Advances in Cement Research Advances in Cement Research

http://dx.doi.org/10.1680/jadcr.17.00032
Paper 1700032
A study of deterioration of cement paste Received 16/02/2017; revised 18/08/2017; accepted 18/08/2017
due to acid attack using X-ray computed Keywords: chemical attack/leaching/microstructure

micro-tomography ICE Publishing: All rights reserved

Ramaswamy and Santhanam

A study of deterioration of cement paste

due to acid attack using X-ray computed
micro-tomography
Koodalloor P. Ramaswamy Manu Santhanam
Doctoral Research Scholar, Department of Civil Engineering, Indian Professor, Department of Civil Engineering, Indian Institute of Technology
Institute of Technology Madras, Chennai, India; Assistant Professor, Madras, Chennai, India (corresponding author: manus@iitm.ac.in)
Department of Civil Engineering, TKM College of Engineering, Kollam,
Kerala, India

Cementitious materials are highly vulnerable to rapid and serious deterioration in aggressive acidic environments. The
aim of this paper is to understand the microstructure deterioration of cement paste due to acid attack using the
non-destructive technique called X-ray computed micro-tomography (micro-CT). To investigate this deterioration,
an accelerated static immersion based leaching test with periodic abrasive action was performed on cement paste
specimens of size 10  10  60 mm, prepared using ordinary Portland cement with a water-to-cement ratio of 0·40.
After 28 d of initial curing in saturated lime water, specimens were exposed to 1% sulfuric, 1% hydrochloric, 0·5 M
acetic and 0·5 M citric acid solutions for a maximum period of 6 weeks. X-ray micro-CT was used for imaging the
deteriorated specimens to investigate the changes in microstructure. Various parameters related to degradation were
also deduced based on the image analysis performed on the CT images in order to evaluate the performance of the
cement paste in various acid solutions. Parameters such as altered depth/area could be considered as primary
indicators of attack, which can be best captured and analysed using X-ray CT.

Introduction mechanical properties on exposure to acid without following a

Cementitious materials, being highly alkaline in nature, are proper test protocol and do not focus on other possible degra-
easily prone to attack by acidic solutions. This disturbs the dation related parameters which could be directly related to
equilibrium of the cement matrix causing calcium leaching the deterioration in microstructure (Aliques-Granero et al.,
and associated degradation (Allahverdi and Škvára, 2000a, 2017). Mass losses and strength changes alone may not give a
2000b; Pavlik, 1994; Zivica and Bajza, 2001). Concrete struc- complete picture of deterioration, as they depend on the type
tures are exposed to aggressive acidic environments in varied of acid and the solubility of salts formed. The technique
situations, but predominantly in industrial environments. adopted and the test parameters developed should be such that
Many industries still remain dependent on the processes that the results are indicative of the microstructural deterioration
produce effluents or wastewater. Such effluents contain a wide due to acid attack. Altered depth measurements are important
range of both organic and/or inorganic acids whose action as they give an idea about the time taken for the degradation
may lead to severe and rapid degradation of concrete structures front to reach the reinforcement to cause corrosion in
(Aiken et al., 2017; Bertron, 2014). According to Scrivener and reinforced concrete structures (Bertron, 2014). Hence, altered
De Belie (2013) and Koenig and Dehn (2016a, 2016b), the depth could be considered as a primary indicator of attack and
deterioration of concrete due to acid attack is also observed in any technique which can capture the altered depth needs to be
concrete sewers, sewage/industrial effluent treatment plants and used for investigating the degradation kinetics. Altered depth
biogas reactors wherein acids are produced by microbiological can be measured roughly using digital calipers or stereo zoom
processes. Spillages and accidental leakages of acids in indus- microscope based on visual changes (changes in colour of
tries which manufacture or use synthetic acids would result in different zones of attack) by cutting the specimen perpendicu-
severe and rapid degradation as the concentration of acids lar to its axis, but this is subject to errors and limitations in
that come in contact with concrete structures may reach very measurement accuracy. Alternatively, phenolphthalein solution
high values. can be sprayed onto a fresh cut surface of the specimen and
altered depth can be measured based on the colour changes.
Degradation of cement-based materials due to acids is a However, the complete altered depth due to decalcification
complex phenomenon influenced by many factors related to and mineralogical zonation may not be picked out correctly
the attacking acid and the binder used, as well as the test using this method and the measurement accuracy will be com-
method (Dyer, 2014). This mechanism of degradation is still promised on smaller sized specimens. As an alternative, X-ray
not properly understood and this problem is aggravated by the tomography could be applied for this same investigation.
fact that there are hardly any codes or standards available for
evaluating the resistance of materials to acid attack. Also, Tomography is a technique for digitally cutting a specimen
most studies focus just on measuring mass loss and changes in open using X-rays to reveal its interior details and is derived

1
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Table 1. Chemical composition of OPC

Calcium Silicon Aluminium Iron (III) Magnesium Sodium Potassium Sulfur Titanium Diphosphorus
Chemical oxide dioxide oxide oxide oxide oxide oxide trioxide dioxide pentoxide
composition (CaO) (SiO2) (Al2O3) (Fe2O3) (MgO) (Na2O) (K2O) (SO3) (TiO2) (P2O5)

