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

net/publication/299494707

Influence of 2D rGO nanosheets on the properties of OPC paste

Article  in  Cement and Concrete Composites · March 2016

DOI: 10.1016/j.cemconcomp.2016.03.005

CITATIONS READS

19 544

5 authors, including:

Murugan Muthu Manu Santhanam

Nanyang Technological University Indian Institute of Technology Madras
14 PUBLICATIONS   35 CITATIONS    178 PUBLICATIONS   1,857 CITATIONS   

SEE PROFILE SEE PROFILE

Soujit Sen Gupta Pradeep Thalappil

Indian Institute of Technology Madras Indian Institute of Technology Madras
31 PUBLICATIONS   528 CITATIONS    398 PUBLICATIONS   14,768 CITATIONS   

SEE PROFILE SEE PROFILE

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

Protective treatments for heritage masonry structures View project

Effect of acid attack on concrete View project

All content following this page was uploaded by Murugan Muthu on 04 September 2018.

The user has requested enhancement of the downloaded file.

 Cement and Concrete Composites 70 (2016) 48e59

Contents lists available at ScienceDirect

Cement and Concrete Composites

journal homepage: www.elsevier.com/locate/cemconcomp

Influence of 2D rGO nanosheets on the properties of OPC paste

M. Murugan a, Manu Santhanam a, *, Soujit Sen Gupta b, T. Pradeep b, Surendra P. Shah c
a
Department of Civil Engineering, Indian Institute of Technology Madras, Chennai 600036, India
b
Department of Chemistry, Indian Institute of Technology Madras, Chennai 600036, India
c
Center for Advanced Cement Based Materials, Northwestern University, Illinois 60208, USA

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

Article history: In this experimental study, the effects of 2D reduced graphene oxide (rGO) sheets on the properties of
Received 9 April 2015 Portland cement paste in comparison to popularly reviewed nanomaterials like aluminium oxide
Received in revised form nanopowder (n-Al2O3) and colloidal silicon dioxide nanoparticles (n-SiO2) were investigated. The
12 February 2016
addition of 0.02% rGO sheets by weight of cement increased the 7 and 28 days flexural strength up to 70%
Accepted 26 March 2016
Available online 29 March 2016
and 23% respectively when compared to control paste. Moreover, its incorporation substantially
decreased the sizes of pores/voids in the paste, even compared to the other nanomaterials, as charac-
terized by Mercury Intrusion Porosimetry (MIP) and 3D X-ray Computed Tomography (CT) aided with
Keywords:
Reduced graphene oxide
image analysis technique. The assessment of Portlandite content by Thermo-gravimetric Analysis did not
Nanocomposite indicate major differences between the pastes, with the exception of the paste incorporating nano-silica.
Pore structure Microstructural analysis by Fourier Transform Infrared Spectroscopy, X-ray diffraction and Scanning
Mechanical properties Electron Microscopy did not reveal any major differences between the control paste and the pastes
Tomography incorporating nanomaterials. The overall results suggest that the performance of rGO was better in
comparison to other two nanomaterials, despite the significantly lower amounts that were used in the
paste.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction surface area, the shape and structure of nanomaterials also play an
important role in microstructural alterations of the cementitious
The diffusion of deleterious ions through the pore network of compounds [8].
cementitious materials can be arrested by incorporating nano-sized Graphene, a 2D planar sheet of single atom thickness having sp2
materials, which have a high surface area to volume ratio. Minor bonded carbon atoms arranged in a honeycomb lattice pattern,
additions of nanomaterials such as carbon nanotubes, nano silica, gives excellent in-plane mechanical (modulus of elasticity
nano alumina and graphene nanosheets, which are effectively 1100 GPa), thermal (conductivity 5300 W/mK), optical transparent,
dispersed in the cement matrix, are beneficial in enriching the and electrical (conductivity 2000 S/cm) properties [9].
mechanical and durability properties [1]. The nanomaterials are Lv et al. [10] have attributed the aforementioned behaviour of
suggested to be super-sorbents that physically adsorb the free morphological changes to shape effects by the dispersion of gra-
water to a large extent. Due to such effects, the nanomaterials are phene oxide (GO) into the cement mortar. It was reported that
perceived to act as nucleation sites for the growth of hydration flower-like planar clusters at low GO dosage and polyhedral and
crystals at early ages, which ultimately results in reducing the lamellar form of hydration crystals were densely formed due to the
fraction of unhydrated cement [2,3]. Further, the issue of reduced effect of high GO dosage into OPC paste. The study claims that such
workability that is caused because of the use of the high surface flower clusters were easily produced in holes and cracks of the
area nanomaterials can be overcome by the use of surfactants, cementitious composites. By performing thermal analysis, Gong
which also help in reducing the agglomeration effect of nano- et al. [11] have reported that 0.03% GO by weight addition into
materials due to van der Waals forces [4e7]. In addition to high cement paste increases the generation of Portlandite and non-
evaporable water up to 6% and 9% respectively. Further, they also
reported that the compressive, tensile and flexural strengths were
* Corresponding author. increased up to 38%, 79% and 60% respectively compared to control
E-mail address: [email protected] (M. Santhanam).

