M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
679–686
CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT
PORE SOLUTIONS
KOROZIJSKA ODPORNOST JEKLA V PORNIH VODAH IZ
ME[ANIH CEMENTOV
Miha Hren
1*
, Tadeja Kosec
1
, Andra` Legat
1
, Violeta Bokan-Bosiljkov
2
1
Slovenian National Building and Civil Engineering Institute, Dimi~eva ulica 12, 1000 Ljubljana, Slovenia
2
Faculty of Civil and Geodetic Engineering, Jamova cesta 2, 1000 Ljubljana, Slovenia
Prejem rokopisa – received: 2019-01-03; sprejem za objavo – accepted for publication: 2019-04-04
doi:10.17222/mit.2019.003
Blended cements might change the chemistry of the pore solution and subsequently affect the corrosion of steel in concrete.
Pore solutions were extracted, analyzed and compared from mortars made of CEM I, CEM II, CEM III and CEM IV cements.
Three combinations of carbonation and chloride states were studied, i.e., non-carbonated without chlorides, non-carbonated
with chlorides and carbonated with chlorides. Different electrochemical and spectroscopic techniques were used to study the
electrochemical properties, the type and the extent of the corrosion products, as well as the type and the extent of the corrosion
damage. It was confirmed that the most corrosive environments were pore solutions extracted from the carbonated mortars with
chlorides. In this environment the highest corrosion rate was observed for the CEM III pore solution, and the lowest for the
CEM I. The extent and the type of corrosion products and the corrosion damage varied according to the environment.
Keywords: corrosion, blended cements, pore solution, mortar
Cementi z mineralnimi dodatki lahko spremenijo strukturo porne vode in posledi~no vplivajo na korozijske lastnosti jekla.
Porne vode so bile iztisnjene in analizirane iz malt, narejenih iz CEM I, CEM II, CEM III in CEM IV cementov. Malte so bile
predhodno izpostavljene kloridom ali pospe{eni karbonatizaciji v treh kombinacijah: brez karbonatizacije in brez kloridov, brez
karbonatizacije s kloridi in v karbonatizaciji s kloridi. Za ugotavljanje korozijskih lastnosti jekla v pornih vodah so bile
uporabljene razli~ne elektrokemijske in spektroskopske metode. Analizirane so bile elektrokemijske lastnosti, tip in obseg
korozijskih produktov ter tip in obseg korozijskih po{kodb. Ugotovljeno je bilo, da najbolj korozivno okolje pripada pornim
vodam v karbonatizirani malti s kloridi, kjer je bila najvi{ja korozijska hitrost izmerjena v pornih vodah iz CEM III cementa,
najmanj{a pa iz CEM I cementa. Obseg in tip korozijskih produktov ter po{kodb se je razlikoval skladno s korozivnostjo okolja.
Klju~ne besede: korozija, me{ani cementi, porna voda, malta
1 INTRODUCTION
The corrosion of rebar has been a longstanding dura-
bility issue in concrete. The penetration of the chloride
ions and the reduction in the alkalinity due to carbon-
ation can destabilize the passive film that protects the
steel from corrosion. The use of certain mineral admix-
tures in blended cements affects both corrosion-propaga-
tion processes. These mineral admixtures primarily re-
fine the concrete pores, which is a well-known beneficial
effect that can slow down the chloride penetration and
the carbonation progression.
1,2
The same reactions that
refine the pores also reduce the alkalinity
3
and affect the
chloride binding,
4
thus creating a different and unknown
corrosive environment.
Multiple corrosion studies have been conducted for
mortars or concretes made of blended cements.
5–14
Most
of them focus on the additions of silica fume, fly ash and
slag, with half-cell potential and linear polarization tech-
niques being mostly used to determine the corrosion ac-
tivity in the concrete or mortar. All the studies report on
blended cements increasing the corrosion resistance due
to the latent pozzolanic reaction. However, not a lot of
research focuses on aged concrete that has undergone
carbonation, nor are there many examples of the corro-
sion rates being measured directly. There are many fewer
corrosion studies available in concrete pore solutions ex-
tracted from blended cements. Most pore-solution exper-
iments involved the use of saturated calcium hydroxide,
with optional additions of potassium and sodium hydrox-
ide.
15–23
Electrochemical impedance spectroscopy and
potentiodynamic polarization are the most common tech-
niques used in these experiments. While chloride con-
centration and alkalinity are the two most important fac-
tors influencing steel corrosion, other ions found in the
pore solution should not be overlooked.
