Cement and Concrete Research 67 (2015) 226236

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: http://ees.elsevier.com/CEMCON/default.asp

Impact of accelerated carbonation on OPC cement paste blended with

y ash
A. Morandeau a,b,c,d, M. Thiry c,e, P. Dangla d,
a

Department of Civil and Environmental Engineering, Princeton University, Princeton, NJ, USA
Andlinger Center for Energy and the Environment, Princeton University, Princeton, NJ, USA
Universit Paris-Est, IFSTTAR, MAT, F-75732 Paris, France
d
Universit Paris-Est, Laboratoire Navier, Ecole des Ponts ParisTech, IFSTTAR, CNRS, F-77455 Marne-la-Valle, France
e
DGAC/STAC, F-94485 Bonneuil-sur-Marne, France
b
c

a r t i c l e

i n f o

Article history:
Received 12 February 2014
Accepted 3 October 2014
Available online xxxx
Keywords:
Carbonation (C)
Calcium-silicate-hydrate (C-S-H) (B)
Fly ash (D)
Microstructure (B)
Portland cement (D)

a b s t r a c t
Cement is a huge carbon dioxide producer. Supplementary cementitious materials can help reduce this outcome.
However, carbonation of these blended cements remains an active subject of research. Accelerated carbonation
tests (10% CO2, 25 C and 62% RH) are performed on y ash blended cement pastes. Experiments are performed
at varying ages of carbonation (1 to 16 weeks) to measure the evolution of the carbonation depth over time and
to quantify key parameters: thermogravimetric analysis (TGA), mercury intrusion porosimetry (MIP) and
gamma ray attenuation method (GRAM). The total porosity decreases with a rearrangement of the microstructure due to carbonation and the creation of big capillary pores for the paste with the highest contents of y
ash (60 vol.%). The C-S-H molar volume evolution during y ash-blended cement carbonation is calculated
using a method combining MIP, TGA and GRAM formerly successfully applied to OPC paste in a paper published
in the same journal.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
The worldwide cement industry accounts for at least 5 to 7% of the
anthropogenic CO2 emissions [1], hence solutions that reduce ordinary
Portland cement (OPC) content in concrete are needed in order to
decrease its environmental impact. Blended cements using supplementary cementitious materials (SCM) (such as y ash, slag, limestone,
metakaolin, silica fume) are increasingly being used in industry to
reduce the OPC content, but it is necessary to understand the long
term performance and durability of these materials in order to optimize
the whole life cycle. For this purpose, an evaluation of the durability of the
reinforced concrete structure including the effect of CO2 (carbonation) is
necessary.
Carbonation is the reaction of gaseous atmospheric CO2 with the
calcium-bearing phases of concrete and is known to cause a lowering
of alkalinity leading to the corrosion of the re-bars. The hydration products which are concerned by carbonation are mainly calcium hydroxide
(CH1) and calcium silicate hydrates (C-S-H). In this work, the inuence
of the substitution of y ash in clinker has been studied in order to
Corresponding author.
E-mail addresses: antoinem@princeton.edu (A. Morandeau),
mickael.thiery@aviation-civile.gouv.fr (M. Thiry), patrick.dangla@ifsttar.fr (P. Dangla).
1
It is recalled that according to the cement notations: C = CaO, H = H2O, S = SiO2, C
CO2 , A = Al2O3 and F = Fe2O3.

http://dx.doi.org/10.1016/j.cemconres.2014.10.003
0008-8846/ 2014 Elsevier Ltd. All rights reserved.

evaluate the resistance to carbonation of blended cement-based materials with y ash.

For OPC, CH is the main buffering compound. In blended cement, the
contribution of C-S-H will be a key point to investigate since the pozzolanic reaction that takes place between portlandite and reactive silica
from y ash decreases the CH content and increases the C-S-H amount
[2]. Mechanically, this reaction is considered as benecial by clogging
the bigger capillary pores and rening the microstructure [3], or lling
cracks [4]. Nevertheless, such concretes blended with y ash may become more sensitive to the presence of carbon dioxide due to a lower
ability of the matrix to bind CO2 in CH. Hence their chemical resistance
against carbonation is decreased. Many studies have been carried out
on OPC systems blended with y ash [511]; however it is still uncertain
how carbonation inuences the pore size distribution (rearrangement
of the microstructure) in the case of y ash-blended cements.
The current literature concerning the microstructure of materials
made of y ash-blended OPC cements and the impact of carbonation
don't show a real consensus. Moreover, the observations seem to be
highly dependent on the conditions of carbonation (CO2 concentration,
relative humidity (RH) and temperature). In the case of OPC cement
pastes (w/c ranging from 0.3 to 0.6), Pihlajavaara [12,13] was the rst
to point out that carbonation affects the capillary pores with diameters
between 10 and 100 nm by globally reducing their volume. Matsusato
et al. [14] observed the same trend on OPC mortars where the porosity
in the range of 20 nm300 nm was reduced by carbonation. Despite a

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

systematic reduction in total bulk porosity, Houst [15] observed using

mercury intrusion porosimetry on OPC cement paste of high water
cement ratio (w/c = 0.8) that there was a shift in the porosity towards
greater pore radii during carbonation. Recently, Miragliotta [16] and
Thiery et al. [17] came to the same conclusion for low grade OPC
concretes. Bier et al. [18] investigated the effect of carbonation on
the pore size distribution of cementitious materials which were
manufactured with different types of cement. It was found that for
OPC cement, accelerated carbonation led to a considerable reduction
in the capillary pore volume, whereas, for cement blended with blast
furnace slag, a coarser capillary porosity was formed during carbonation. Moreover, it is worth noting the investigations of carbonation
of air lime mortars [19,20]. These studies showed an increase in pore
volume at around a pore diameter of 100 nm which was attributed to
the transformation of CH macrocrystals to CC microcrystals. Furthermore, the authors observed a monotonic increase in the volumes of
pores with diameters below 30 nm which was attributed to the attachment of CC crystals at the surface of aggregate particles and/or at the
surface of CH macrocrystals.
Thus the development of blended cements with supplementary
cementitious materials (such as y ash, slag, limestone, metakaolin
and silica fume) is a large area of research and innovation in the eld
of cement sciences [21]. In this work, destructive methods including
phenolphthalein spray test, thermogravimetric analysis, and mercury
intrusion porosimetry are combined to gamma ray attenuation (nondestructive) in order to investigate the evolution of porosity and pore
size distribution according to the carbonation level and the amount of
y ash incorporated in the cementitious system. This work follows a
recent study provided by the same authors [22] where a method was
presented to determine the molar volume evolution of C-S-H during
its carbonation. The same method will be applied in the current study:
OPC cement paste blended with y ash is submitted to accelerated conditions of carbonation (10% CO2).
2. Materials
2.1. Formulation and fabrication
The cement used in this study is an ordinary Portland cement CEM I
(CEM I 52,5 N CE PM-ES CP2 NF). The Bogue composition is 59% C3S, 19%
C2S, 2% C3A, 14% C4AF and 5% gypsum. Class F y ash (Silicoline) is used
as a supplementary cementitious material (chemical composition given
by the manufacturer: 50 wt.% SiO2, 29 wt.% Al2O3, 8.5 wt.% Fe2O3). Five
formulations were studied with a y ash volume substitution xed at
0%, 30% or 60% (cf. Table 1). The highest substitution ratio is out of
standards according to EN 206. CN and CP are two well-known formulation compositions previously studied in the literatures [23,24] in
terms of moisture (ad- and desorption) and transport properties and
more recently [22] in terms of microstructural and chemical changes