http://dx.doi.org/10.1016/j.cemconcomp.2016.03.005
0958-9465/© 2016 Elsevier Ltd. All rights reserved.
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 49

specimens. 2.2. Sample preparation

Bulk synthesis of graphene is difficult to achieve because of the
significant van der Waals forces in these sheets of high surface Four different cement paste mixes with a water to cement (w/c)
area (2630 m2/g) that cause irreversible agglomeration or even ratio of 0.32 and dispersion agent (PCE) of 0.05% by weight of
restacking to form graphite. Kuila et al. [12] have suggested that cement were prepared. Table 2 specifies the type and amount of
such an effect was overcome with the help of using Chemical nanomaterials used in the different mixes.
Oxidation/Modified Hummers' method to synthesize graphene For better dispersion, pre-mixing of the nanomaterials along
oxide. This method proves effective in synthesizing chemically with deionized water and PCE were carried out using a magnetic
converted graphene sheets which are able to form stable colloids stirrer for 15 min, followed by ultrasonication in an ultrasonic bath
through electrostatic stabilization. Without the use of surfactants, for next 30 min Fig. 2 shows the well-dispersed rGO solution (black
effective dispersion is achieved through hydrazine reduction of coloured liquid) used in this study.
graphene oxide under controlled conditions [13]. The resultant For paste preparation, the ingredients were mixed in a high
material is termed as ‘Reduced’ Graphene Oxide (rGO). speed Hobart mixer for 12 min. After mixing, the mini slump test
In this study, the influence of reduced graphene oxide (rGO), dry was conducted to determine the flowability; the remaining paste
aluminium oxide nanopowder (n-Al2O3) and colloidal silicon di- was used for casting specimens of size 25  25  25 mm and
oxide nanoparticles (n-SiO2) on the properties of cement pastes 20  20  160 mm for determining the compressive and flexural
such as workability, compressive and flexural strength, mass strength respectively. Consolidation was performed using a vibra-
porosity and water sorptivity index were determined after curing tion table. The specimens were demolded after 24 h and then cured
for 7 and 28 days. Then the effect of nanomaterials on the pore in saturated limewater solution at room temperature (25 ± 2
C) for
refinement of OPC paste was investigated using Mercury Intrusion 7 and 28 days.
Porosimetry (MIP) and 3D X-ray Computed Tomography (CT)
studies. Further, the alterations in microstructure of cement paste 2.3. Testing methods
due to the addition of nanomaterials were characterized using
Fourier Transform Infrared spectroscopy (FTIR), X-ray Diffraction A mini slump cone as shown in Fig. 3a was used for conducting
technique (XRD), Thermo-gravimetric Analysis (TGA) and Scanning the mini slump test in order to determine the workability of cement
Electron Microscopy (SEM) techniques. paste. The procedure specified by Collins et al. [18] was followed in
The motivation behind choosing the popularly reviewed nano- this study. To determine compressive and flexural strength, the
materials such as n-Al2O3 and n-SiO2 for the comparative evalua- specimens were tested at a loading rate of 143 N/sec and 50 N/sec
tion against the rGO cement composites was due to their effective respectively. Both tests were performed using Controls C9842
dispersion into the cementitious matrix without the use of any testing machine on different frames having the capacities of 250 kN
surface modification agents [14,15]. and 15 kN respectively.
Water sorptivity test was performed on the specimens of size
2. Experimental 25  25  12.5 mm (sliced from the 25 mm cubes) after 7 and 28
days. After curing, the specimens were stored in acetone for the
2.1. Materials next two days to arrest the hydration followed by air drying for
two days. Further, the dried specimens were kept in a shallow tray
A top-down approach was followed to extract the rGO from of size not less than 150  75 mm with roller arrangement as
natural graphite by redox reaction. Three steps such as pre- shown in Fig. 3b. Then the tray was filled with limewater solution