% by mass 64·59 19·01 4·17 3·89 0·89 0·16 0·59 1·70 0·24 0·11

from the Greek word ‘tomos’ which means ‘cut’ or ‘section’. Dehn, 2016a, 2016b; Rougelot et al., 2010; Wan et al., 2013).
X-ray computed tomography (CT) is a non-destructive However, most of the leaching studies were qualitative
technique for visualising the interior features within solid in nature, restricted to visualisation of degradation. Also, the
objects, and for obtaining digital information on their three- majority of the leaching studies were conducted using
dimensional (3D) geometries and properties. CT data have ammonium nitrate solution, which is a powerful leaching
been applied to virtually every discipline of science and tech- agent. The kinetics and mechanism of attack due to
nology, and new applications are being continually discovered. ammonium nitrate differ from the other acids that attack the
Various researchers have used CT in studying the microstruc- concrete. Limited studies are available on the application of
ture of building materials (Boel et al., 2008; Cesen et al., 2013; tomography in studying the deterioration due to acid attack. It
Chotard and Smith, 2003; Chotard et al., 2003; Diamond and is evident from the literature that tomographic imaging is an
Landis, 2007; Gallucci et al., 2007; Kim et al., 2012, 2013; appropriate technique to investigate the attack due to various
Lu et al., 2006; Ponikiewski et al., 2014; Promentilla, 2011; acids. Hence, this study is aimed at understanding the deterio-
Promentilla et al., 2009, Provis et al., 2012; Stock, 2008; Wang ration of cement paste specimens when exposed to various
et al., 2014; Wei et al., 2013; Wong and Chau, 2005). X-ray acids (organic and inorganic) that are commonly encountered
CT has also been used in creating models to predict the per- in field conditions using X-ray CT. Also, the possible ways in
meability and elastic properties of cementitious materials which the tomography technique can be applied to quantify
(Zhang, 2017; Zhang and Jivkov, 2016). The tomography various parameters to explain the deterioration due to acid
technique, being a non-invasive imaging technique, has the attack are detailed in the study. The extraction of these par-
potential to be used as one of the prime techniques to charac- ameters from CT could help in the performance evaluation of
terise the microstructure and durability associated problems in materials, thus aiding in the determination of acid resistance.
cementitious systems such as sulfate attack (Naik et al., 2006;
Stock et al., 2002), carbonation (Lydzba et al., 2014;
Wan et al., 2014) and leaching (Burlion et al., 2006; Sugiyama
Experimental programme
et al., 2010). Optical and electron microscopy have been com-
monly used, for instance, as research tools to examine the Materials
microstructure of concrete, especially while investigating dura- The materials used include ordinary Portland cement (OPC)
bility issues. However, the destructive or invasive way in which 53 grade (conforming to IS 12269 (BIS, 2013)) and distilled
specimens are prepared prior to imaging of the cross-section water. Cement was tested for its physical characteristics as per
may produce problematic artefacts and these techniques may the relevant Indian standards (BIS, 2013). OPC with specific
not be suitable for quantifying altered depth. In addition, the gravity and Blaine’s fineness values of 3·15 and 318 m2/kg,
limitations of observing a two-dimensional (2D) section of a respectively, and having a minimum 28th day compressive
3D structure from 2D microscopy must be borne in mind. strength (tested in accordance with IS 4031 (part 6)
According to Diamond and Landis (2007), tomography being (BIS, 2005a) using mortar cubes of 70 mm size) of 53 MPa
a 3D imaging technique, the microstructural features present in was used and the details of oxide composition are presented in
a 2D image such as air voids, sand grains, presence of cracks, Table 1. Particle size distribution of the cement was tested
hydrated and anhydrous phases and so on can be extended to using a laser diffraction technique and the D50 (average size) of
trace such features in the third dimension, a capability un- the cement particles was 19·4 μm. The loss on ignition of the
available to conventionally used microscopic techniques. cement when tested as per IS 4032 (BIS, 2005b) was 3·43%.

Deterioration of concrete in acidic environments results in The paste mix was prepared using distilled water with water-
premature degradation in terms of microstructural alteration to-cement ratio of 0·40. The mixing was done using a hand
of phases (mineralogical zonation) leading to mass changes, blender. The paste was poured into acrylic moulds (lubricated
weakening of mechanical properties, increase in porosity due using thin oil) and specimens of size 10  10  60 mm
to calcium leaching and so on. In past studies, tomography has were demoulded after 24 h. No chemical admixture was used
been successfully used by researchers to characterise the micro- for the OPC mix investigated in the current study. A water-
structure evolution and deterioration associated with calcium to-cement ratio of 0·40 was chosen so that most of the cement
leaching in cement-based materials (Dyer, 2016; Koenig and grains were hydrated at the time of acid exposure (after 28 d of

2
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

curing in lime water) and the mix was workable without

bleeding.

Methods
Currently, there are hardly any codes or standards available for
evaluating the resistance of materials to acid attack. Most of
the standards classify the aggressive environment based on pH,
which may not hold true in case of acidic environments, OPC
especially in the case of organic acids wherein the aggressive- 0·5M acetic
ness is governed by factors such as the type and concentration
of acid, polyacidity, buffer action of acid and so on (Bertron,
2014). According to De Belie (2013), accelerated tests are pre-
ferred in chemical attack studies, in which the degradation rate
can be increased by means of higher concentrations of the acid
solution, larger contact surfaces, smaller sized specimens and
so on. The studies performed on concrete have their own draw-
Figure 1. Experimental set-up
backs owing to the large specimen size involved. The altered
depth in concrete is much less when compared to the dimen-
sions of the specimen and any useful results can only be
obtained over the long term. Also, the results obtained may specimens after the brushing was measured using a weighing
not be distinguishable, if larger numbers of alternative binders balance with a precision of 0·1 mg, in saturated condition,
are tested for their suitability. Hence, accelerated tests need to after wiping the excess water from the specimen. Mass
be developed. Moreover, it is the paste phase that is attacked loss/gain was then calculated with respect to the initial mass
by the acid, hence, it is useful to study kinetics in the cement (before the acid exposure, i.e. after 28 d initial curing in satu-
paste instead of the concrete itself. Paste study also allows rated lime water) on a weekly basis, just before the renewal of
uniform deterioration to be measured without the disturbance the acid solution. The degradation was stopped (by solvent
of aggregates and quantification of parameters related to exchange process using isopropanol) after 6 weeks of exposure
kinetics may be more meaningful, especially while comparing for sulfuric, hydrochloric and acetic acids and after 3 weeks in
the performance of various binders in resisting acid attack. the case of citric acid (due to rapid mass loss and loss of thick-
Hence, a static immersion based accelerated leaching test with ness of specimens); the specimens were then stored in sealed
periodic abrasive action was conducted on prismatic cement zip lock covers inside a vacuum desiccator for further testing.
paste specimens of size 10  10  60 mm. The specimen size
was selected so as to have a high surface area to volume ratio X-ray micro-tomography
(0·43 mm2/mm3) to enable accelerated leaching. Following the Computed tomography is a technique for obtaining volumetric
initial 28 d of curing in saturated lime water, specimens were measurements of the X-ray attenuation coefficient, creating
exposed to 1% sulfuric, 1% hydrochloric (HCl), 0·5 M acetic images that map the variation of the X-ray attenuation co-
and 0·5 M citric acids. These acids are commonly encountered efficient within the objects. A schematic illustration of various
in practice; such acids differ in their mechanism of attack and processes involved in CT imaging is depicted in Figure 2.
hence are considered for the study. The volume of acid solution
to volume of specimen was kept fixed at 7. An X-ray beam is directed towards the specimen and the trans-
mitted beam is recorded on a detector. According to the
Two specimens of the mix were kept fully immersed in a boro- Beer–Lambert law, the ratio of the number of transmitted to
silicate glass beaker containing acid solution (as shown in incident X-ray photons is related to the integral of the linear
Figure 1) and the beaker was then covered tightly with an alu- attenuation coefficient (LAC) of the material along the path
minium foil wrap in order to avoid carbonation and exposure that the photons follow through the specimen. The resulting
to the outside environment as much as possible. Brushing of image (or radiograph) is a superimposed information or pro-
specimens (with a soft nylon brush) was done on a weekly jection of a volume in a 2D plane. To obtain the 3D infor-
basis before the acid replenishment to remove the loosely held mation, radiographic projections of the specimen are taken at
corroded particles (formed as a result of reaction of acids with many angles or projection views. In industrial CT scanners, the
cement hydrates) and also to accelerate the testing by exposing specimen is rotated and the X-ray source and detector are
fresh surface for the reaction to proceed inward further. Also, fixed in position, as shown in Figure 3. Once these projections
periodic abrasive action is expected to simulate the action of are available, a reconstruction algorithm can be used to
flowing effluents on cementitious surfaces, thereby eroding produce the contiguous 2D images, which provide a discrete
the degraded outer layer. Acid solutions were renewed every approximation of the distribution of X-ray attenuation coeffi-
week to maintain the aggressiveness of the acid. The mass of cient within the volume of the imaged specimen. This 2D