227

due to accelerated carbonation (10% CO2). This last article detailed all
the experimental protocols we used in the present work; hence this
article will present the experimental methods only briey.
In order to avoid early drying, sealed plastic bottles were used as
moulds for the preparation of the cement pastes. The plastic bottles
were: (i) lled in 3 steps and air was removed by vibrating the samples
between each step, (ii) sealed using a cork and paralm, and (iii) put in
an anti-segregation rotating system for 24 h. They were then left for at
least 6 months at 20 C before being subjected to carbonation. The y
ash-blended cement pastes were left for at least 11 months so that the
pozzolanic reaction had time to stabilize [25].

2.2. Pretreatment
After the sealed curing period (between 6 and 11 months, depending on the formulation), plastic bottles were cut in two pieces in order
to obtain two cylinders (height 60 mm, = 70 mm). The samples
were preconditioned prior to carbonation by covering each sample
with a self-adhesive aluminium foil over the lateral face and bottom
faces, so that all transport phenomena occurred unidimensionally. The
specimens were placed in a temperature-controlled oven at 45 C for
56 days. Finally, the specimens were stored for 56 days at 20 C and
62 5% RH over a saturated NH4NO3 solution. This last pretreatment
was used to accelerate the carbonation mechanism by controlling the
moisture content of the specimens as close as possible to the optimum
RH in the range of 40%70% [8,2629]. Note that the pretreatment duration was too short to obtain homogeneous RH in the samples, but these
heterogeneous moisture proles were quantied by gamma-ray attenuation measurement as reported in Section 3.2 and in [22].

2.3. Accelerated carbonation test

Accelerated carbonation was performed using an incubator (Sanyo
MCO5-AC) which controlled the temperature (T = 25 0.5 C) and
CO2 concentration. The relative humidity was xed using a saturated
salt solution of NH4NO3 (RH = 63 5%). Temperature, relative humidity and [CO2], were monitored using HM70 and GM70 probes (Vaisala).
A 10% CO2 concentration was chosen in order to be consistent with our
previous investigation [22] where we showed that X-ray diffraction
patterns were similar compared with natural carbonation and 10%
CO2 accelerated carbonation on mortars made of the same cement.
This choice was also made to accelerate the carbonation test in a reasonable time limit and thus avoid strong drying during the experiment, but
without leading to a dramatically different microstructure than the one
observed in tests conducted at a lower CO2 concentration. Obviously,
this may be questioned when adding y ash since these materials
bring aluminium in the system.

Table 1
Formulations used in this work and water-to-binder ratio (cement and y ash), water-to-cement ratio, and volumic and mass fraction of y ash. The highest substitution ratio is out of
standards (cf. EN 206). The hydration degree is calculated using CH content (TGA). Porosity is measured either by MIP or by GRAM.
Characteristics

Units

CN

CN30

CN60

CP

CP30

Fly ash (fa)

Fly ash (fa)
w/(c + fa)
w/c

vol.%
wt.%

0
0
0.45
0.45

30
23
0.45
0.64

60
51
0.45
1.13

0
0
0.6
0.6

30
23
0.6
0.86

Noncarbonated
Porosity (MIP)
Porosity (GRAM)
Hydration degree (TGA)

%
%

19.2 0.7
37.9 0.7
0.86 0.02

29.2 1.1
47.4 1.2
0.83 0.01

37.8 1.0
52.9 1.7
0.98 0.01

31.0 2.1
47.1 0.6
0.84 0.04

42.7
54.9 1.0
0.93 0.01

Carbonated
Porosity (MIP)
Porosity (GRAM)

%
%

12.3 1.1
30.2 1.1

23.4 2.7
36.0 0.8

36.4 1.2
46.5 0.9

23.1 3.1
36.5 1.3

34.2 2.1
48.8 0.8

228

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

2.4. Techniques of carbonation monitoring

Several experimental techniques had been used in order to quantify
the evolution of carbonation after 1 to 16 weeks of exposure of the specimens inside the CO2 incubator.
2.4.1. Phenolphthalein spray test
Phenolphthalein is a pH indicator which changes progressively
from pink-fuchsia to colourless when the pH drops from 10 to 8.3. The
specimens dedicated to the phenolphthalein test were split in two
half-cylinders. Phenolphthalein was sprayed upon both exposed fresh
surfaces. The carbonation depth was determined at each test age in six
points along the 6 cm-diameter exposed surface. The median result is
used to dene the carbonation depth XC in order to get rid of the side
effect due to potential leakage effects and thus 2-D diffusion at the interface between the aluminium fold and the sample.

to the so-called ink bottle effect, but as long as these limitations are
taken into account during interpretation of the results, the technique
is still valuable in making comparative assessments of the pore size
changes occurring in a given cementitious system [40], as it is the case
in this work.
2.4.4. Thermogravimetric analysis
Thermogravimetric analysis (TGA) has been carried out with a
Netzsch STA 449 F1 Jupiter coupled with a mass spectrometer (MS)
Netzsch QMS 403 C Quadrupole in order to determine the composition
of cementitious materials in terms of CH and CC. Samples were freeze
dried, ground with mortar and pestle, and sieved at 315 m. The heating
rate was 10 C/min from 20 C to 1100 C.
Note that the computed CH and CC molar contents are corrected by
subtracting the minimal value observed in each prole.
3. Results

2.4.2. Gamma-ray attenuation method

The gamma-ray attenuation method (GRAM) is a non-destructive
test able to determine density proles in building materials. GRAM
has commonly been used to quantify porosity and water content proles within concrete [3033]. Moreover, by making reasonable assumptions (e.g. no signicant water transfer occurs by drying in the short run
of an accelerated carbonation test), GRAM can also be used to assess
carbonation proles within concrete by measuring the local density increase due to carbonation (CO2 binding) [32,34,22,35].
2.4.3. Mercury intrusion porosimetry
Mercury intrusion porosimetry (MIP) was performed with a
Micromeritics Autopore IV which determines pore sizes in the range
of 3 nm to 500 m (corresponding to a 0.0035 MPa400 MPa range of
applied pressure). At different times, cylindrical specimens of cement
paste were sawn underwater using a wheel saw (0.5 cm thickness).
MIP requires complete removal of water from the sample prior to
intrusion of the mercury. Moreover, in order to avoid re-hydration of
the cementitious system during the sawing step, the slices were
freeze-dried just after sawing. It is known that the choice of the pretreatment method strongly impacts the obtained result [36,37]. Moreover, the true pore structure cannot be described by MIP [38,39] due