reduction, sulfonation and post-reduction were followed to syn- (5 g per 1 L of distilled water). The final water level was main-
thesize rGO. Sodium borohydride was used for the pre-reduction tained slightly above the bottom edge of the specimen. The mass
of rGO at 80
C for 1 h in order to remove the major amount of of the specimen was measured at specified intervals such as 0, 3, 5,
oxygen functionalities from graphene oxide. Then the water sol- 7, 9, 12, 16, 20 and 25 min. Before weighing, the adsorbed water at
ubility (dispersion in water) was improved by adding sulfanilic the specimen surface was removed with a moist cloth. Finally, the
acid in an ice bath for 2 h. In this stage, controllable structural vacuum saturated mass of the specimens was measured.
defects were allowed to avoid alteration in the properties of gra- The influence of rGO, n-Al2O3 and n-SiO2 on the microstructure
phene. In the post reduction process, hydrazine solution was of cement paste cured for 28 days was characterized using XRD,
mixed with as-synthesized graphene oxide and stirred at 100
C FTIR, TGA and SEM. The paste samples were ground using mortar
for 24 h using rapid synthesizer. A majority of oxygenated groups and pestle and sieved through 75 mm sieve for XRD, FTIR and TGA
are removed other than acid functionalities during such a reaction. testing. Bruker XRD machine was used to determine the crystalline
Before its use, the synthesized graphene oxide of sheet sizes phases formed in the nanocomposites and the control paste. IR
ranging from 5 nm to 1500 nm was kept under observation for 6 h absorption spectra were collected using PerkinElmer FTIR instru-
to examine the stability of colloids [16,17]. The transformed colour ment for all paste samples. Netzsch TGA testing machine was used
of GO to rGO during its synthesis is shown in Fig. 1a. The TEM for determining the amount of Portlandite in pastes cured for 28
image of synthesized rGO as presented in Fig. 1b, indicating a days. For SEM investigation, the fractured surface of paste sample
wrinkled surface texture (image obtained from JEOL 3010 Trans- was tested using FEI Quanta 400 FEG-SEM device. To enhance its
mission electron microscope). 53 grade ordinary Portland cement conductivity, a 3 nm thick Gold/Palladium layer was coated over the
(OPC) conforming to IS 12269-1987 and polycarboxylic ether surface of the specimen.
(PCE) based superplasticizer as dispersion agent were used in this The effect of nanomaterials on the pore structure of OPC paste
study. Table 1 shows the chemical composition of the cement. was characterized using MIP and 3D X-ray CT. Firstly, the distri-
Dry aluminium oxide nanopowder (n-Al2O3) of mean size less bution of gel pores (<10 nm) and capillary pores (10 nme10 mm) in
than 50 nm and silicon dioxide nanoparticles of average size 12 nm a paste sample of size 7  7  7 mm was determined using Thermo
dispersed in deionized water (n-SiO2) having surface area of 40 m2/ Scientific Pascal 140/440 series MIP instrument. Secondly, the dis-
g and 225 m2/g respectively were obtained from Sigma-Aldrich. tribution of pores/voids (>100 mm) in a paste sample of size
Energy dispersive X-ray analysis was performed on the nano- 10  10  10 mm was determined using tomographic images ac-
materials, and the results are shown in Fig. 1c. quired through GE phoenix vjtomejx s 3D X-ray CT testing machine.
50 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59

Fig. 1. a. GO transformed into rGO, b. TEM image of the synthesized rGO and c. EDS spectrum of the nanomaterials used in the study.

Table 1 Additionally, multiple slices from the acquired tomographic images

Chemical composition of 53 grade OPC (conforming to IS 12269-1987) used in this were processed and analysed using the software named Image Pro
study.
Premier 9.1 to determine the pore area fraction of the slices which
Name of the compound CaO SiO2 Al2O3 Fe2O3 SO3 MgO Cl is better explained in subsection 3.7.
% by weight 62.16 20.95 5.14 3.06 3.08 1.33 0.009

3. Results and discussion

Table 2
3.1. Mini slump spread, compressive and flexural strength
Amounts of the different nanomaterials added into OPC paste.