24
So far, there is
no information in the literature as to what extent the
pore-solution’s composition influences the corrosion of
steel in blended cements.
In this paper the effect of the presence of chloride
and carbonation was studied on the corrosion behavior of
carbon steel in pore solutions extracted from mortars
made of blended cements containing different amounts
of fly ash, natural pozzolana and slag. Multiple electro-
chemical techniques, including corrosion-potential mea-
surements, linear polarization and potentiodynamic
Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686 679
UDK 620.193:691.714.018.8:666.9.022.6 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 53(5)679(2019)
*Corresponding author's e-mail:
miha.hren@zag.si

scans, were used to study the corrosion properties. SEM,
optical microscopy and Raman spectroscopy were em-
ployed to analyze the corrosion products formed in such
media.
2 MATERIALS AND METHODS
Mortar prisms, 100 mm × 60 mm × 30 mm in size,
were made to extract and analyze their pore solutions.
Each prism had a water-cement ratio of 0.75 and a ce-
ment-aggregate ratio of 0.33. Standard sand, as described
in EN 196-1:2016, was used and mixed with one of four
different cements: CEM I 42.5 N, CEM II/B-M (LL-V)
42.5 N, CEM III/B (S) 32.5 N – LH/SR and CEM IV/A
(V-P) 42.5 R SR. These cements differ in terms of the
compressive strength and the admixtures used. Mineral
admixtures are designated by the labels in parentheses,
with some cements having multiple mineral admixtures.
Label V is used for Fly ash, S is used for slag, LL is used
for limestone and P is used for natural pozzolans. The
cements were provided by a local cement factory, Salonit
Anhovo.
All the specimens were cured for 28 d in a humidity
chamber and one specimen of each cement also under-
went accelerated carbonation for 10 w. These specimens
are referred to in the text as carbonated, while the natu-
rally carbonated specimens are referred to as non-car-
bonated. External chlorides were added through a
35-week cyclic ponding of 3.5 % sodium chloride solu-
tion, where the wetting period lasted for 3 d and the dry-
ing period lasted for 4 d. Reference specimens of each
cement were also made to obtain pore water of
uncarbonized mortars without external chlorides imme-
diately after 28 d of curing in a humidity chamber.
The pore solution was extracted from the specimens
using an extraction technique described in the litera-
ture.
25
A device was used that produces compressive
pressure on the specimen in the range of 500 MPa to
1000 MPa and allowed the solution to be drained
through a filter. A few ml of solution was extracted, di-
luted and analyzed using ion chromatography. Based on
these results, a total of 12 synthetic pore solutions were
prepared, where the cement type, chlorides and carbon-
ation state were varied.
All the pore solutions were used as an electrolyte in 3
consecutive electrochemical experiments, with each set
of tests repeated at least 3 times with a fresh working
electrode and solution. The open-circuit potential (OCP)
was measured first for about 22 h, until the potential of
the working electrode stabilized. The linear polarization
was measured at ±20 mV vs. OCP at a scan rate of
0.1 mV/s. The cyclic polarization followed from
–250 mV vs OCP to 1.2 V above the reference potential,
or until the current density reached 10 mA/cm
2
. The scan
rate was set at 1 mV/s. Tafel fitting was done in Gamry
Echem Analyst software to obtain the corrosion poten-
tial, the corrosion-current density and both Tafel slopes.
In order to obtain the corrosion rates, the corrosion-cur-
rent density was first calculated from the linear
polarization using the Stern-Geary Equation (1). Anodic
( a
) and cathodic ( c
) Tafel slopes were acquired with the
cyclic polarization. The corrosion rates were consecu-
tively calculated from the corrosion-current density us-
ing Equation (2), where a is the atomic weight of iron
(55.8 g/mol), n is the oxidation number of iron (2+) and
 is the density of the steel (7.9 g/cm
3
).
i
R
corr
ac
pa c
=
⋅
⋅+

 2303 .( )
[μm/cm
2
] (1)
CR
ai
n
=
⋅
⋅
327 .