3.1. Evolution of the carbonation depth detected by the phenolphthalein

spray test
Fig. 1 shows XC as a function of time for the studied cement pastes
CN, CN30, CN60, CP and CP30 (error bars represent the minimal and
maximal values). It is evident that, as more y ash is substituted, the
depth of carbonation increases. However, this is only partially true
when comparing CN and CN30, especially after 16 weeks of accelerated
carbonation. Indeed, XC is stable for CN30 between 8 and 16 weeks. This
unexpected behaviour can highlight the limits of the phenolphthalein
spray test which is only showing the region where the pH drops
below 8.2 without further information on the chemical composition of
the interstitial solution.
3.2. Carbonation proles obtained by GRAM
Fig. 2 illustrates the density variation as a function of depth for each
formulation analysed by GRAM. Each curve is the average of 3 different
samples (only 2 samples for 16 weeks). While the CO2 diffuses (from
left to right in the illustrated gures), it increases the density of the
cementitious matrix. This density can be calculated and related to the

CN
CN30
CN60

50

40

XC (mm)

40

XC (mm)

CP
CP30

50

30

30

20

20

10

10

0
0

10 12

t (days1/2)

t (days1/2)

(a)

(b)

10 12

Fig. 1. Carbonation depth XC as a function of the square root of time measured by phenolphthalein spray test.

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

0.4

0.4

1 week
2 weeks
4 weeks
8 weeks
16 weeks

0.3

-3
- 0 (g.cm )

229

0.3

0.2

0.2

0.1

0.1

0
-0.1

-0.1
-0.2

20

40

x (mm)
Fig. 2. Density evolution prole after 1 week,
4 different samples.

weeks,

20
40
x (mm)
weeks,

weeks and

20
40
x (mm)

20
40
x (mm)

20
40
x (mm)

-0.2

weeks of accelerated carbonation for the ve cement paste formulations. Each curve is the average value of

formation of calcium carbonate (CC). The CN and CN30 samples behave

in a similar manner as the density increases during the testing period
(16 weeks). This trend can also be observed for CP until 8 weeks, but
after 16 weeks of accelerated carbonation, the density begins to decrease. For CP30 and CN60 the data show a decrease in the density for
the depth x N 20 mm. The most plausible explanation is that the intense
drying during carbonation drained the liquid water produced by carbonation from the centre of the sample to the exposed surface (as the
surface is in equilibrium with the outer environment).
This assumption is conrmed by the results presented in Fig. 3
where the liquid water saturation prole (Sw) corresponding to the
water-lled porosity determined by vacuum saturation and drying at
105 C is illustrated. For each formulation, at a given carbonation time
t, one sample was dedicated to the calculation of Sw in three steps:
(i) GRAM measurement was made at time t, (ii) after the sample was

immersed in water while in a vacuum environment for 72 h a second

GRAM is performed, and (iii) the last measurement was made on the
dried sample (105 C until mass stabilization, i.e. m 0.05 % in 24 h
[41]). Porosity (w) and liquid water saturation proles were obtained
using this same 3-step procedure [22,32]. Since this is based on the
quantication of the water contained in the pores, we will refer to
GRAM porosity as accessible-to-water porosity (as opposed to mercury
porosity).
All the formulations present a highly heterogeneous initial Sw prole
as illustrated in Fig. 3, meaning that the pretreatment only partially
dries the samples and the homogenization step is not efcient enough.
However, this is not seen as a problem here since this initial state has
been quantied by GRAM and the effect of the heterogeneity of the
initial moisture prole on the measured carbonation penetration can
be assessed.

Fig. 3. Liquid water saturation evolution prole at the initial state, after week, weeks, weeks and
weeks of accelerated carbonation for the ve cement paste formulations plus a
non-carbonated (NC) control sample (dashed line) exposed to the same T and RH without CO2 for at least 16 weeks.

230

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

Fig. 4. Water content (in molL1 of porous material) evolution prole at the initial state, after week, weeks, weeks and
weeks of accelerated carbonation for the ve cement paste
formulations plus a non-carbonated (NC) control sample (dashed line) exposed to the same T and RH without CO2 for at least 16 weeks.

Given that carbonation also affects the evolution of the porosity, the
pertinent value to quantify a potential drying effect during the carbonation test is not Sw, but nw the molar content of free water molecules
per unit volume of porous material (molL1).
nw w Sw

w
Mw

where w and Mw are the density and molecular weight of water,

respectively.
Fig. 4 illustrates the nw prole for the ve formulations at various
stages of carbonation. The control samples are kept in a RH-controlled
desiccator at 62 5% and 20 C for 16 weeks. The curve shape for
CN30, CN60, CP and CP30 in Fig. 4 agrees with Fig. 3. However, this is
not so for the 16 week CN curves where the maximum at 15 mm in

Fig. 3 (8 and 16 weeks) has almost vanished in Fig. 4. This is due to

the change in the porosity. While nw is only impacted by drying, the
liquid water saturation ratio Sw is also impacted by the change in the
volume of the solid phase which articially increases Sw. For the CN
paste, nw increases slightly due to carbonation of the hydrated phases.
nw also increases for CN30, but only after 1 week of carbonation in the
vicinity of the exposed surface. For CP, nw increases at the edge of the
sample (x b 10 mm) but this is more likely due to transfer of water by
capillary gradient from the sample centre to the edge, rather than
water released by carbonation. In particular, a signicant drying occurs
for CN60 and CP30 in the course of the 16 weeks of accelerated carbonation. On the other hand, the moisture prole for the control samples
(compared to the initial state) does not evolve in a signicant way:
slight drying is observed for CN, CN30, CP30, and wetting for CN60
and CP. This proves that the observed changes of the proles of moisture

Fig. 5. Porosity (water porosity) evolution prole at the initial state, after week, weeks, weeks and
weeks of accelerated carbonation for the ve cement paste formulations plus a
non-carbonated (NC) control sample (dashed line) exposed to the same T and RH without CO2 for at least 16 weeks.