3
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Visible
X-rays light

Specimen on Series of projection

rotating stage Scintillator Photo detector
images (radiographs)

3D image
represented Tomographic
as a series reconstruction using
of 2D slices special algorithm

Figure 2. Schematic illustration of X-ray CT acquisition and reconstruction processes (adapted from Landis and Keane, 2010)

X-ray source Sample Detector xis

Z-a
Y-axis

Rotation
step by step

µ11
µ12
µ 21 µ13
µ22
X-rays µ 23

Y X-a 6 voxels
xis
256 pixels

X
Z

Data acquisition 256

Tube control Sample stage CNC pix
els
1 cm
Computed tomography/volume reconstruction
Figure 4. CT image illustrating the concept of pixel and voxel
(adapted from Jacobs et al., 1995)
Figure 3. Cone beam configuration for CT data acquisition

non-destructive 3D visualisation and characterisation of the

image is commonly referred to as a slice because it corre- internal structure without the time-consuming and difficult sec-
sponds to the cross-section of what would be seen if the speci- tioning of a specimen, as is done in 2D microscopy techniques.
men were sliced along the scan plane. Each slice is a matrix of As long as the spatial resolution is small with respect to the
voxels (volume element or 3D pixel) in which each voxel is microstructural feature of interest, the volumetric image
associated with a greyscale value (GSV) that is related to the obtained from these measurements can provide valuable 3D
measured LAC (Figure 4). For example, in an 8-bit CT image structural information. Reconstruction is generally attenuation-
with voxels, with GSV that ranges from 0 to 255, the higher conservative, allowing sub-voxel level details to be extracted.
the X-ray attenuation coefficient of the material component, Quantitative measurements can be made from the resulting
the higher the GSV of the corresponding voxel. Since the LAC digital image data, including the spatial distribution and
is sensitive to the atomic composition and density (Bossa volume fraction of phases, as well as the changes in phases due
et al., 2015), the CT technique can therefore provide to a range of mechanical or chemical phenomena.

4
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Table 2. Voxel resolution and details of CT images

Voxel
Concentration resolution:
Solution of acid Legend μm

Saturated lime water — OPC L 15·18

Sulfuric acid 1% OPC S 15·51
Hydrochloric acid 1% OPC H 12·80
Acetic acid 0·5 M OPC A 13·80
Citric acid 0·5 M OPC C 9·77

2 2

Front view
1 1
Figure 5. The v|tome|x s – tomography set-up

4 4 Top view

3 3

Figure 7. Schematic diagram of specimen showing sections for

obtaining CT slice images

420 mm high and 260 mm dia. A caesium iodide flat panel

detector (pixels of 1024  1024) is used (for converting X-rays
10 mm
to light).
Figure 6. Specimen for X-ray CT imaging
Cubical samples of size approximately 10 mm were cut, using
a diamond saw blade, from the end of the degraded specimens
(as illustrated in Figure 6) and the control specimens immersed
The present study attempts to characterise the microstructure in lime water. The ends of the specimens are generally
of the acid-degraded specimens using X-ray computed micro- degraded more and that is why specimens are cut from the end
tomography (X-ray micro-CT). Earlier, Ramaswamy et al. to extract the maximum information about the deterioration.
(2015) attempted tomography to understand the microstructure A solvent exchange process using isopropanol was used for
of cementitious materials modified using nano-materials. The specimen conditioning, followed by drying. The dried speci-
tomography test set-up at IIT Madras used for imaging mens were then taken for CT imaging. A micro-focus tube was
(Figure 5), ‘v|tome|x s’, is a versatile high-resolution system used to generate X-rays (conical beam as shown in Figure 3) at
consisting of both micro- and nano-focus tubes. To allow high a voltage of 65 kV and current of 120 μA. The voxel resolution
flexibility, the v|tome|x s is equipped with both a 180 kV/15 W obtained was in the range of 9·77–15·51 μm, as per the details
high-power nano-focus X-ray tube and a 240 kV/320 W micro- presented in Table 2. The specimen was rotated 360° and a
focus tube. This unique combination means that the system is total of 900 images were captured during the complete rotation
an ideal tool for a wide range of applications from extremely of the specimen on the stage, to reconstruct the 3D image. The
high-resolution scans of low-absorbing materials to the data acquisition time was 333 ms per image captured with an
3D analysis of high-absorbing objects up to 260 mm dia. averaging of four frames per image. The total time for scan-
Maximum voxel resolution for micro-focus tube is less than ning a specimen was about 25 min. After the reconstruction
2 μm and for a nano-focus tube is less than 1 μm. The process, the 3D image and 2D slice images were obtained.
maximum specimen mass that can be inspected is limited to Figure 7 is a schematic diagram of a specimen, showing the
10 kg, while the maximum specimen size that can be imaged is position of sections for obtaining 2D slice images from the 3D

5
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Before acid After acid exposure

exposure Sulfuric Hydrochloric Acetic Citric

10 mm

Figure 8. Visual appearance of specimens before and after the acid exposure

image. Sections 1–1 and 2–2 were cut to obtain the slice due to decalcification) and GSV when compared to the sound
images (top view) through the inside of the specimen and on zones that have not been attacked. An attempt was also made
the surface, respectively. Sections 3–3 and 4–4 were cut to to evaluate the porosity changes due to leaching. However, as
obtain slice images (front view) through the surface and from the voxel resolution obtained is in the range of 9·77–15·51 μm
the inside of the specimen, respectively. for a specimen size of 10 mm, only macro voids were visible,
the effect of which was minimal in the deterioration due to
Results and discussion acid attack and, hence, is not considered for the detailed analy-
The visual appearance of specimens before and after the degra- sis. Three distinct zones could be identified in all acids:
dation are shown in Figure 8. A white-yellowish-coloured an inner zone termed the core, a middle layer named as the
gypsum was precipitated on the specimen exposed to sulfuric progression front and the outer layer, which is degraded. The
acid. The specimen exposed to hydrochloric acid experienced a outer layer is degraded and decalcification is almost complete.
slight reduction in thickness and appeared orange brown in The middle layer could be called the progression/reaction front
colour. The specimen exposed to acetic acid was dark brown in for the attack to proceed inward, wherein the decalcification is
colour. It was noted that the deterioration was less on the ongoing and is partly completed. The inner core zone has not
external appearance in the case of the specimens exposed to been attacked and the microstructure is very similar to that
hydrochloric and acetic acids. Citric acid was aggressive, as it before the acid exposure.
led to the precipitation of white salt in enormous quantities.
A thin layer of gypsum (white-yellow-coloured precipitate on
The CT slice images, 3D rendered image and illustrative image visual observation) bound with a silica–alumina gel skeleton
showing void porosity of the control mix cured in saturated can be seen as an outer layer covering the specimen in the
lime water are shown in Figure 9. The slice images obtained 2D slice images of specimens exposed to sulfuric acid
from both the inside (section 1–1 in Figure 7) and the surface (Figure 10(a)). Calcium from the hydrates, such as portlandite
of the sample (section 2–2 in Figure 7) show uniform micro- and calcium–silicate–hydrate (C–S–H) gel, leaches out to the
structure. The 16 bit images obtained from CT were converted exterior and combines with sulfate ions from the acid solution
to 8 bit images before performing the image analysis. The 8 bit to form gypsum. As the solubility of the gypsum in the acid
CT images thus obtained had a GSV ranging from 0 to 255. solution is less, it precipitates and adheres to the surface of the
Pure black regions in the slice images refer to the voids (pixels specimen. Decalcification of hydrates leads to formation of
with GSV less than 50), white regions refer to the very amorphous gel rich in silica and alumina, which can be seen
dense/anhydrous particles (pixels with GSV greater than 250) along with gypsum. The thickness of the gypsum layer was
and grey regions indicate the various hydrated phases greater at the edges (due to penetration of acid from more than
formed, depending on the GSV. The void porosity as shown in one direction) and less thick on the lateral surfaces. The
Figure 9(d) is obtained by performing image segmentation. average thickness of the layer at the edges was 545 μm and
The voxels with GSV less than 50 were considered to be voids 180 μm on the surfaces. It was soft in nature and could be
and are visualised in Figure 9(d). removed by brushing, especially at the edges. Edges that
were more deteriorated due to decalcification appeared dark.
The specimens after the degradation (Figure 10) have clearly However, no cracks were observed on the gypsum that was
shown the mineralogical zonation due to alteration in phases adhered on the lateral surface, thus confirming the protective
(demarcated by white/black lines in Figure 10), which is a nature of the gypsum (Gutberlet et al., 2015), which played a
characteristic feature of acid attack. Darker portions in slice role in limiting the degradation kinetics against further attack
images are mostly deteriorated, and have lower density (mainly for the concentration of the acid studied (1%). Gypsum can be