corr
 [μm/year] (2)
Electrochemical tests were done in Gamry Frame-
works using the Gamry Reference 600 potentiostat. All
the experiments were performed in a standard three-elec-
trode cell, using a saturated calomel reference electrode
(SCE), graphite counter electrode and a working elec-
trode made of cold-rolled carbon-steel sheet with an ex-
posed surface area of 1 cm
2
. All the potentials in the text
refer to the SCE potential. Prior to exposure, the steel
surface was degreased with acetone and grinded using
600-grit, 1200-grit and 2500-grit papers. Immediately af-
ter the 24-hour electrochemical measurements, the elec-
trodes were rinsed with distilled water, dried and exam-
ined under a Tagarno HD microscope. A low-vacuum
JEOL 5500 LV SEM using an accelerating voltage of 20
kV and Horiba Yvon HR800 RAMAN spectrometer with
a laser at = 632 nm and scanning range 50 cm
–1
to 1000
cm
–1
were also used to provide additional information
about the surface of the steel and the corrosion products.
3 RESULTS AND DISCUSSION
3.1 Pore water extraction
The pore solutions were prepared according to the re-
sults of the mortar extraction (Table 1). Components
were chosen so that the pH, chloride and sulfate concen-
trations (not shown) matched the values of the extracted
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
680 Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686
Table 1: Extracted pore solutions. Ion concentrations are in g/L
CEMENT AC
1
Cl
2
pH Cl
–
[Cl
–
]/ [OH
–
]
CEM I
13.06 0.07 0.017
X 12.05 61.04 153
X X 10.45 82.83 8.29×10
3
CEM II
13.03 0.10 0.026
X 11.81 84.86 371
X X 7.52 106.37 9.06×10
6
CEM III
12.51 0.40 0.349
X 11.45 47.99 480
X X 7.54 117.08 9.53×10
6
CEM IV
13.01 0.06 0.017
X 12.06 74.91 184
X X 9.83 109.99 4.59×10
4
1 – Accelerated Carbonation, 2 – External chlorides

solutions, while the cations were chosen to reflect the so-
lutions as close as possible, due to solubility limitations.
All the pH values were measured with a Mettler Toledo
MP220 pH meter before and after exposure. They
remined stable throughout the electrochemical experi-
ments.
The pH of the untreated mortars, prepared by using
different blended cements (CEM II, CEM III and CEM
IV) and standard cement (CEM I) was as high as 13. The
pore waters from the mortars that were exposed to cyclic
ponding with chlorides contained around 48 g/L to 85
g/L chlorides and the pH was reduced to about 12, while
the [Cl
–
]/[OH
–
] ratio was between 150 and 480. The ac-
celerated carbonation and cyclic wetting with chlorides
resulted in a very high [Cl
–
]/[OH
–
] ratio and a substantial
decrease of the pH. Carbonation had the least effect on
the CEM I pore solution’s alkalinity (around pH 10), fol-
lowed by CEM IV, CEM II and CEM III (around pH 7).
The CEM III cement had the least amount of lime in its
composition, which could be the reason for its lower al-
kalinity in all the exposed conditions (Table 1). All the
cements showed a much higher chloride concentration
after the accelerated carbonation. Although there are not
many publications about carbonation and chloride bind-
ing, there seems to be a consensus that carbonation re-
leases bound chlorides due to the reduction of pores for
physical binding and the dissolution of Friedel’s salt due
to the lowered alkalinity.
4
The latter would explain the
higher concentration of chlorides in all the carbonated
blended cements, especially CEM III. When naturally
carbonated, CEM II had the largest amount of chlorides
in its pore solution, followed by CEM IV, CEM I and
CEM III. The chloride-binding capability of a cement is
a combination of the alumina content, the C/S ratio, the
level of carbonation, the sodium ions present and other
factors.
4
If one parameter does not stand out, a combina-
tion of all the factors influences the concentration of the
bound chlorides. It is hard to determine to what degree
each factor contributes to the chloride release.
3.2 Open-circuit potential
The open-circuit potential on steel in the pore solu-
tions was measured for 22 h. In all the pore waters from
the non-carbonated cement without chlorides, the poten-
tial increased over time, indicating that the high pH val-
ues made the steel more corrosion resistant compared to
its in-air state. In some cases, the potential experienced
more than a 100-mV drop after being stable for many
hours. This may be due to the breakdown processes of
the passive film formed in air, as a new, more stable pas-
sive film in alkaline solution forms. This process is
knowntotakeupto3dinalkaline solutions without ag-
gressive ions.
16
Similar trends are rarely observed for ei-
ther non-carbonated or carbonated pore water with chlo-
rides, as the potential decreased with time and remained
relatively stable after a couple hours.