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

3.3. Impact of carbonation on the microstructure

In our previous work on OPC cement pastes [22], the pore size distribution (PSD) is signicantly affected during accelerated carbonation by
the clogging of the whole range of pores accessible to MIP (1 m to
4 nm). In the case of OPC systems blended with y ash the results are
quite different (see Fig. 6 for average curves).
In Fig. 6a, we do not observe any obvious pore blocking on the non
carbonated sample for a pore radius of 400500 nm, However, we can
see that the porosity below 50 nm is increasing. In Fig. 6b and c, the
pore structure of the non-carbonated sample (average value) also
shows a large population at very ne pore sizes and a much smaller
population at somewhat coarser pore sizes. This is not surprising and
it is a typical result of the pore blocking effect of C-S-H created by pozzolanic reactions that can be observed on well cured OPC samples
blended with SCM containing reactive silica [3,42]. This effect is less
obvious for CN30 because of the lower y-ash substitution ratio.
If we now compare these results to the carbonated sample, the
threshold radius (the radius before which the cumulative pore volume

200

60
55
50

40
35
30
25
20
15
10
5
0
CN

CN30 CN60

CP

CP30

Fig. 7. Total porosity (%) comparison measured by mercury intrusion (Hg) or by GRAM
(w) between the non carbonated (NC) and carbonated (C) states.

rises sharply) has increased by one order of magnitude (from 30 nm

to 200 nm). This can be explained by the carbonation of the pozzolanic
C-S-H gel that forms during the pozzolanic reaction in the large pores.
This C-S-H suffers a severe decalcication, and, even if the total porosity
decreases, the formation of big capillary pores (typically greater than
50 nm) is observed. This observation is consistent with the results
obtained on similar formulations (blended y ash or blended y ash
and slag) for various carbon dioxide concentrations [10,4346].
In Fig. 7 we nally compare the porosity results, for all the formulations, between NC and C states, for both MIP and GRAM porosity. MIP
porosity results are typically lower than GRAM because mercury does
not access the whole range of pores (usual maximal mercury pressure
is 400 MPa which corresponds to 3 nm pore), the ner C-S-H gel
pores (3 nm) being only reached by water. The MIP and GRAM results
agree with previous carbonation investigations on CEM I [12,47,43] and
supplementary cementitious materials [44,10], performed over a wide
range of CO2 concentration (1.5 to 100%). When the substitution ratio
increases, the porosity variation between a non-carbonated sample
and a carbonated sample is less important. However, we did not observe
any increase in porosity due to carbonation (related most likely to

300

CN30
NC
C

w NC
Hg NC
w C
Hg C

45
porosity (%)

content are certainly due to the carbonation process itself, and not to a
process of redistribution of the moisture content within the specimens.
In Fig. 5, the porosity proles measured by GRAM (w) are presented. For the initial non-carbonated states, as the y ash content increases,
so does the porosity (comparison of CN, CN30 and CN60 for instance).
The initial porosity proles are evenly distributed, except for CN60.
Furthermore, the difference in the porosity between the initial and
nal control samples is at most 23%, which highlights that the variability from one sample to another is minimal compared to the effect of
carbonation.
As carbonation proceeds, the decreasing porosity proles indicate
a systematic clogging effect, even for CN60. This illustrates a wellidentied phenomenon for OPC systems where the formation of CC
clogs the pores since its molar volume is higher than the parent hydration products (CH and C-S-H). For CN60, the average non-carbonated
porosity decreases from 53% to 46% when carbonated. For CP30, it
decreases from 55% to 46%. It is also important to note that the carbonation reaction is stabilized for both CN60 and CP30 after 16 weeks
which makes it possible to dene an average porosity representing
the carbonated state (see Table 1). Indeed, a reduction in porosity
should lead to a decrease of the moisture transfer properties and a
slow down of the drying process during the course of the carbonation
test. On the contrary severe drying is observed for CN60 and CP30
when compared to non-carbonated control sample.

231

250

CN60
NC
C

250

CP30
NC
C

200

100

50

V/ log(rp) (mm3/g)

3
V/ log(rp) (mm /g)

V/ log(rp) (mm3/g)

150
200
150
100

0
10

100
1000
rp (nm)

(a)

10000

100

50

50
0

150

0
10

100
1000
rp (nm)

(b)

10000

10

100
1000
rp (nm)

(c)

Fig. 6. Average pore size distribution for non carbonated (NC) and carbonated (C) samples for CN30 (a), CN60 (b) and CP30 (c).

10000

232

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

nCaCO
3
nCH
nCaCO (CSH)

CN30 16 weeks

-1

CN30 4 weeks

nCaCO or nCa(OH) (mol.L )

CN30 2 weeks

0
0

10

20

30

40

50

0
0

10

x (mm)

20

30

40

50

10

x (mm)

20

30

40

50

x (mm)

Fig. 8. TGA results for CN30 after 2, 4 and 16 weeks of accelerated carbonation: molar content of portlandite nCH, of calcium carbonate nCC and of calcium carbonate issued from C-S-H
decalcication nCSH as a function of the sample depth.
CC

pozzolanic C-S-H carbonation) as observed by other authors [48,45,46],

for higher CO2 concentration than the 10% used here.
3.4. Carbonation proles obtained by thermogravimetric analysis
Fig. 8 depicts the TGA proles obtained for CN30 and Fig. 9 presents
the same results for CN60 after 2, 8 and 40 weeks of accelerated carbon
ation.
 The molar content of portlandite (nCH), of total calcium carbonate
nCC and of calcium carbonate issued from the carbonation of other
hydration products than CH is illustrated after 2, 4 and 16 weeks of
accelerated carbonation as a function of the sample depth. These other

4.5

nCaCO
3
nCH
nCaCO (CSH)

nCaCO or n Ca(OH) (mol.L -1)

4.5

CN60 2 weeks

carbonates come from the carbonation of the solid phases containing

calcium, i.e., mainly C-S-H and some amounts of AFm, AFt and
hydrogarnet such as C3AH6. For OPC with a low substitution ratio of
y ash (CN30) the aluminium content is likely to be negligible compared to the calcium content, since the cement has a low C3A content.
Hence, the carbonates coming from the carbonation of solid phase
other that portlandite can be approximated to nCSH [33,22]. This
CC
assumption may be challenged for CN60 as seen in the discussion (see
Section 4).
In both Figs. 8 and 9, carbonation of portlandite and C-S-H occurs
simultaneously. Residual portlandite (non-carbonated) can be observed

4.5

CN60 4 weeks

3.5

3.5

2.5

2.5

2.5

1.5

1.5

1.5

0.5

0.5

0.5

3.5

10

20

30

x (mm)

40

50

10

20

30

x (mm)

40

50

CN60 40 weeks
nCaCO
3
nCH
nCaCO (CSH)
3

10

20

30

40

50

x (mm)

Fig. 9. TGA results for CN60 after 2, 4 and 40 weeks of accelerated carbonation: molar content of portlandite nCH, of calcium carbonate nCC and of calcium carbonate issued from C-S-H
decalcication nCSH as a function of the sample depth.
CC