Type of mix Control rGO n-Al2O3 n-SiO2 The relative increase/decrease (expressed as %) of the experi-
Nanomaterials (% by wt.) e 0.02 0.20 4.0 mental results with different nanocomposites compared with the
control paste is presented in Table 3.
The addition of rGO, n-Al2O3 and n-SiO2 in OPC paste was seen
to reduce the workability by 13%, 11% and 24% respectively

Fig. 2. Well dispersed rGO solution (black coloured liquid) used in this study.
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 51

Fig. 3. a. Mini slump cone test apparatus and b. Water sorptivity test.

Table 3
Relative change in performance in pastes with nanomaterials (expressed as % of control paste).

No Paste mixes % By wt. of cement Percent increment or decrement

Compressive strength Flexural strength Slump spread Mass porosity Sorptivity Index

7/28 days 7/28 days 7/28 days 7/28 days

1 Control e 100/100 100/100 100 100/100 100/100

2 rGO 0.02 120/122 170/123 87 80/96 82/104
3 n-Al2O3 0.20 132/123 204/123 89 89/93 70/151
4 n-SiO2 4.0 145/141 218/104 76 76/81 89/170

compared to the control mix. As mentioned earlier, these nano- 3.3. XRD pattern and IR spectra
materials are super sorbents having high surface area, and lead to
the absorption of maximum amount of available free water in the The X-ray diffraction pattern and the Infra-red spectra shown in
early ages [19e21]. Due to such an effect, the early strength Fig. 4a and Fig. 4b were obtained from the different nano-
enhancement achieved by the pastes with different nano- composites and the control paste cured in limewater solution for 28
composites (rGO, n-Al2O3 and n-SiO2) was found comparatively days.
higher when compared to the 28 days strength enhancement re- No major differences could be seen in the diffraction patterns of
sults as mentioned in Table 3. In particular, the 2-dimensional the different pastes, indicating similar mineralogical compositions.
(sheet-like) rGO even at small concentrations (0.02% by weight) In case of IR spectra, the peaks attributed to OeH bend ~1645 cm1,
proved efficient in comparison with the 3-dimensional spherical/ SeO stretch ~1118 cm1, SiO4 ~ 950 cm1 vibration representing the
angular nanomaterials such as n-Al2O3, and n-SiO2. The good per- presence of free water, ettringite and silicates were observed in all
formance of rGO can be attributed to its quality of better dispersion pastes cured for 28 days. Secondly, the peaks attributed to CeO
in a polar solvent like water, the shape and size factors, and the stretch ~1414 cm1 and CeO bend ~874 cm1 vibration repre-
presence of functional groups such as carboxyl and hydroxyl on the senting the presence of CO2 3 were also obtained in the pastes with
structure of the graphene sheets [9,10]. nanocomposites (rGO, n-Al2O3, and n-SiO2) and the control paste,
which is indicative of carbonation.

3.2. Mass porosity and water sorptivity index

3.4. Thermogravimetric analysis
The relative increase/decrease (expressed as %) of the mass
porosity and water sorptivity index results with different nano- The TGA curves of the pastes with different nanocomposites and
composites and the control paste are presented in Table 3. The the control paste cured for 28 days are shown in Fig. 5. Table 4
results indicate that the nanomaterials (rGO, n-Al2O3, and n-SiO2) presents the gravimetric changes that were obtained from the
in OPC paste at initial ages led to a decrease in the water sorptivity pastes.
(and porosity) compared to the control mix. Pore refinement due to The peaks P1 and P2 specified in the TGA plot indicate the
the presence of nanomaterials and increased production of hy- gravimetric changes due to decomposition of Ca(OH)2 and CaCO3
dration compounds such as CeSeH, CeH, AFt and AFm at early ages respectively. The presence of the CaCO3 can be attributed to the
are the possible reasons for this effect. However, at later ages, the original cement (~4% in the cement used), as well as to the con-
high surface area nanomaterials cause an increase in the water version from hydrated lime due to carbonation. The total Ca(OH)2 is
sorptivity (porosity in all the pastes is almost the same) possibly therefore calculated by adding the Ca(OH)2 after 28 days from the
owing to the super-absorbing behaviour. TGA study along with the approximate quantity of Ca(OH)2 that
52 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59

Fig. 4. a. XRD pattern and b. IR spectra of the different pastes cured for 28 days.

decreased value when compared to the other pastes after 28 days.

Such an effect is well supported with the help of compressive
strength results of n-SiO2 paste.