The mean OCP error bar chart is shown in Figure 1
at the end of the exposure for pore waters of all 4 ce-
ments in the non-carbonated state without chlorides, the
non-carbonated state with chlorides and the carbonated
state with chlorides. Large differences were observed
when comparing the corrosion potentials of the steel in
cements in different chloride and carbonation states. The
highest potentials were found in the non-carbonated pore
water without chlorides, with a potential drop of around
350 mV observed for all the cements as the chlorides
were added. The lowest potentials were measured in car-
bonated pore water with chlorides. These were between
–700 mV and –620 mV.
The differences in the corrosion potentials of the steel
in different cement types under the same mortar condi-
tions were less expressed, but still differentiative. In the
non-carbonated state without additional chlorides, the re-
sults cannot be correlated to either chloride concentra-
tions or alkalinity. However, the highest concentrations
of potassium, calcium, magnesium and sulfate ions can
be found in the CEM I cement, which could explain its
lower and less-stable potential.
19
In the non-carbonated
state with additional chlorides, the corrosion potential
was the most positive for CEM I, followed by CEM IV
and CEM II, while for CEM II it was the lowest
(–570 mV). Results in the non-carbonated state with
chlorides were in line with the [Cl]/[OH] ratio (Table 1),
where a higher ratio is reflected in a lower potential. In
the carbonated pore solutions with chlorides, CEM I
stood out with its lowest chloride content, highest pH
and the resulting highest potential (–640 mV, Figure 1).
3.3 Polarization resistance
The results of the polarization resistance (R
p
) mea-
surements for 4 cements in 3 carbonation and chloride
states are presented in Figure 2. The values of R
p
are
shown on a logarithmic scale. The polarization resistance
is often translated to the corrosion rate through the
Stern-Geary Equation (1) and Equation (2), where a
higher polarization resistance results in a lower corrosion
rate.
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686 681
Figure 1: Open circuit potential values for all cements in all exposure
states

Large differences in the polarization resistance can be
seen between the different chloride and carbonation
states. In the non-carbonated pore solution without chlo-
rides, the polarization resistances were an order of mag-
nitude higher for all the cements compared to R
p
in the
non-carbonated pore solution with chlorides. Thus, the
corrosion rates in the non-carbonated pore solutions with
chlorides were 10-times higher than for the steel in the
non-carbonated pore solution without chlorides. The po-
larization resistance values between 1.5 k cm
2
and 5
k cm
2
in carbonated pore solutions with chlorides were
2.5-times lower than in the non-carbonated pore solu-
tions with chlorides.
In the non-carbonated state without external chlo-
rides, CEM III had a lower polarization resistance com-
pared to the other cements, most likely due to the higher
chloride concentration and the lower pH (Table 1). Al-
most no differences between the cements could be ob-
served in the non-carbonated state with chlorides, apart
from CEM II and CEM III having more scattered results.
The performance of the cements in the carbonated state
with chlorides was in line with the [Cl
–
]/[OH
–
] ratio,
where a higher ratio results in a lower resistance. The
CEM III cement had the lowest polarization resistance,
followed by CEM II, CEM IV and CEM I. In the carbon-
ated state with chlorides, the steel in CEM I had the most
positive R
p
(5 k cm
2
).
3.4 Cyclic polarization
Potentiodynamic scans for the steel in pore solutions
simulating the non-carbonated pore solution with and
without chlorides and the carbonated pore solution with
chlorides are presented in Figure 3. The corrosion-cur-
rent density j
corr
and the corrosion potential E
corr
were de-
duced from the polarization curves and both are pre-
sented in Table 2. Using Equation (2), the corrosion rate
v
corr
was calculated from the j
corr
values.
In the non-carbonized pore solution without chlo-
rides, the polarization curves showed good corrosion re-
sistance for the steel (Figure 3a). The corrosion poten-
tial was between –250 mV for CEM I and –130 mV for
CEM IV. The corrosion-current densities were similar at
1 μA/cm
2
, and the resulting corrosion rates did not ex-
ceed 20 μm/y (Table 2). CEM IV showed the best corro-
sion resistance with both the lowest corrosion current
and passive current densities (Figure 3a), followed by
CEM I, CEM II and CEM III. The steel remained passive
in all the pore solutions and the breakdown potential (E
b
)
was observed at roughly +600 mV vs. SCE, while E
b
for
CEM III was slightly more positive. During the reverse
polarization cycle, the currents returned towards zero,
following the same path as during anodic polarization.