6
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

1 mm 1 mm

(a) (b)

1 mm 1 mm

(c) (d)

Figure 9. CT images of OPC specimen before acid exposure: (a) top view (slice image) – inside; (b) front view (slice image) – surface;
(c) 3D rendered image; (d) voids in the specimen

considered to be an expansive precipitate (solubility of 2·4 g/l observed, as leaching and acidolysis were the driving processes
at 20°C) which might have blocked the pore spaces left behind of deterioration. Also, as the silica gel was adherent to the
by the decalcification of hydrated products, predominantly matrix and was not soft when compared to gypsum, the
portlandite, thus restricting the ingress of the acid further. The deterioration at the edges that was observed in sulfuric acid
thickness of the middle layer, which was in the process of dec- was absent in this case. Hence, the mass loss was predomi-
alcification (partly decalcified), was found to be less, at about nantly due to the combined effect of leaching and abrasive
257 μm. action. Calcium and aluminium salts are highly soluble (solu-
bility of 745 g/l and 458 g/l, respectively, at 20°C) and therefore
In the case of hydrochloric acid, a thick layer of degraded pro- leach out into the acid solution, leaving the silica gel porous,
ducts comprising mainly amorphous silica gel (about 869 μm) which results in further acid ingress, thus leading to a greater
was formed uniformly over the edges and on the surfaces degraded depth of the outer layer. As a result, the observed
(Figure 10(b)). The outer layer (orange brown) was darker and mass loss was higher (34·43%). However, despite the thick
porous (lower GSV of pixels in the outer layer), which could outer layer, the thickness of the middle layer was found to be
be attributed to the severe decalcification of hydrates and con- smaller – at about 107 μm.
sequently weak properties of the silica gel layer thus formed.
Decalcification-associated shrinkage cracks (map cracking due Even though acetic acid is generally considered as a weak acid,
to changes in volume on conversion of hydrated products to it was found to be highly detrimental to the cement matrix in
silica gel) were seen on the external surface of the altered zone. this study. Both the outer and the middle layer were observed
Precipitate formation as in the case of sulfuric acid was not to be thicker, as evident from Figure 10(c) (1825 μm and

7
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

545 µm

Gypsum

180 µm

1 mm 1 mm

(a) (b)

1 mm 0·65 mm

(c) (d)

Figure 10. CT slice images (cut through the inside of the specimen, section 1–1 in Figure 7) of OPC paste exposed to various acids:
(a) OPC paste exposed to sulfuric acid; (b) OPC paste exposed to hydrochloric acid; (c) OPC paste exposed to acetic acid; (d) OPC paste
exposed to citric acid

1550 μm, respectively), leaving behind only a small area of acetic acid. The outer zone and the middle layer were found to
core that had not been attacked. The outer layer was severely be very weak with respect to their mechanical properties such
decalcified, as evident from the low GSV of pixels from CT as strength and hardness (Bertron et al., 2004). The mechan-
images. Even anhydrous grains (white dots in images) were ism of attack was leaching/acidolysis and therefore similar to
absent in the outer layer and this indicates the instability of hydrochloric acid. The calcium acetate salt that forms is
anhydrous grains and that the outer layer had turned comple- soluble (solubility of 347 g/l at 20°C) and leaches out, making
tely amorphous. In addition to these three zones observed in the outer layer porous (undissolved), with primarily silica gel
all cases, a thin, dense, intermittent layer between the outer (Bertron et al., 2007). Hence, the mass loss observed after
and the middle layer was observed in the specimens exposed to 6 weeks was also higher (30·04%). The higher aggressiveness
acetic acid and this may be ascribed to the formation of and alteration depth could also be due to properties such as
aluminium and/or iron bearing products, which are deposited specificity, buffer action and complexation properties exhibited
as a thin layer of hydroxides. Aluminium and iron also leach by acetic acid (De Windt et al., 2015; Koenig and Dehn,
from the hydrates at a very low pH and are finally converted to 2016b).
corresponding hydroxides, which are seen as a thin, brown ring
encircling the middle layer. More importantly, the altered Citric acid was found to be the most aggressive among the
depth in this case was found to be the highest among the acids acids studied for the concentration investigated. On exposure
studied (about 3375 μm), thus confirming the aggressiveness of to citric acid, a white salt due to calcium citrate tetrahydrate