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

at the edge of the sample after 4 and 16 weeks for CN30 and after 2
and 4 weeks for CN60. This conrms the results obtained by GRAM
(see Fig. 2) which shows a lower carbonation degree at the edge of
the specimen, mainly due a too low liquid water saturation in this
region as conrmed by Turcry et al. [49]. A lack of water near the surface
allows for faster diffusion of CO2 in the gas phase until the depth where
the carbonation reaction can occur, i.e. where water content is high
enough for CO2 dissolution. This residual portlandite may also be the
result of accelerated conditions since it is not always observed in the
case of naturally carbonated samples.
4. Discussion
4.1. Quantication of the available calcium content
As expected, the initial CH content in the NC samples is lower when
y ash is substituted in the cement (see Fig. 10a). This is due to not only
the dilution effect (lowering of the clinker content), but also the effect
of the pozzolanic reactions (which consume CH). In the case of CN60,
there is almost no portlandite in the NC state. The main source of calcium
is C-S-H, conrmed by the TGA proles.
Whatever the formulation, carbonation of C-S-H and CH is not instantaneous. This is highlighted by the broad calcium carbonate prole
(see Figs. 8 and 9). Due to calcium carbonate coating around CH crystals
all the portlandite will not be dissolved during carbonation [29]. This
phenomenon is also observed in natural conditions [50]. By subtracting
the minimal remaining CH content in the carbonated samples from
the initial CH content of NC materials we quantify the calcium content
available from CH carbonation. This amount is added to the calcium content in the C-S-H (assessed through a hydration model [51] using
Papadakis's work [52] and the cement oxide composition) in order to
calculate the theoretical total available calcium content. In Fig. 10a, we
compared it to the total experimental CC content when the carbonation
reaction is stabilized. We are clearly missing calcium (yellow bar higher
than green bar). This illustrates that all the calcium in the C-S-H is
not available for carbonation. Since the carbonation state is stabilized
(e.g. after 40 weeks of accelerated carbonation for CN60, see Fig. 9), it
is assumed that this difference is not due to kinetics effect, but mostly

13
11

Calcium content (mol.L-1) (TGA or model)

-1
Content (mol.L ) (TGA + model)

12

attributed to the fact that (i) the ner water-saturated gel pores are
hardly accessible to carbon dioxide by aqueous diffusion and that
(ii) C-S-H can suffer an accessibility reduction that does not lead to
their total carbonation. Note that Visser [53] considers that all calcium
is available for carbonation and that experimental results showing the
contrary either correspond to not well carbonated samples or are due
to accelerated carbonation side effects. In this work, an almost carbonation state has been reached, this presence of non carbonated C-S-H
may thus be explained by the accelerated conditions.
In the following section, we will try to quantify the a priori amount
of calcium available and thus the amount of C-S-H that is theoretically
available for carbonation as a function of the total calcium content
deduced from the cement oxide composition. We need hypothesis on
the C-S-H carbonation mechanism. The rst one being that the inner
product (or high density C-S-H) will not carbonate. The second idea is
to consider that the more porous C-S-H gel is the only one which can
be easily reached by carbon dioxide. MIP results on blended cement
pastes (on CN60 and CP30 where macropores were created during
carbonation) also support the idea that pozzolanic C-S-H carbonates
easily. Fig. 10b presents the experimental total CC content compared
to the addition of three contributors to calcium release in carbonation:
(i) the carbonated CH content (assessed by TGA), (ii) the low density
C-S-H content (assessed according to [54]) and (iii) the pozzolanic
C-S-H content (hydration model [51] using Papadakis work [52]).
The gure shows that the calcium content from CH + LD-C-S-H +pozzolanic C-S-H (see Fig. 10b) is higher than the total experimental calcium content from CC (green bars): the hypothesis stating that all
pozzolanic C-S-H are readily carbonated or that all of the low density
C-S-H are accessible to carbon dioxide is thus not fully satisfactory
for all ve formulations. For CN, CN30 and CN60, this approach appears
correct, but it does not hold true for CP and CP30. The experimental
results show that not all the C-S-H is able to carbonate, even for the
high w/b formulations, and that the LD-C-S-H hypothesis including
pozzolanic C-S-H is only relevant for low w/b ratios. It may also be
due to (i) the fact that for high w/b ratios, the LD-C-S-H model overestimate the real quantity of C-S-H that has a low density, or (ii) simply
because this classication cannot be applied to carbonation since it is
based on nitrogen sorption experiments (the size of nitrogen molecules

13

nCa in CH
nCa in CSH (model)
nCa CSH + CH
Total CaCO3 (exp)

10
9
8
7
6
5
4
3
2
1
0

233

12

Total Ca (exp)
CSH-LD model

11
10
9
8
7
6
5
4
3
2
1
0

CN

CN30 CN60

(a)

CP

CP30

CN CN30 CN60 CP

CP30

(b)

Fig. 10. Calcium content in molL1 of porous material. Comparison between (i) available calcium in CH (TGA) for carbonation, (ii) total calcium in C-S-H (hydration model) (iii) total
amount of calcium available a priori (iv) nal CC when the carbonation reaction is stabilized (a). In (b) is illustrated the comparison between (i) the nal CC (carbonated) and (ii) sum
of the available calcium in CH (TGA) for carbonation, in LD-C-S-D and in the pozzolanic C-S-H (see Section 4.1).

234

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

is slightly smaller than carbon dioxide), (iii) one part of the pozzolanic
C-S-H is not carbonated or, (iv) because the distinction between low
and high density C-S-H is too simple to picture the real diversity in
the C-S-H structure. It is also important to note that the initial CS ratio
of the C-S-H has an inuence on its ability to release calcium [55,56]
and that the lower the CS is, the less likely the C-S-H will be releasing
calcium (threshold value of CS = 0.75 corresponding to an increasing
structural disorder in the C-S-H structure).
Fly ash contains aluminium, and it may be pertinent to evaluate
the impact of the hypothesis neglecting solid phases other than CH
and C-S-H (AFm, AFt and hydrogarnet C3AH6). Indeed, carbonation of
hydrogarnet produces hydrated aluminium oxide and CC [57] and
carbonation of ettringite produces CC, gypsum, hydrated aluminium
oxide and releases water [58]. We can easily conclude that a small
part of CC measured by TGA comes from the carbonation of these last
phases. The amount of calcium that we assume available in C-S-H for
carbonation is even lower than the amount we quantify because we
include a small part of CC coming from these secondary solid phases.
This reinforces the previous conclusion stating that C-S-H does not
fully carbonate.
4.2. Porosity evolution vs. chemical changes
For CN (w/b = 0.45 CEM I), it has been shown [22] that CH and C-S-H
contribute to the formation of CC. If one considers that the porosity variation is only due to CH (CH) and C S H (CSH), this variation can
1
37 cm3  mol stands
be expressed as proposed in Eq. (2) where V CH
CC
for the molar volume of the calcium carbonate formed by CH carbonation
(mainly calcite), and VCH = 33 cm3 mol1 is the CH molar volume. The
calcium carbonate content produced by CH carbonation, nCH comes from
CC
TGA and comes from GRAM.