The shape of the potentiodynamic curves showed that no
local type of corrosion is expected in such conditions.
In the non-carbonized pore solution with chlorides
(Figure 3b), the corrosion potentials of the working
electrodes were more negative due to the presence of
chlorides. The values ranged between –650 mV and
–500 mV. In terms of corrosion-current densities and the
corresponding corrosion rates, CEM I, II, III and IV
showed a similar average value of 3.5 μA/cm
2
(40 μm/y).
The breakdown potential for all 4 cement pore solutions
was reached at about –300 mV and the repassivation pro-
cess required much lower potentials for the current den-
sity to reach values in the passive state. This is common
for localized, chloride-induced corrosion. After the re-
verse polarization cycle, the potentials decreased by
about 100 mV, on average.
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
682 Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686
Figure 3: Cyclic polarization curves of select specimens in: a) non-
carbonated pore water without chlorides, b) non-carbonated pore wa-
ter with chlorides and c) carbonated pore water with chlorides
Figure 2: Polarization resistance values for all cements in all exposure
states

CEM IV and CEM I exhibited lower passive current
densities compared to CEM II and CEM III (Figure 3b).
These results correlate well with the [Cl
–
]/[OH
–
] ratio, as
both CEM II and CEM III have an about 2–3 times
higher ratio compared to CEM I and CEM IV.
In carbonated pore water with chlorides, the steel un-
derwent general corrosion. In the polarization curves
(Figure 3c) this was indicated by a lack of a passive re-
gion as the current density was increasing with the po-
tential in the anodic direction. The corrosion rate was the
smallest for CEM I and the highest for CEM II and CEM
III (Figure 3c, Table 2). The specimens made of pore
water from CEM II, III and IV showed little difference in
terms of the average current densities and corrosion rates
(Table 2, values around 9 μA/cm
2
and 100 μm/y), with
CEM IV having less-scattered results and a lower maxi-
mum expected corrosion rate as a result. The CEM I pore
water stood out with a noticeably lower average and
maximum expected corrosion rate (70±10 μm/y), which
is the result of its lower chloride content and higher alka-
linity. Its corrosion potential was also higher by about 60
mV, in line with the OPC results. The optical microscopy
and SEM images in Figure 5a and 5b of the exposed
working electrode in the carbonated pore solution with
chlorides showed large areas of uniformly formed corro-
sion products. The size of the corroded area was larger
when the pH was lower.
3.5 Optical microscopy, SEM and Raman
With the aim to follow the development of the corro-
sion products, the steel specimens were immersed in
three different simulated environments (non-carbonated,
non-carbonated with chlorides and carbonated with chlo-
rides) of CEM I, CEM II, CEM III and CEM IV pore so-
lutions. Only the steel specimens in the most corrosive
CEM III environment are presented in Figure 4. The
non-carbonated state (pH 12.54) of the CEM III pore so-
lution was not aggressive and was observed to have no
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686 683
Table 2: Corrosion potential, corrosion current and corrosion rate for all cements obtained from cyclic polarization
CEMENT E
corr
[mV] j
corr
[μA/cm
2
] v
corr
[μm/y]
non-carbonated
no chlorides
CEM I –250 ± 50 1.0 ± 0.5 11 ± 5
CEM II –190 ± 30 1.2 ± 0.5 14 ± 5
CEM III –160 ± 40 1.7 ± 0.3 19 ± 4
CEM IV –170 ± 20 0.8 ± 0.3 9 ± 3
non-carbonated
chlorides
CEM I –590 ± 20 3.0 ± 2.0 40 ± 20
CEM II –580 ± 20 4.0 ± 1.0 40 ± 20
CEM III –590 ± 20 4.0 ± 1.0 40 ± 10
CEM IV –610 ± 40 3.0 ± 2.0 30 ± 20
Carbonated
chlorides
CEM I –646 ± 1 6.0 ± 1.0 70 ± 10
CEM II –719 ± 9 9.0 ± 5.0 100 ± 50
CEM III –710 ± 10 9.0 ± 5.0 110 ± 50
CEM IV –705 ± 8 9.0 ± 3.0 110 ± 30
Figure 4: Photographs of steel specimen immersed in pore waters extracted from CEM III mortars exposed to different carbonation and chloride
states: a) after 7 d, b) and c) after 18 d of exposure and d), e) and f) the same surfaces after cleaning of corrosion products in HCl with urotropine

visible signs of corrosion (Figure 4a). In the non-car-
bonated state with chlorides (pH 11.46), the corrosion
was already visible after 18 d of exposure (Figure 4b),
where approximately one-third of the steel surface was
covered in corrosion products. Almost the entire exposed
surface of the steel was covered in corrosion products
(Figure 4c) in the CEM III pore solution that was ex-
tracted from carbonated mortars soaked with chlorides
(pH 7.50). The corrosion damage is shallow and not lo-
calized (Figure 4e and 4f).