8
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

(Ca3(C6H5O7)2.4H2O) of very low solubility (0·95 g/l at 25°C) the decalcification as the average GSV of the layer was found
was precipitated in substantial quantities on the specimens to be lower when compared to the core layer. Figure 13(c) was
(Ramaswamy and Santhanam, 2016). The salt was totally non- obtained by performing the thresholding operation on the 3D
protective as it was loosely attached to the matrix, and there- image (showing only voxels having GSV greater than 120) and
fore this led to almost complete dissolution of the outer portrays the smaller core area left behind after the attack by
degraded layer. Hence, it was seen only as a thin layer of about acetic acid. The slice image obtained by cutting through the
303 μm (dark layer with lower GSVs for pixels) in the CT centre of the specimen exposed to citric acid (Figure 13(d))
images, as a majority of this zone underwent dissolution and confirms the non-protective nature of the salt formed. The thin
was eroded away due to leaching and/or brushing. The shape outer layer and cracking, as explained earlier, is also evident in
of the specimen was highly distorted due to the formation of this image.
loosely adhered precipitates, which were further removed
by brushing/leaching. As the salt possesses a higher molar The image analysis software called ImageJ was used for the
volume compared to hydrated products such as portlandite detailed analysis of the tomographic images to extract quanti-
and C–S–H (Larreur-Cayol et al., 2011), it resulted in the for- tative data. Histograms were plotted from the 8 bit images
mation of cracks in the outer layer, as shown in Figure 10(d). obtained from 2D slice images (based on Figures 10(a)–10(d))
As a result, the precipitate is exfoliated from the specimen, and are shown in Figure 14. It can be seen that the histograms
goes into the acid solution and is deposited in the beakers. of acid-attacked specimens shift to the left when compared to
Hence, the outer layer was restricted close to the surfaces only. the reference (OPC in lime water for 28 d), indicating the
This process of deterioration resulted in a severe loss of mass reduction of GSVs. Histograms of all specimens are standard-
(74·51%) and thickness loss, which is evident from the CT ised with respect to the GSV of anhydrous grains; in other
images. The higher aggressiveness could also be attributed words, the histograms are adjusted so that the GSV of anhy-
to polyacidity and complexation properties of citric acid drous grains in all the cases is equal to 255. As the GSVs are
(Dyer, 2016; Ramaswamy and Santhanam, 2016). However, proportional to the attenuation coefficient and hence the
the thickness of the middle layer was observed to be less density of the phases, this validates the fact that there is density
(196 μm) as in the case of sulfuric acid and hydrochloric acid. reduction of cementitious phases associated with decalcification
As the outer degraded layer undergoes dissolution, the altered and weakening of the microstructure associated with the degra-
depth alone cannot be used as an indicator of degradation. In dation process. However, as the GSVs may vary slightly on
this case, loss of thickness is also important and must be taken adjusting the brightness and contrast of the images after the
into consideration. reconstruction process, histograms are better used only qualitat-
ively to assess the degree of deterioration. Similarly, the vari-
In order to confirm the mechanism and to know the nature of ation of GSVs of pixels of CT slice images can be plotted with
the outer degraded layer, a section was cut through it in the the distance from the external surface of the specimen by
3D image (section 2–2 in Figure 7) and the resulting CT slice selecting and analysing the image using rectangular selections
images for various cases are shown in Figure 11. A thick of the area of image using the ‘plot profile’ feature in ImageJ.
gypsum formation without cracks was observed on these slice In the case of citric acid, as the shape is not well defined, the
images of the paste exposed to sulfuric acid. Extensive cracks average of six plot profile data obtained by drawing lines on
were observed, along with severe decalcification, in the slice images are taken to draw the GSV plot profile. The GSV
images of specimens exposed to both hydrochloric and acetic profile data thus obtained are shown in Figures 15(a)–15(e).
acid. The fragmentation of the matrix and subsequent delami- On exposure to acids such as hydrochloric, acetic and citric
nation due to higher molar volume of salt was evident in the acids, there was a considerable reduction in GSV near the
slice images of specimens exposed to citric acid. external surface, which indicates a remarkable reduction in
density and constitutes the outer degraded layer. A gradual
The 3D rendered images of the specimens are shown in increase of GSV was observed towards the inner part of the
Figure 12. Mass loss and damage at the edges are clearly image after the degraded layer and this could be attributed to
visible on the specimens exposed to sulfuric acid. As the salts the gradual decalcification of portlandite and other hydrated
are soluble, the external visual deterioration was less in the products; this forms the progression front for the attack to
case of hydrochloric and acetic acid. Higher aggressiveness and proceed inward. The GSV of the core zone is higher compared
rapid loss of mass and thickness are evident from the images to the middle and outer layers. In the case of sulfuric acid, an
of the specimens exposed to citric acid. increase of GSV can be observed near the external surface and
this could be due to the formation of calcium sulfate salts. This
Figures 13(a) and 13(b) show the CT images obtained by method of analysis could be more interesting if more binder
cutting through the various layers on specimens exposed to systems are investigated to study the effect in detail.
acetic acid. A slice image through the inner part of the outer
degraded layer reveals that the layer was dark due to silica gel Using ImageJ, various parameters were determined to study
formation. A slice image through the middle layer confirms the deterioration of OPC in acids and these are presented in

9
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Thick layer of gypsum

with silica gel

Cracks

1 mm
1 mm

(a) (b)

Cracking and
delamination
Cracks

1 mm 0·65 mm

(c) (d)

Figure 11. CT slice images of OPC paste specimens cut through the external surface: (a) OPC paste exposed to sulfuric acid;
(b) OPC paste exposed to hydrochloric acid; (c) OPC paste exposed to acetic acid; (d) OPC paste exposed to citric acid

Tables 3–5. The area of various layers was quantified and the area of each layer is normalised with respect to the final area
data presented in Table 3 are the average of ten slice images. after the attack. Normalising with respect to the core area
The terms Ao, Am and Ac denote the area of the outer layer, would result in bloated area values and this was hence avoided.
middle layer and core layer, respectively. Af represents the final The percentage of area decalcified is arrived at by adding the
area of the specimen after the acid exposure. The terms Ho, area of outer and middle layers and expressing this as a percen-
Hm and Hc denote the average GSV of pixels of areas Ao, Am tage of the final area. The percentage of area shrunk or
and Ac, respectively. The term DD refers to the degraded expanded was obtained by comparing the areas before and
depth, that is, the average thickness of the outer layer, whereas after the attack. It can be noted that the Ao/Af ratio is lower
DM refers to the average thickness (depth) of the middle layer. for specimens exposed to sulfuric acid. This reinforces the fact
The depth of alteration/altered depth (DA) is defined as the that the gypsum layer formed is protective and restricts the
sum of DD and DM. It is the depth for which the decalcifica- further penetration of acid into the specimen. The evidence of
tion is effective in causing deterioration. The parameters gypsum is also evident from the high density ratio, Ho/Hc.
defined above are illustrated in Figure 16. These depths Also, the progressing front (low value of Am/Af ) was found to
are measured using the software ImageJ after calibrating the be less in the case of sulfuric acid for the 1% concentration
images (using the value of voxel resolution). studied. Only 13·84% of the area was observed to be decalci-
fied and 86·16% of the final area remained that had not been
As the area of the core layer varies with the extent of degra- attacked. The total depth of decalcification was also found to
dation, a parameter called area ratio is defined wherein the be lower (437 μm) when compared to the other acids. Another

10
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

1 mm
1 mm

(a) (b)

1 mm 1 mm

(c) (d)