3.5. Scanning electron microscopy images

High resolution SEM images were obtained for the pastes with
the different nanocomposites to observe the morphological
changes at high magnification due to the incorporation of nano-
materials (rGO, n-Al2O3 and n-SiO2). Fig. 6a-b shows the
morphology of rGO cement composite cured for 28 days. It was
found that the production of thin non-uniform platelets and
entangled network of rod like crystals were observed at various
locations in the rGO sample after 28 days.
Fig. 7a-b shows the morphology of n-Al2O3 cement composite
cured for 28 days. The images show the presence of rod like crystals
within the CeSeH matrix and the production of cloud-like CeSeH
which is non-uniform in texture due to the addition of crystalline
alumina nanomaterials into the cementitious matrix.
Fig. 8a-b shows the morphology of n-SiO2 cement composite
cured for 28 days. Here, the pozzolanic conversion due to the
addition of reactive silica nanoparticles caused the formation of a
Fig. 5. TGA curves of the different pastes cured for 28 days. homogeneous textured CeSeH matrix at various places in the n-

Table 4
Gravimetric changes (in percentage) obtained from the TGA study.

Mix designation Control rGO n-Al2O3 n-SiO2

P1 e % Dehydroxylation of Ca(OH)2 ~430e480 C 6.04 7.15 8.06 4.11
P2 e % Decarbonation of CaCO3 ~600e780
C 13.98 11.77 11.57 10.32
Total Ca(OH)2 in % by mass 14.14 13.62 14.38 9.50

Note: P1 corresponds to the removal of water from Ca(OH)2, so the mass change in TGA is adjusted to the Ca(OH)2 content by multiplying by the ratio of molecular masses of
Ca(OH)2 and H2O. Similarly, as P2 corresponds to the removal of CO2 from CaCO3, the mass change in TGA is adjusted to the CaCO3 content by multiplying by the ratio of
molecular masses of CaCO3 and CO2. Further, the total Ca(OH)2 is then calculated as the sum of P1 and stoichiometric quantity of CH from P2 (assuming the reaction as:
Ca(OH)2 þ CO2 / CaCO3 þ H2O).

transforms to calcite by carbonation. SiO2 sample after 28 days.

From the results in Table 4, it can be seen that the potential CeH
is similar in rGO and n-Al2O3 pastes when compared to control
paste. In the case of n-SiO2, the rapid conversion of CeH to CeSeH 3.6. Porosimetry results
products due to pozzolanic reactivity was the reason behind the
Fig. 9 shows the MIP result of the different nanocomposites and
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 53

Fig. 6. Secondary electron images of the pastes with rGO nanocomposite after 28 days.

Fig. 7. Secondary electron images of the paste with n-Al2O3 nanocomposite after 28 days.

the control paste cured for 28 days e in each plot, two curves are of CeSeH gel. Gong et al. [11] have observed similar MIP results in
presented e the cumulative intrusion curve and the differential OPC paste containing 0.03% GO addition by weight cured for 28
intrusion curve. Table 5 presents the evaluation of the pore struc- days. In case of n-Al2O3 and n-SiO2, the gel and capillary pores
ture features of the pastes such as amount of gel and capillary were found increased to a significant extent when compared to
pores, threshold pore diameter and most likely diameter obtained control paste due to agglomeration/bundling of such nano-
from MIP data. While the ranges for gel and capillary pores were materials within the cementitious matrix. However, as reported
classified based on Aligizaki [22], the threshold diameter was earlier in the paper, the mechanical performance of such mixes
selected as the pore size beyond which there was a sudden increase was found to be increased compared to the control paste. This
in the intrusion, and the most likely diameter was chosen as the might be due to better particle packing between the nano-
peak from the differential intrusion curve. additives and the cementitious compounds.
It can be seen from the results that the well-dispersed rGO in
OPC paste was found to decrease the capillary pores up to 32.1%
compared to the control paste due to factors such as pore refine- 3.7. 3D X-ray CT images
ment and increased production of cement hydrates as observed
from the TGA results. However, the rGO addition was found to The CT scans were performed at the Centre for Non-destructive
increase the gel pores in cementitious matrix up to 36.5% when Evaluation (CNDE), Indian Institute of Technology Madras. Different
compared to control paste possibly due to the increased formation pastes of sample size 10  10  10 mm were scanned using an X-ray
source generated at a voltage and current of 120 kV and 70 mA
54 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59

Fig. 8. Secondary electron images of the paste with n-SiO2 nanocomposite after 28 days.