CH
CH
CH CSH nCC V CC V CH CSH

Fig. 11 illustrates that CH is really small and that the majority of

the porosity variation is due to the difference in molar volume between
non-carbonated C-S-H and calcite plus decalcied C-S-H. This trend is
even higher for the other four formulations. Note that the contribution

14

w
Hg
from Ca(OH)2

12

- (%)

10
8
6

of CH in the porosity variation is negligible in the case of CN60.

These results conrm the need to take into account the important
effect of C-S-H in the microstructure evolution process during carbonation of any cement-based material, especially for high substitution in
clinker.
The difference between GRAM (i.e. water) porosity variation
(w) and MIP porosity variation (Hg ) is not signicant for CN,
CP and CP30. However for the formulations CN30 and CN60 there
are evident differences between w and Hg. Since MIP does not
allow investigating the microstructure below the nanometric scale,
this conrms that the intrinsic C-S-H porosity is affected during the
carbonation of these formulations of blended-y ash cements and that
their nanostructure is more accessible to CO2. The results describing
the pore size distribution show that formulations containing y ash
suffer from a severe redistribution of the PSD. The pore clogging effect
observed commonly on CEM I [12,13] is here affecting only the nest
pores (rp 30 nm) while the biggest capillary pores are opening
(rp 100 nm). This corresponds to a volumetric redistribution since
for CN60 the total mercury porosity does not change signicantly after
carbonation whereas the total GRAM porosity decreases by w
6 %. This can only mean that the gel porosity (rp 3 nm) is reduced
by the formation of small calcium carbonate crystals that are lling
the space between the low density C-S-H and are responsible for the
accessibility reduction of C-S-H [59,60] which is being transformed
into hydrous silica.
These results are consistent with the observations made in other
works [43,29,44,61,10,45,46] where the porous structure of CEM I
with a low w/c is changed so that small capillary pores close, and
where large capillary pores open when the w/c formulation increases.
One part of the pozzolanic C-S-H dissolves during carbonation,
reopening capillary porosity which results in the opening of the
macropores observed in the carbonated PSD, while another part of
C-S-H is coated with CC, and thus remains non-carbonated.
In order to better assess the cause of the porosity variation during
carbonation, MIP, GRAM and TGA have been combined. This study
was rst carried on OPC samples [22] at a macroscopic scale, i.e. by combining molar balance, TGA results and water porosity results, in order to
identify the links between the changes of the chemical composition of
the material due to carbonation (degree of carbonation of each hydration compound and amount of calcium carbonate formed) and the evolution of the global porosity. Given that MIP does not give access to the
whole range of pore size, we consider here that it is more relevant to assess the total porosity variation by GRAM. This variation can be written
as proposed previously in Eq. (2).
Using a hydration model and a global molar calcium balance (TGA),
makes it possible to calculate a macroscopic CS ratio [62] as a function
of the calcium carbonate content produced by C-S-H carbonation nCSH
CC
and the initial calcium content in the C-S-H nCSH
Ca (t0) (see Eq. (3)). It is
C
essential to keep in mind that the assessment of S provided here in
case of non carbonated and carbonated systems is based on a molar
balance of calcium and silicon, and not on a Q1/Q2 investigation by 29Si
Nuclear Magnetic Resonance which leads to different values.
!
nCSH
C
C
CC
t t 0 1 CSH
S
S
nCa t 0

4
2
0
CN CN30CN60 CP CP30
Fig. 11. Porosity variation measured by mercury intrusion (Hg) or by GRAM (w) between
the non carbonated (NC) and carbonated (C) states and contribution to CH carbonation to
this porosity evolution.

CxS1Hz is chosen as the stoichiometry of the C-S-H. The formed silica

gel corresponds to C-S-H with CS ratio (C over S molar ratio) equals to
zero. It can be considered that the amount of solid silicon is held constant over time, i.e., nCSH(t0) = nSi(t0) nSi(t) between an initially
non-carbonated state (t0) and a given age of carbonation (t). This assumption is mainly based on the fact that the amount of dissolved silica
in the pore solution remains negligible in comparison with the solid
amount of silicon in C-S-H.

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

0.04

-1

VCSH(initial)-VCSH (L.mol )

The molar volume variation of classic C-S-H and pozzolanic C-S-H

may not be the same during carbonation. For OPC and low substitution
ratio (below 30% y ash), we can assimilate these two in order to quantify the average molar volume evolution of C-S-H during carbonation as
a function of C/S without any hypothesis on H/S. It is an useful result
for modelling carbonation since volumic modications are the key
parameters modifying the transport properties. For high y-ash substitution ratio over OPC, the calculation is not accurate and highlight the
necessity to take into account the presence of AFm, AFt and hydrogarnet
phases.
All these conclusions are based on accelerated carbonation test
results. We showed that all the calcium initially in the cement are not
available for carbonation. The high CO2 concentration used here for
the tests (10%), but also the drying pretreatment, may be responsible
for this behaviour which is not systematically observed in natural
conditions.

poro G to ATG
CN
CN30
CN60
CP

0.03

235

0.02

0.01

0
References

-0.01
0

0.4

0.8

1.2

1.6

C/S ratio in CSH

Fig. 12. C-S-H molar volume variation as a function of C/S.

If we consider the porosity variation due to C-S-H we can add two

contributions in Eq. (4), CC precipitation and C-S-H decalcication.
CSH

CSH

CSH nCC V CC nCSH t 0 V CSH t V CSH t 0 

Eventually, combining Eqs. (3), (4) and the fact that CS  nCSH nCSH
Ca ,
we get Eq. (5) which allows plotting of the evolution of the C-S-H molar
volume at a given carbonation time t, VCSH(t), as a function of the C-S-H
decalcication state quantied by the macroscopic CS (see Fig. 12). We
clearly see that, in this case, the trend followed by CN and CP (no y
ash) is not obvious for CN30 and CN60.
The molar volume variation of classic C-S-H and pozzolanic C-S-H
might not be the same. The uncertainties are also important and it is
hard to obtain an investigation of the full range of CS since the y-ah
blended samples carbonates quickly.
0
CSH

CSH B CSH
nCC @V CC

1
C
C
t t
S 0
S

C
V CSH t V CSH t 0 A

5. Conclusion
Methods have been developed in order to combine TGA, MIP and
GRAM measurements. Following a previous study on OPC based cement
pastes and mortars. We have monitored carbonation of y-ash blended
cement pastes.
For blended carbonated pastes, when the substitution ratio is high
(above 30%) combined with a high water to binder ratio (above 0.6),
the pastes develop coarser capillary pores which allows for extensive
drying to occur, even if the total porosity decreases. The moisture transfer properties change and the materials can easily dry, especially water
released by carbonation of the hydration product. This can be even more
problematic for durability concerns that are not limited to carbonation
by increasing the speed of chloride or sulfate penetration in concrete.
This is not the case for OPC based cement paste where the whole
range of pores is clogging, and thus free water is not easily transported.