Figure 12. The 3D rendered images of OPC paste specimens exposed to various acids: (a) OPC paste exposed to sulfuric acid;
(b) OPC paste exposed to hydrochloric acid; (c) OPC paste exposed to acetic acid; (d) OPC paste exposed to citric acid

observation is that the ratio Hm/Hc is higher (1·03) and this higher percentage of area degraded could be associated with
may be ascribed to the formation of gypsum or ettringite in the higher solubility of the salts being formed. Acetic acid
the middle layers that are undergoing decalcification. attack resulted in strong calcium leaching, leading to serious
Formation of gypsum resulted in a 7·46% expansion in area, deterioration of the microstructure, which is evident from the
despite a mass loss of 7·13% observed. fact that the area of the core was the least (only 6·5% of the
final area). The area ratios, Ao/Af and Am/Af, were significantly
Calcium leaching associated with the high solubility of calcium higher when compared to the other acids. However, the density
chloride salts resulted in higher Ao/Af ratio of 0·32 for speci- ratios are comparable to those found for hydrochloric acid.
mens exposed to hydrochloric acid. The thickness of the Higher alteration depth and comparable density ratios could
middle layer was lowest (107 μm), resulting in an Am/Af ratio be due to the higher concentration of acetic acid investigated
of only 0·05. The outer layer was severely deteriorated and this (0·5 M) when compared to the hydrochloric acid (1%). The
is evident from a considerable reduction in density (Ho/Hc thickness of the outer layer and the middle layer was the
value of 0·47) and leads to higher degraded layer depth of highest of all (1825 μm and 1550 μm, respectively), which cor-
869 μm. The decalcification of the middle layer was clear from relates with the weak mechanical properties. As the salt formed
the lower value of Hm/Hc (0·81). It has been found that in the case of citric acid is non-protective and loosely bound to
37·22% of the area was observed to be decalcified and 62·78% the matrix, the outer degraded layer was thin (303 μm), as is
of the final area remained which had not been attacked. The evident from the lower area ratios and lower alteration depth.

11
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

1 mm 1 mm

(a) (b)

1 mm 0·65 mm

(c) (d)

Figure 13. CT images of OPC paste specimens exposed to acetic and citric acids showing details of various layers formed due to acid
attack: (a) OPC paste exposed to acetic acid (cut through the inner part of degraded layer, section 3–3 in Figure 7) (b) OPC paste exposed
to acetic acid (cut through middle decalcified layer) (c) 3D image of OPC paste exposed to acetic acid showing the unattacked core area
(d) OPC paste exposed to citric acid (cut through middle of the specimen, section 4–4 in Figure 7) showing the various layers of attack

12 000 (1) OPC lime water The lower density ratio (Ho/Hc of 0·56) means that the outer
(2) OPC sulfuric acid
(3) OPC hydrochloric acid (2) layer is very weak and porous. Figure 17 shows the area of
10 000 (4) OPC acetic acid each layer expressed as a percentage of the final area after the
(5) OPC citric acid
exposure to various acids. Hydrochloric acid was more aggres-
8000
(4) (1) sive when compared to sulfuric acid for the 1% concentration
(5)
Counts

6000 studied. The area of the core was lower for hydrochloric acid
when compared to sulfuric acid. The detrimental effect of
4000 acetic acid was further confirmed by the fact that the area of
(3)
the core was the least of all acids (only 6·5%). Even though the
2000 core area was greater for citric acid (69·73% of the final area),
the value of the final area was the least when compared to the
0
0 50 100 150 200 250 other acids. Rapid loss of mass and thickness led to a 78·43%
GSV reduction in the area when compared to the initial area
before the exposure. The parameters defined, such as area
Figure 14. Histogram of CT images. A full-colour version of ratios, density ratios and alteration depth, are quantitative data
this figure can be found on the ICE Virtual Library
(www.icevirtuallibrary.com) extracted from CT images. These parameters serve as more
direct indicators of degradation and, hence, should be used for

12
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Gypsum formation bound by

silica–alumina gel
200 200
Grey value

Grey value
Core layer (not attacked)
Middle decalcified layer
100 100

0 0
0 2000 4000 6000 8000 0 2000 4000 6000 8000 10 000
Distance: µm Distance: µm
(a) (b)

Thin layer of aluminium and iron hydroxides

200 200
Core layer (not attacked)
Grey value

Core layer (not attacked)

Grey value

Reaction front Middle decalcified layer

100 100

Outer degraded layer

Outer degraded layer
0 0
0 2000 4000 6000 8000 0 2000 4000 6000 8000 10 000
Distance: µm Distance: µm
(c) (d)

200
Grey value

Reaction front

100 Core layer (not attacked)

Outer degraded layer

due to salt formation
0
0 1000 2000 3000 4000
Distance: µm
(e)

Figure 15. GSV profile curve of CT images: (a) OPC in saturated lime water; (b) OPC in sulfuric acid; (c) OPC in hydrochloric acid;
(d) OPC in acetic acid (e) OPC in citric acid

Table 3. Degradation parameters obtained from image analysis of CT images

Area ratio
Percentage area Final area: Initial area: Percentage area
Legend A0/Af Am/Af Ac/Af decalcified: % mm2 mm2 shrunk/expanded: %

OPC S 0·09 0·05 0·86 13·84 107·94 100·44 +7·46

OPC H 0·32 0·05 0·63 37·22 77·77 96·73 −19·60
OPC A 0·62 0·31 0·07 93·50 83·12 97·93 −15·12
OPC C 0·18 0·13 0·70 30·27 22·37 103·73 −78·43

performance evaluation of materials while investigating degra- Conclusions

dation kinetics. Thus, CT images can be used to quantify par- This paper describes the application of X-ray micro-CT to
ameters that are linked with the alteration in microstructure, to investigate the alterations in the microstructure of cement paste
explain the behaviour of cement-based materials when exposed by acid attack. The qualitative and quantitative information
to various acids. generated from the tomographic data are helpful in assessing

13
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Table 4. Density ratios of specimens on exposure to acids

Average GSV of the area

Legend Ho Hm Hc Ho/Hc Hm/Hc

OPC S 196·26 ± 4·99 165·85 ± 7·12 161·14 ± 7·12 1·22 1·03

OPC H 85·08 ± 13·40 146·30 ± 16·48 179·84 ± 20·71 0·47 0·81
OPC A 112·00 ± 10·68 194·35 ± 14·25 219·98 ± 12·28 0·51 0·88
OPC C 71·48 ± 7·72 117·11 ± 7·14 126·74 ± 13·11 0·56 0·92

Table 5. Altered depth of specimens on exposure to acids Core area (not attacked)
Middle layer area (decalcified)
Depth of Depth of Outer degraded layer area
Degraded middle layer, alteration,
100
Legend depth, DD: μm DM: μm DA: μm Area of layers as percentage of

