork pressure of concrete [24].
Fig. 5. (a) Delayed and (b) instantaneous formwork pressure response of SCC with
and without a 0.33% nanoclay addition.
S. Kawashima et al. / Cement & Concrete Composites 36 (2013) 8–15
11
the effect is more pronounced for the sonicated sample, where the
peak is higher and occurs earlier.
4.2. Setting
The setting time of pastes with a 50% replacement of cement
with fly ash were measured and compared against a plain OPC paste.
All samples had a water-to-binder (w/b) ratio of 0.4, with or without
a 5% nanoCaCO
3
addition. The results are shown in Table 4. In both
cases, blended and sonicated, the addition of 5% nanoCaCO
3
helped
accelerate setting. Further, the sonicated nanoCaCO
3
completely
offset the delay caused by the 50% fly ash replacement, where it
exhibited the same setting times as the plain OPC paste sample.
4.3. Compressive strength
The compressive strength gain (1, 3, and 7 d) of 50 mm cube
samples were compared for an OPC paste and 30% fly ash–cement
pastes with and without a 5% nanoCaCO
3
addition. All pastes had a
w/b = 0.43. The results are shown in Fig. 7. At 3 and 7 d, the soni-
cated sample showed a greater improvement than the blended
sample. However, neither reached the strength of the OPC sample.
Work on modifying the sonication protocol to improve the stability
of nanoCaCO
3
suspensions is ongoing.
5. NanoSiO
2
Although high volume fly ash introduces the advantage of
increasing the workability of cementitious materials, its slow
early-age strength gain is a major drawback and has hindered its
application. With the incorporation of nanoSiO
2
, the strength gain
of cement-based materials can be improved due to its hydration
seeding effect and high pozzolanic activity [8,10,26,27], which re-
sult in a higher amount of C–S–H gel and more densified bulk
structure [28]. However, the shortcoming of nanoSiO
2
-modified
cement-based material is that the nanoSiO
2
adversely affects its
workability due to the high specific surface area [29]. In consider-
ing the characteristics of the two constituents (fly ash and nano-
SiO
2
), the benefit of each material can help counteract the
shortcoming of the other: reactive nanoSiO
2
can improve the
early-age mechanical properties while fly ash can improve
flowability.
When considering the enhancing effect of nanoSiO
2
on strength
gain, most studies thus far have focused on the early-age strength
gain [9,26]. However, some contradicting results have been re-
ported for the later-age strength development [7]. The lower
strength of nanoSiO
2
-modified mixes compared to the control at
later ages is typically attributed to the dispersion of the nanoparti-
cle and the production technique of the nanoSiO
2
[7,30]. Consider-
ing that high amounts of Ca(OH)
2
can be consumed by nanoSiO
2
in
the early age, resulting in a lower Ca(OH)
2
content for later-age
pozzolanic reaction of fly ash, the influence of nanoSiO
2
on the la-
ter-age strength of high volume fly ash cementitious materials may
be significant. Thus, their effect on the later-age mechanical prop-
erties needs further investigation. The results of an in-depth study
on the mechanisms underlying the effect of colloidal nanoSiO
2
on
the fresh and hardened properties of fly ash–cement system are
discussed here.
5.1. Fresh properties
The effect of CNS dosage on the slump flow of fly ash–cement
mortars is shown in Fig. 8. Mortars were prepared with a sand-
Fig. 6. Semi-adiabatic calorimetry results for cement pastes with a 5% nanoCaCO
3
addition, prepared by sonication or blending, and plain cement paste.
Table 4
Setting time of 50% fly ash OPC pastes with and without a 5% nanoCaCO
3
addition,
prepared by blending or sonication, compared to plain OPC paste.
Initial setting/h Final setting/h
50FA 6.2 7.4
50FA 5CaCO
3
(Blended) 4.8 6.1
50FA 5CaCO
3
(Sonicated) 4.4 5.4
Plain OPC 4.4 5.4
Fig. 7. Compressive strength gain of 30% fly ash OPC pastes with and without a 5%
nanoCaCO
3
addition, prepared by blending or sonication, compared to plain OPC
paste.
Fig. 8. Influence of CNS on the slump flow of fly ash–cement mortar (w/b = 0.5,
cement to sand ratio = 1:3).
12 S. Kawashima et al. / Cement & Concrete Composites 36 (2013) 8–15
![Fig. 2. Formwork pressure response of SCC with and without a 0.33% nanoclay addition [20].](/figures/fig-2-formwork-pressure-response-of-scc-with-and-without-a-0-ynem20z8.webp)