[1] C. Chen, G. Habert, Y. Bouzidi, A. Jullien, Environmental impact of cement production:

detail of the different processes and cement plant variability evaluation, J. Clean.
Prod. 18 (2010) 478485.
[2] V. Papadakis, Effect of y ash on portland cement systems Part I. Low-calcium y
ash, Cem. Concr. Res. 29 (11) (1999) 17271736.
[3] P.K. Mehta, O.E. Gjrv, Properties of portland cement concrete containing y ash and
condensed silica-fume, Cem. Concr. Res. 12 (5) (1982) 587595.
[4] H. Ranaivomanana, Transferts dans les milieux poreux ractifs non saturs: application la cicatrisation de ssure dans les matriaux cimentaires par carbonatation
(PhD thesis) Univ Paul Sabatier Toulouse, 2010.
[5] C. Atis, Accelerated carbonation and testing of concrete made with y ash, Constr.
Build. Mater. 17 (3) (2003) 147152.
[6] J. Khunthongkeaw, S. Tangtermsirikul, T. Leelawat, A study on carbonation depth
prediction for y ash concrete, Constr. Build. Mater. 20 (9) (November 2006)
744753.
[7] D. Burden, The Durability of Concrete Containing High Levels of Fly Ash(PhD thesis)
University of New Brunswick, 2006.
[8] E. Drouet, Impact de la temprature sur la carbonatation des matriaux cimentaires prise en compte des transferts hydriques(PhD thesis) ENS Cachan, 2010.
[9] A. Younsi, Carbonatation de btons forts taux de substitution du ciment par des
additions minrales(PhD thesis) Universit de La Rochelle, 2011.
[10] M. Saillio, Interactions physiques et chimiques ions-matrice dans les btons sains et
carbonats, Inuence sur le transport ionique, Universit Paris-Est-Marne-La-Valle,
2012. (PhD thesis).
[11] X. Wang, Modlisation du transport multi-espces dans les matriaux cimentaires
saturs ou non saturs et ventuellement carbonats(PhD thesis) Universit ParisEst, 2012.
[12] S.E. Pihlajavaara, Some results of the effect of carbonation on the porosity and pore
size distribution of cement paste, Mater. Struct. 1 (6) (1968) 521526.
[13] S.E. Pihlajavaara, E. Pihlman, Effect of carbonation on microstructural properties of
cement stone, Cement Concr. Compos. 4 (1974) 149154.
[14] H. Matsusato, K. Ogawa, M. Funato, T. Sato, Studies on the carbonation of hydrated
cement and its effect on microstructure and strength, 9th International Congress
on the Chemistry of Cement (ICCC). Volume 5: Performance and Durability of
Concrete and Cement Systems, New Dehli, India, National Council for Cement and
Building Materials, 1992, pp. 363369.
[15] Y. Houst, Diffusion de gaz, carbonatation et retrait de la pte de ciment durcie
(PhD thesis) EPFL, Lausse (Switzerland), 1992.
[16] R. Miragliotta, Modlisation des processus physico-chimiques de la carbonatation
des btons prfabriqus - Prise en compte des effets de parois(PhD thesis)
Universit de La Rochelle, juin, 2000.
[17] M. Thiery, G. Villain, G. Platret, Effect of carbonation on density, microstructure and
liquid water saturation of concrete, in: D.A. Lange, K.L. Scrivener, J. Marchand (Eds.),
Advances in Cement and Concrete, Copper Mountain, Colorado, U.S.A., 2003.
Engineering Conferences International (E.C.I.), 2003, pp. 481490.
[18] T.A. Bier, J. Kropp, H.K. Hilsdorf, Carbonation and realcalinisation of concrete and hydrated cement paste, in: J.C. Maso (Ed.), Durability of Construction Materials, Chapman and Hall, London-New York, 1987, pp. 927934.
[19] D.R. Moorehead, Cementation by the carbonation of hydrated lime, Cem. Concr. Res.
16 (1986) 700708.
[20] R.M. Lawrence, T.J. Mays, S.P. Rigby, P. Walker, D. D'Ayala, Effects of carbonation on
the pore structure of non-hydraulic lime mortars, Cem. Concr. Res. 37 (7) (2007)
10591069.
[21] K.L. Scrivener, R.J. Kirkpatrick, Innovation in use and research on cementitious
material, Cem. Concr. Res. 38 (2) (2008) 128136.
[22] A. Morandeau, M. Thiry, P. Dangla, Investigation of the carbonation mechanism of
CH and CSH in terms of kinetics, microstructure changes and moisture properties,
Cem. Concr. Res. 56 (2014) 153170.
[23] V. Baroghel-Bouny, Water vapour sorption experiments on hardened cementitious
materials. Part I: essential tool for analysis of hygral behaviour and its relation to
pore structure, Cem. Concr. Res. 37 (3) (March 2007) 414437.