OPC S 180 257 437

80
final area after attack: %

OPC H 869 107 976

OPC A 1825 1550 3375
OPC C 303 196 499
60

40

Initial area
20

Ac Am Final area
Ao
0
OPC S OPC H OPC A OPC C
DM
Acid solution
DD DA
Figure 17. Areas of various layers formed on exposure to the
four acids
Figure 16. Schematic figure of slice image of specimen showing
various areas/depths

gypsum layer with silica–alumina gel had a protective effect in

the case of sulfuric acid attack for the 1% concentration of
the kinetics of degradation. The main drawback of this method the acid studied. As a result, the total depth of alteration was
is its spatial resolution, which is at present quite low relative to found to be only 437 μm, which is the least for the acids
the size of pores present in the cementitious systems. Also the studied. Owing to the limited penetration of sulfuric acid, only
specimen size is restricted in terms of obtaining increased 13·84% of the final area was decalcified and the remaining
resolution. Despite these limitations, this technique has a 86·16% of the final area was not attacked. Acidolysis/leaching
strong potential to gain widespread popularity as a useful tool is prominent in hydrochloric and acetic acid and, hence, the
for analysing the microstructure in cementitious systems. altered depth was found to be higher (976 μm and 3375 μm,
respectively). In the case of hydrochloric acid, 62·78% of the
The study validates the fact that X-ray micro-CT is an appro- final area of the specimen was not attacked, whereas it was
priate technique to understand the deterioration of cement- only 6·5% in the case of acetic acid, confirming the aggressive
based materials subjected to acid attack. Mineralogical nature of acetic acid. As citric acid is aggressive, the non-
zonation, altered depth/area and other degradation related par- protective nature and higher molar volume of the loosely
ameters are best captured by CT. The parameters defined in bound precipitate formed resulted in rapid loss of mass and
the paper could serve as more direct indicators of degradation thickness; the outer layer underwent complete dissolution
when compared to physico-mechanical test parameters such as and could be seen in the specimen only as a thin layer. Hence,
mass change, thickness and strength change. Based on the CT even though the total depth of alteration was found to be
images, it can be inferred that the depth of alteration and lower (499 μm), considering the thickness loss of 78·43%, citric
the extent of damage mainly depend upon the type and con- acid was found to be the most aggressive to the cementitious
centration of acid, and the solubility of salts formed by the matrix. Acid attack in cementitious systems is still a relatively
reaction of the acids with the cement hydrates. Formation of a unexplored area of research. There are hardly any codes or

14
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

standards available, thus far, for assessing the material resist- Chotard T, Smith A, Boncoeur M, Fargeot D and Gault C (2003)
ance to acid attack. Hence, the quest is on to develop new, Characterisation of early stage calcium aluminate cement
hydration by combination of non-destructive techniques: acoustic
reliable and flexible test methods and test parameters, which
emission and X-ray tomography. Journal of the European Ceramic
could then be applied to study a wide variety of materials, Society 23(13): 2211–2223.
especially on alternative binders, to arrive at suitable mixture De Belie N (2013) General considerations: methods for testing concrete
formulations. The results presented in this paper indicate that degradation in aggressive aqueous environments. In Performance
X-ray CT may be used in conjunction with other characteris- of Cement-Based Materials in Aggressive Aqueous Environments
(Alexander M, Bertron A and De Belie N (eds)). Springer,
ation techniques, such as X-ray diffraction, scanning electron
Dordrecht, the Netherlands, RILEM State-of-the-Art Reports,
microscopy, porosimetry, thermogravimetric analysis and so vol. 10, pp. 221–234.
on, for a complete understanding of the acid attack phenom- De Windt L, Bertron A, Larreur-Cayol S and Escadeillas G (2015)
enon in cement-based materials. Interactions between hydrated cement paste and organic acids:
thermodynamic data and speciation modeling. Cement and
Concrete Research 69: 25–36, https://doi.org/10.1016/j.cemconres.
2014.12.001.
REFERENCES Diamond S and Landis EN (2007) Microstructural features of a mortar
Aiken TA, Sha W, Kwasny J and Soutsos MN (2017) Resistance of as seen by computed microtomography. Materials and Structures
geopolymer and Portland cement based systems to silage effluent 40(9): 989–993.
attack. Cement and Concrete Research 92: 56–65, https://doi.org/ Dyer T (2014) Concrete Durability. CRC Press (Taylor & Francis
10.1016/j.cemconres.2016.11.015. Group), Boca Raton, FL, USA.
Aliques-Granero J, Tognonvi TM and Tagnit-Hamou A (2017) Durability Dyer T (2016) Influence of cement type on resistance to organic
test methods and their application to AAMs: case of sulfuric-acid acids. Magazine of Concrete Research 69(4): 175–200,
resistance. Materials and Structures 50(1): 1–14. http://dx.doi.org/10.1680/jmacr.16.00271.
Allahverdi A and Škvára F (2000a) Acidic corrosion of hydrated Gallucci E, Scrivener K, Groso A, Stampanoni M and Margaritondo G
cement-based materials: part 1: mechanisms of the phenomenon. (2007) 3D experimental investigation of the microstructure of
Ceramics-Silikáty 44(3): 114–120. cement pastes using synchrotron X-ray microtomography (μCT).
Allahverdi A and Škvára F (2000b) Acidic corrosion of hydrated cement Cement and Concrete Research 37(3): 360–368.
based materials: part 2: kinetics of the phenomenon and Gutberlet T, Hilbig H and Beddoe RE (2015) Acid attack on hydrated
mathematical models. Ceramics-Silikáty 44(4): 152–160. cement – effect of mineral acids on the degradation process.
Bertron A (2014) Understanding interactions between cementitious Cement and Concrete Research 74: 35–43, https://doi.org/10.1016/
materials and microorganisms: a key to sustainable and safe j.cemconres.2015.03.011.
concrete structures in various contexts. Materials and Structures Jacobs P, Sevensa E and Kunnenb M (1995) Principles of computerised
47(11): 1787–1806. X-ray tomography and applications to building materials.
Bertron A, Escadeillas G and Duchesne J (2004) Cement pastes alteration The Science of the Total Environment 167(1–3): 161–170.
by liquid manure organic acids: chemical and mineralogical Kim KY, Yun TS, Choo J, Kang DH and Shin HS (2012) Determination
characterization. Cement and Concrete Research 34(10): 1823–1835. of air-void parameters of hardened cement-based materials using
Bertron A, Duchesne J and Escadeillas G (2007) Degradation of cement X-ray computed tomography. Construction and Building Materials
pastes by organic acids. Materials and Structures 40(3): 341–354. 37: 93–101, https://doi.org/10.1016/j.conbuildmat.2012.07.012.
BIS (Bureau of Indian Standards) (2005a) IS 4031 (part 6) – 2005: Kim KY, Yun TS and Park KP (2013) Evaluation of pore structures
Methods of physical tests for hydraulic cement. BIS, New Delhi, and cracking in cement paste exposed to elevated temperatures by
India. X-ray computed tomography. Cement and Concrete Research 50:
BIS (2005b) IS 4032–2005: Method of chemical analysis of hydraulic 34–40, https://doi.org/10.1016/j.cemconres.2013.03.020.
cement. BIS, New Delhi, India. Koenig A and Dehn F (2016a) Biogenic acid attack on concretes in
BIS (2013) IS 12269–2013: Specification for 53 grade ordinary Portland biogas plants. Biosystems Engineering 147: 226–237,
cement. BIS, New Delhi, India. https://doi.org/10.1016/j.biosystemseng.2016.03.007.
Boel V, Cnudde V, De Schutter G and Van Meel B (2008) X-ray computed Koenig A and Dehn F (2016b) Main considerations for the
microtomography on cementitious materials: possibilities and determination and evaluation of the acid resistance of
limitations. In Proceedings of the International RILEM Symposium cementitious materials. Materials and Structures 49(5): 1693–1703.
on Concrete Modeling – CONMOD’08 (Schlangen E and Landis EN and Keane DT (2010) X-ray microtomography. Materials
De Schutter G (eds)). RILEM, Paris, France, pp. 487–494. Characterization 61(12): 1305–1316.
Bossa N, Chaurand P, Vicente J et al. (2015) Micro- and nano-X-ray Larreur-Cayol S, Bertron A and Escadeillas G (2011) Degradation of
computed-tomography: a step forward in the characterization of cement-based materials by various organic acids in agro-industrial
the pore network of a leached cement paste. Cement and Concrete waste-waters. Cement and Concrete Research 41(8): 882–892.
Research 67: 138–147, https://doi.org/10.1016/j.cemconres.2014. Lu S, Landis EN and Keane DT (2006) X-ray microtomographic studies
08.007. of pore structure and permeability in Portland cement concrete.
Burlion N, Bernard D and Chen D (2006) X-ray microtomography: Materials and Structures 39(6): 611–620.
application to microstructure analysis of a cementitious material Lydzba D, Rajczakowska M, Stefaniuk D and Kmita A (2014)
during leaching process. Cement and Concrete Research 36(2): Identification of the carbonation zone in concrete using X-ray
346–357. microtomography. Studia Geotechnica et Mechanica 36(1): 47–54.
Cesen A, Korat L, Mauko A and Legat A (2013) Microtomography Naik NN, Jupe AC, Stock SR et al. (2006) Sulphate attack monitored by
in building materials. Materials and Technology 47(5): 661–664. microCT and EDXRD: influence of cement type, water-to-cement
Chotard TJ and Smith A (2003) Application of X-ray computed ratio and aggregate. Cement and Concrete Research 36(1): 144–159.
tomography to characterise the early hydration of calcium Pavlik V (1994) Corrosion of hardened cement paste by acetic and
aluminate cement. Cement and Concrete Composites 25(1): nitric acids; part I: calculation of corrosion depth. Cement and
145–152. Concrete Research 24(3): 551–562.