Fig. 9. Porosimetry results of the different pastes cured for 28 days.

respectively. Additionally, the scanning rate was maintained 20 mm approximately during reconstruction of images. Caesium
around 4 ms per image, in order to have a voxel resolution up to Iodide based flat panel detector was equipped inside the CT testing
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 55

Table 5
Pore structure features of the pastes cured for 28 days obtained from MIP data.

Paste mixes Gel pores <10 nm in mm3/g Capillary pores (10 nme10 mm) in mm3/g Threshold diameter in nm Most likely diameter in nm

Control 10.69 48.58 26.8 14.7

rGO 14.60 33.01 20.7 12.9
n-Al2O3 26.97 60.58 39.1 27.1
n-SiO2 23.42 57.76 28.6 20.4

Fig. 10. a. GLV adjusted, b. Reconstructed and c. Transparent CT image of the paste sample.

Fig. 11. Colour-coded 3D CT images of the cement paste obtained before and after Porosity analysis using the software VGStudio MAX. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
56 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59

Fig. 12. Histogram representing the distribution of voids present in the different pastes determined using the Porosity/Inclusion Module available in the software VGStudio MAX.

machine to convert the X-rays into light. The output data obtained vgstudio-max/porosityinclusion-analysis). This provides a range
from the X-ray CT run after the end of scanning period was a 2D of algorithms allowing porosity/inclusions to be located within
radiograph of the paste specimen. The radiographs were extracted the object and yields quantitative information about these defects
with the help of a data acquisition software named Phoenix datosjx (voids/pores). The module is a bundle of different algorithms and
CT. Later, the 2D data were reconstructed and processed with the modes which allows to specially investigate on pores or imper-
software named VGStudio MAX version 2.2 which helps to visualise fections with high contrast (air voids) or low contrast (loose
the 3D imaging of the samples. material). It also helps to examine the material with open defects
The reconstructed CT image which is of ‘n’ number of voxels is a that have cut through the surface or enclosed defects. Different
spatial distribution of the linear attenuation coefficient, which is parameters such as pore to pore distance, sphericity, and user-
expressed in terms of Grey Level Value (GLV), with brighter and defined surface can be estimated using this analysis module. Af-
darker regions corresponding to higher and lower value of atten- ter the analysis run, the detailed information about each indi-
uation coefficient. To visualise the presence of pores/voids inside vidual defect (pores/inclusion) such as its volume, projected size
the cement paste, the GLV of the CT image was slightly adjusted as (Feret's diameter) and position in the material are obtained. The
shown in Fig. 10a. The reconstructed and transparent version of CT pores/inclusions present in the material are colour-coded in all
image of the paste sample subjected to GLV adjustment which 2D and 3D views and are also acquired along with the analysis
helps to visualize the distribution of pores/voids inside the sample data.
is shown in Fig. 10b-c. In this study, the analysis module was tested on the CT image
The image reconstruction software VGStudio MAX is supplied of the paste specimens to determine the volume and diameter of
with an optional add-on module named Porosity/Inclusion the pores/voids present in the system. The only input parameters
Analysis (http://www.volumegraphics.com/en/products/ that are required to run the analysis are lower and upper bound of
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 57

Fig. 13. CT slice of the paste sample before and after analysis performed using Image analysis technique (Image Pro Premier 9.1).