236

A. Morandeau et al. / Cement and Concrete Research 67 (2015) 226236

[24] Vronique Baroghel-Bouny, Water vapour sorption experiments on hardened

cementitious materials. Part II: essential tool for assessment of transport properties
and for durability prediction, Cem. Concr. Res. 37 (3) (2007) 438454.
[25] V.G. Papadakis, M.N. Fardis, C.G. Vayenas, Hydration and carbonation of pozzolanic
cements, ACI Mater. J. 89 (2) (1992).
[26] G. Verbeck, Carbonation of hydrated portland cement, ASTM Spec. Publ. (205)
(1958) 1736.
[27] H. Wierig, Longtime studies on the carbonation of concrete under normal outdoor
exposure, In R.I.L.E.M. Seminar, R.I.L.E.M, 1984, pp. 239249.
[28] V.G. Papadakis, C.G. Vayenas, M.N. Fardis, Physical and chemical characteristics
affecting the durability of concrete, ACI Mater. J. 88 (2) (1991).
[29] M. Thiry, Modlisation de la carbonatation atmosphrique des matriaux
cimentaires Prise en compte des effets cintiques et des modications microstructurales et hydriques(PhD thesis) ENPC, Champs sur Marne, 2005.
[30] M.A. Esam, A compton scattering method for inspecting concrete structures, Nucl.
Instrum. Methods Phys. Res. Sect. A 283 (1) (1989) 100106.
[31] M.C. Da Rocha, L.M. Da Silva, C.R. Appoloni, O. Portezan Filho, F. Lopes, F.L.
Melquiades, E.A. Dos Santos, Moisture prole measurements of concrete samples
in vertical water ow by gamma ray transmission method, Radiat. Phys. Chem. 61
(36) (2001) 567569.
[32] G. Villain, M. Thiry, Gammadensimetry: a method to determine drying and carbonation proles in concrete, NDT E Int. 39 (4) (2006) 328337.
[33] G. Villain, M. Thiry, G. Platret, Measurement methods of carbonation proles in
concrete: thermogravimetry, chemical analysis and gammadensimetry, Cem. Concr.
Res. 37 (2007) 11821192.
[34] M. Thiry, P. Dangla, P. Belin, G. Habert, N. Roussel, Carbonation kinetics of a bed of
recycled concrete aggregates: a laboratory study on model materials, Cem. Concr.
Res. 46 (2013) 5065.
[35] A. Morandeau, Carbonatation atmosphrique des systmes cimentaires a faible
teneur en portlandite(PhD thesis) Universit Paris Est, Ifsttar, 2013.
[36] C. Gall, Effect of drying on cement-based materials pore structure as identied by
mercury intrusion porosimetry: a comparative study between oven-, vacuum-,
and freeze-drying, Cem. Concr. Res. 31 (10) (2001) 14671477.
[37] N.C. Collier, J.H. Sharp, N.B. Milestone, J. Hill, I.H. Godfrey, The inuence of water
removal techniques on the composition and microstructure of hardened cement
pastes, Cem. Concr. Res. 38 (6) (2008) 737744.
[38] S. Diamond, Mercury porosimetry an inappropriate method for the measurement
of pore size distributions in cement-based materials, Cem. Concr. Res. 30 (2000)
15171525.
[39] S. Diamond, A discussion of the paper effect of drying on cement-based materials
pore structure as identied by mercury porosimetrya comparative study between
oven-, vacuum-, and freeze-drying by C Gall, Cem. Concr. Res. 33 (1) (2003)
169170.
[40] C. Gall, Reply to the discussion by s. diamond of the paper effect of drying on
cement-based materials pore structure as identied by mercury porosimetrya
comparative study between oven-, vacuum-, and freeze-drying, Cem. Concr. Res.
33 (1) (2003) 171172.
[41] G. Arliguie, H. Hornain, Grandeurs associes la durabilit des btons GranDuB,
Presses Ecole Nationale Ponts et Chausses2007.
[42] P. Chindaprasirt, C. Jaturapitakkul, T. Sinsiri, Effect of y ash neness on microstructure of blended cement paste, Constr. Build. Mater. 21 (7) (2007) 15341541.

[43] V.T. Ngala, C.L. Page, Effects of carbonation on pore structure and diffusional properties of hydrated cement pastes, Cem. Concr. Res. 27 (7) (1997) 9951007.
[44] M. Castellote, C. Andrade, X. Turrillas, J. Campo, G.J. Cuello, Accelerated carbonation
of cement pastes in situ monitored by neutron diffraction, Cem. Concr. Res. 38 (12)
(2008) 13651373.
[45] M. Thiry, P. Faure, A. Morandeau, G. Platret, J.-F. Bouteloup, P. Dangla, V. BaroghelBouny, Effect of Carbonation on the Microstructure and Moisture Properties of
Cement-Based Materials, DBMC, 2011.
[46] M. Thiry, V. Baroghel-Bouny, A. Morandeau, P. Dangla, Impact of Carbonation on
the Microstructure and Transfer Properties of Cement-Based Materials, In Transfert,
Lille, 2012.
[47] Y.F. Houst, F.H. Wittmann, Inuence of porosity and water content on the diffusivity of
CO2 and O2 through hydrated cement paste, Cem. Concr. Res. 24 (6) (1994) 11651176.
[48] S. Lammertijn, N. De Belie, Porosity, gas permeability, carbonation and their interaction in high-volume y ash concrete, Mag. Concr. Res. 60 (7) (2008) 535545.
[49] P. Turcry, L. Oksri-Nela, A. Younsi, A. At-Mokhtar, Analysis of an accelerated
carbonation test with severe preconditioning, Cem. Concr. Res. 57 (2014) 7078.
[50] I. Galan, C. Andrade, M. Castellote, Thermogravimetrical analysis for monitoring
carbonation of cementitious materials, J. Therm. Anal. Calorim. 110 (1) (2012)
309319.
[51] M.D. Nguyen, Modlisation des couplages entre hydratation et dessiccation des
matriaux cimentaires l'issue du dcoffrage, Etude de la dgradation des
proprits de transfert, ENPC, Universit Paris Est, 2009. (PhD thesis).
[52] V. Papadakis, Experimental investigation and theoretical modeling of silica fume
activity in concrete, Cem. Concr. Res. 29 (1) (January 1999) 7986.
[53] J.H.M. Visser, Inuence of the carbon dioxide concentration on the resistance to
carbonation of concrete, Constr. Build. Mater. 67 (2013) 813 Part A.
[54] P.D. Tennis, H.M. Jennings, A model for two types of calcium silicate hydrate in the
microstructure of Portland cement pastes, Cem. Concr. Res. 30 (6) (2000) 855863.
[55] L. Black, C. Breen, J. Yarwood, K. Garbev, P. Stemmermann, B. Gasharova, Structural
features of CSH (I) and its carbonation in air a raman spectroscopic study. Part
II: carbonated phases, J. Am. Ceram. Soc. 90 (3) (2007) 908917.
[56] L. Black, K. Garbev, I. Gee, Surface carbonation of synthetic CSH samples: a comparison between fresh and aged CSH using X-ray photoelectron spectroscopy,
Cem. Concr. Res. 38 (6) (2008) 745750.
[57] Z. Sauman, V. Lach, Long term carbonation of the phases 3CaOAl2O36H2O and
3CaOAl2O3SiO24H2O, Cem. Concr. Res. 2 (4) (1972) 435446.
[58] T. Nshikawa, K. Suzuki, S. Ito, K. Sato, T. Takebe, Decomposition of synthesized
ettringite by carbonation, Cem. Concr. Res. 22 (1) (1992) 614.
[59] G.W. Groves, D.I. Rodway, I.G. Richardson, The carbonation of hardened cement
pastes, Adv. Cem. Res. 3 (11) (1990) 117125.
[60] I.G. Richardson, Tobermorite/jennite and tobermorite/calcium hydroxide-based
models for the structure of CSH: applicability to hardened pastes of tricalcium
silicate, -dicalcium silicate, Portland cement, and blends of Portland cement
with blast-furnace slag, metakaolin, or silica fume, Cem. Concr. Res. 34 (2004)
17331777.
[61] N. Hyvert, Application de l'approche probabiliste la durabilit des produits
prfabriqus en bton(PhD thesis) Universit Paul Sabatier, 2009.
[62] M. Castellote, L. Fernandez, C. Andrade, C. Alonso, Chemical changes and phase analysis of OPC pastes carbonated at different CO2 concentrations, Mater. Struct. 42 (4)
(2009) 515525.