15
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research A study of deterioration of cement paste
due to acid attack using X-ray computed
micro-tomography
Ramaswamy and Santhanam

Ponikiewski T, Katzer J, Bugdol M and Rudzki M (2014) Determination Stock SR, Naik NN, Wilkinson AP and Kurtis KE (2002) X-ray
of 3D porosity in steel fibre reinforced SCC beams using X-ray microtomography (microCT) of the progression of sulfate attack of
computed tomography. Construction and Building Materials cement paste. Cement and Concrete Research 32(10): 1673–1675.
68: 333–340, https://doi.org/10.1016/j.conbuildmat.2014.06.064. Sugiyama T, Promentilla MA B, Hitomi T and Takedam N (2010)
Promentilla MAB (2011) Characterizing the microstructure of Application of synchrotron microtomography for pore structure
deteriorated cementitious materials with X-ray microtomography. characterization of deteriorated cementitious materials due to
ASEAN Engineering Journal 1(4): 12–21. leaching. Cement and Concrete Research 40(8): 1265–1270.
Promentilla MAB, Sugiyam T, Hitomi T and Takeda N (2009) Wan K, Xu Q, Li L and Sun W (2013) 3D porosity distribution of partly
Quantification of tortuosity in hardened cement pastes using calcium leached cement paste. Construction and Building Materials
synchrotron-based X-ray computed microtomography. Cement 48: 11–15, https://doi.org/10.1016/j.conbuildmat.2013.06.073.
and Concrete Research 39(6): 548–557. Wan K, Xu Q, Wang Y and Pan G (2014) 3D spatial distribution of the
Provis JL, Myers RJ, White CE, Rose V and Van Deventer JSJ (2012) X-ray calcium carbonate caused by carbonation of cement paste. Cement
microtomography shows pore structure and tortuosity in and Concrete Composites 45: 255–263.
alkali activated binders. Cement and Concrete Research 42(6): Wang J, Dewanckele J, Cnudde V et al. (2014) X-ray computed
855–864. tomography proof of bacterial-based self-healing in concrete.
Ramaswamy KP and Santhanam M (2016) Degradation kinetics of Cement and Concrete Composites 53: 289–304, https://doi.org/
cement-based materials in citric acid. Proceedings of the 10th 10.1016/j.cemconcomp.2014.07.014.
Structural Engineering Convention (Rao ARM (ed.)). Sigma Wei S, Yiqiang C, Yunsheng Z and Jones MR (2013) Characterization and
Publications, Chennai, India, pp. 1690–1695. simulation of microstructure and thermal properties of foamed
Ramaswamy KP, Santhanam M and Murugan M (2015) Characterization concrete. Construction and Building Materials 47: 1278–1291,
of cement paste modified with nano-materials using X-ray https://doi.org/10.1016/j.conbuildmat.2013.06.027.
computed microtomography. Proceedings of the 3rd International Wong RCK and Chau KT (2005) Estimation of air void and aggregate
Conference on Modeling and Simulation in Civil Engineering spatial distributions in concrete under uniaxial compression using
(Bindu S and Thomas AV (eds)). Excel India Publishers, New computer tomography scanning. Cement and Concrete Research
Delhi, India, pp. 279–289. 35(8): 1566–1576.
Rougelot T, Burlion N, Bernard D and Skoczylas F (2010) About Zhang M (2017) Pore-scale modelling of relative permeability of
microcracking due to leaching in cementitious composites: X-ray cementitious materials using X-ray computed microtomography
microtomography description and numerical approach. Cement images. Cement and Concrete Research 95: 18–29, https://doi.org/
and Concrete Research 40(2): 271–283. 10.1016/j.cemconres.2017.02.005.
Scrivener K and De Belie N (2013) Bacteriogenic sulfuric acid attack Zhang M and Jivkov AP (2016) Micromechanical modelling of
of cementitious materials in sewage systems. In Performance of deformation and fracture of hydrating cement paste using X-ray
Cement-Based Materials in Aggressive Aqueous Environments computed tomography characterisation. Composites Part B:
(Alexander M, Bertron A and De Belie N (eds)). Springer, Engineering 88: 64–72, https://doi.org/10.1016/j.compositesb.2015.
Dordrecht, the Netherlands, RILEM State-of-the-Art Reports, 11.007.
vol. 10, pp. 305–318. Zivica V and Bajza A (2001) Acidic attack of cement based materials – a
Stock SR (2008) Recent advances in X-ray microtomography applied to review (part 1): principle of acidic attack. Construction and
materials. International Materials Reviews 53(3): 129–181. Building Materials 15(8): 331–340.

How can you contribute?

To discuss this paper, please submit up to 500 words to the
editor at journals@ice.org.uk. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial board, it will be published as a
discussion in a future issue of the journal.

16
Downloaded by [ University of Saskatchewan] on [10/10/17]. Copyright © ICE Publishing, all rights reserved.