the pore/void volume which are expected to be present in the 4. Conclusions

material. Hence, the minimum voxel resolution of the CT image
was up to 20 mm, and the safe void volume (minimum threshold The mechanical, durability and microstructure performance of
value) was fixed to have a voxel resolution of 100 mm to avoid cementitious paste incorporating nanomaterials such as rGO, n-
misleading results. After running the analysis using the module, Al2O3 and n-SiO2 at concentrations of 0.02%, 0.20% and 4.0% by
the output data consists of parameters such as void volume, void weight respectively was reported in this paper. The key findings are
diameter (Feret diameter), voxel counts, and coordinates of the summarized as follows:
each defect present in the system. In addition, the colour-coded
3D views of the analysed CT image were also collected from the 1. The addition of nanomaterials (rGO, n-Al2O3 and n-SiO2) re-
software. Figs. 11 and 12 present the different colour-coded CT duces the workability of OPC paste.
images of the paste sample analysed with the Porosity/Inclusion 2. The use of nanomaterials increases the compressive and flexural
module and histogram of the distribution of pores/voids present strength of cementitious paste at early and later ages. In
in the different pastes cured for 28 days. particular, the mechanical performance of 0.02% rGO nano-
The histogram results suggest that the addition of nano- sheets was found efficient than the spherical/angular shaped
materials (rGO, n-Al2O3 and n-SiO2) in OPC paste was found to nanomaterials (n-Al2O3 and n-SiO2).
refine the pores/voids to a significant extent when compared to 3. The nanomaterials decrease the mass porosity and water sorp-
control paste. In particular, the incorporation of rGO was found tivity of OPC at 7 days compared to the control paste. However,
efficient than the other nanomaterials (n-Al2O3 and n-SiO2) in there is a marginal increase at 28 days, possibly owing to the
refining the voids distribution in cement paste as clearly evident sorbing nature of these materials.
from the histogram results shown in Fig. 12. 4. Morphological alterations in the microstructure of OPC paste
The influence of nanomaterials (rGO, n-Al2O3 and n-SiO2) on due to the addition of nanomaterials (rGO, n-Al2O3 and n-SiO2)
refining the pores/voids was further established by determining the were observed from the SEM images. Thermal analysis, IR
pore area fraction of CT slices of the paste sample. Such determi- spectra and XRD pattern indicate that the pastes with nano-
nation was conducted with the help of image analysis technique composites (rGO, n-Al2O3 and n-SiO2) after 28 days were iden-
performed over the CT slices. Image Pro Premier version 9.1 was the tical compositionally to OPC paste, with the exception of the
software used for this analysis. The reconstructed CT image was lower CeH content in n-SiO2 by thermal analysis because of the
sectioned into multiple slices (around 200). Then the pore area pozzolanic reaction.
fraction of the slices was acquired with the use of area measure- 5. The porosity analysis (surface and volume) by X-ray tomography
ment and batch processing tools present in the image analysis suggests that the incorporation of nanomaterials was found
software. Fig. 13 shows an unanalysed and analysed image of the useful in refining the pores/voids (100 mme3 mm) to a signifi-
paste sample obtained from the software Image Pro Premier 9.1. cant extent than the control paste cured for 28 days. The
The output data obtained from the software were converted into a improved performance of rGO paste was also seen in the MIP
histogram representing the distribution of pores/voids in the studies. However, the gel (<10 nm) and capillary pores
different pastes as shown in Fig. 14. (10 nme10 mm) were found increased due to agglomeration of
The results reconfirm the observations of the porosity analysis nanomaterials such as n-Al2O3 and n-SiO2 in OPC paste, as
presented earlier, that the addition of nanomaterials (rGO, n-Al2O3 observed from MIP results. The overall results clearly show that
and n-SiO2) in OPC paste refines the pores/voids to a significant the performance of 2D rGO was more convincing than n-Al2O3
extent compared to control paste. It is evident that the porosity and n-SiO2 in terms of refinement of pores/voids and capillary
seen in the rGO slices is at a lower level compared to the n-Al2O3 pores in OPC paste as determined from tomography and
and n-SiO2 pastes. porosimetry results.
58 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59

Fig. 14. Histogram representing the pore area fraction of the CT slices of the different pastes determined using the image analysis technique (Image Pro Premier 9.1).

Acknowledgement (2) (2012) 344e357.

[6] Z.S. Metaxa, M.S. Konsta-Gdoutos, S.P. Shah, Carbon nanofiber cementitious
composites: effect of debulking procedure on dispersion and reinforcing ef-
Partial financial support from the Nanoscience and Nano ficiency, Cem. Concr. Compos. 36 (0) (2013) 25e32.
Mission of the Department of Science and Technology (DST) (Grant [7] Z. Pan, L. He, L. Qiu, A.H. Korayem, G. Li, J.W. Zhu, et al., Mechanical properties
number: SR/NM/NS-12/2011(G)), Government of India, for sup- and microstructure of a graphene oxideecement composite, Cem. Concr.
Compos. 58 (0) (2015) 140e147.
porting our research program is sincerely acknowledged. [8] X. He, X. Shi, Chloride permeability and microstructure of portland cement
mortars incorporating nanomaterials, Transp. Res. Rec. J. Transp. Res. Board
2070 (2008) 13e21.
References [9] V. Singh, D. Joung, L. Zhai, S. Das, S.I. Khondaker, S. Seal, Graphene based
materials: past, present and future, Prog. Mater. Sci. 56 (8) (2011) 1178e1271.
[10] S. Lv, J. Liu, T. Sun, Y. Ma, Q. Zhou, Effect of GO nanosheets on shapes of cement
[1] S. Chuah, Z. Pan, J.G. Sanjayan, C.M. Wang, W.H. Duan, Nano reinforced cement
hydration crystals and their formation process, Constr. Build. Mater. 64 (0)
and concrete composites and new perspective from graphene oxide, Constr.
(2014) 231e239.
Build. Mater. 73 (0) (2014) 113e124.
[11] K. Gong, Z. Pan, A.H. Korayem, L. Qiu, D. Li, F. Collins, et al., Reinforcing effects
[2] J.M. Makar, G.W. Chan, Growth of cement hydration products on single-
of graphene oxide on portland cement paste, J. Mater. Civ. Eng. 27 (2) (2015).
walled carbon nanotubes, J. Am. Ceram. Soc. 92 (6) (2009) 1303e1310.
A4014010.
[3] S. Lv, Y. Ma, C. Qiu, T. Sun, J. Liu, Q. Zhou, Effect of graphene oxide nanosheets
[12] T. Kuila, S. Bose, A.K. Mishra, P. Khanra, N.H. Kim, J.H. Lee, Chemical func-
of microstructure and mechanical properties of cement composites, Constr.
tionalization of graphene and its applications, Prog. Mater. Sci. 57 (7) (2012)
Build. Mater. 49 (0) (2013) 121e127.
1061e1105.
[4] M.S. Konsta-Gdoutos, Z.S. Metaxa, S.P. Shah, Highly dispersed carbon nano-
[13] D. Li, M.B. Muller, S. Gilje, R.B. Kaner, G.G. Wallace, Processable aqueous dis-
tube reinforced cement based materials, Cem. Concr. Res. 40 (7) (2010)
persions of graphene nanosheets, Nat. Nano 3 (2) (2008) 101e105.
1052e1059.
[14] A. Nazari, S. Riahi, Al2O3 nanoparticles in concrete and different curing media,
[5] G. Quercia, G. Hüsken, H.J.H. Brouwers, Water demand of amorphous nano
Energy Build. 43 (6) (2011) 1480e1488.
silica and its impact on the workability of cement paste, Cem. Concr. Res. 42
 M. Murugan et al. / Cement and Concrete Composites 70 (2016) 48e59 59

[15] S. Kawashima, P. Hou, D.J. Corr, S.P. Shah, Modification of cement-based €rnstro
[19] J. Bjo €m, A. Martinelli, A. Matic, L. Bo
€rjesson, I. Panas, Accelerating effects
materials with nanoparticles, Cem. Concr. Compos. 36 (0) (2013) 8e15. of colloidal nano-silica for beneficial calciumesilicateehydrate formation in
[16] S. Sen Gupta, V. Manoj Siva, S. Krishnan, T.S. Sreeprasad, P.K. Singh, T. Pradeep, cement, Chem. Phys. Lett. 392 (1e3) (2004) 242e248.
et al., Thermal conductivity enhancement of nanofluids containing graphene [20] P. Hou, S. Kawashima, D. Kong, D.J. Corr, J. Qian, S.P. Shah, Modification effects
nanosheets, J. Appl. Phys. 110 (8) (2011) 084302. of colloidal nanoSiO2 on cement hydration and its gel property, Compos. Part
[17] D. Konios, M.M. Stylianakis, E. Stratakis, E. Kymakis, Dispersion behaviour of B Eng. 45 (1) (2013) 440e448.
graphene oxide and reduced graphene oxide, J. Colloid Interface Sci. 430 [21] E. Horszczaruk, E. Mijowska, R.J. Kalenczuk, M. Aleksandrzak, S. Mijowska,
(2014) 108e112. Nanocomposite of cement/graphene oxide e Impact on hydration kinetics
[18] F. Collins, J. Lambert, W.H. Duan, The influences of admixtures on the and Young's modulus, Constr. Build. Mater. 78 (0) (2015) 234e242.
dispersion, workability, and strength of carbon nanotubeeOPC paste mix- [22] K.K. Aligizaki, Pore Structure of Cement-based Materials: Testing, Interpre-
tures, Cem. Concr. Compos. 34 (2) (2012) 201e207. tation and Requirements, Taylor & Francis, London, 2006.

View publication stats