Optical microscopy and SEM were also used to ex-
amine the type of corrosion damage occurring on the
specimens in each cement pore solution. No corrosion
damage was detected on any of the specimens exposed to
non-carbonated pore solutions without chlorides. In the
non-carbonated pore solutions with chlorides, large iso-
lated pits (CEM I and CEM II) and areas of localized
corrosion along the edges of the specimen (CEM III and
CEM IV) were detected. Two pits for CEM I and CEM II
are shown in Figure 5c and 5d, respectively. The extent
and type of damage varied greatly between identical
specimens in non-carbonated cements with chlorides, so
no generalizations could be made. In the carbonated pore
solutions with chlorides, general corrosion was observed
for all the cements. Roughly 75 % and 35 % of the ex-
posed area was corroded for the CEM I and CEM IV
specimens, respectively. In the case of CEM II and CEM
III, almost the entire exposed area was corroded (Figure
5a), which is in line with the lower pH values of these
pore solutions.
SEM and Raman spectroscopy were used to study the
morphology and the type of corrosion products present.
Similar corrosion products were found across multiple
cements’ pore solutions, so only the results from CEM
III were evaluated in detail. The Raman spectra of the
corrosion products formed on steel were compared to the
Raman spectra reported in the literature.
26
On a specimen
exposed to a non-carbonized pore solution with chlo-
rides, strong hematite and weak goethite spectra were
measured (Figure 6a, 6b and 6c). Hematite was charac-
terized by two strong bands at 221 cm
–1
and 286 cm
–1
,
one moderate band at 398 cm
–1
and two weaker bands at
484 cm
–1
and 600 cm
–1
. The band around 240 cm
–1
was
not as pronounced as in the literature. Goethite showed a
moderate band at 674 cm
–1
and two very weak bands at
238 cm
–1
and 387 cm
–1
. Both the detected corrosion
products are considered stable, adhere to the steel sur-
face and slow down the corrosion reaction.
18
The steel specimens in the carbonized pore solution
with chlorides predominantly contained lepidocrocite on
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
684 Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686
Figure 6: SEM images and Raman spectra of corrosion products
found in: a), b) and c) non-carbonated CEM III pore solution with
chlorides, d) and e) carbonated CEM III pore solution with chlorides
Figure 5: Optical and SEM images after7dofe xposure: a) and b)
steel surface in carbonated CEM III pore solution with chlorides, c) pit
in non-carbonated CEM I pore solution with chlorides and d) pit in
non-carbonated CEM II pore solution with chlorides

their surface (Figure 6d and 6e). Lepidocrocite was
characterized by two strong bands at 251 cm
–1
and 381
cm
–1
, and three moderate bands at 530 cm
–1
, 649 cm
–1
and 713 cm
–1
. Unlike goethite and hematite, this phase is
unstable and porous.
18
With time, it tends to dissolve and
transform to other forms of oxides, thus offering weaker
corrosion protection.
4 CONCLUSIONS
Pore waters were extracted from 4 cements exposed
to a combination of carbonation and chlorides for 35 w.
The following conclusions can be drawn.
Large differences in the corrosion rates, corrosion po-
tentials, corrosion damage and corrosion products were
observed when comparing different carbonation and
chloride states of the pore solutions. The influence of
these states on the corrosion conditions was significantly
higher than that of the different cements.
When comparing different cement types (CEM I,
CEM II, CEM III and CEM IV), it could be observed
that the electrochemical properties varied, to some ex-
tent. However, these variations could be more significant
when taking different degrees of corrosion localization
into account.
In the non-carbonated state without chlorides the
steel in the pore solution from all the cements showed
passive behavior, with corrosion current densities as low
as 1 μA/cm
2
. The post-exposure microscopic observa-
tions confirmed these results.
The corrosion damage in all the pore solutions from
the carbonated cements with chlorides was roughly gen-
eral, whereas the damage in the pore solutions from
non-carbonated cements with chlorides was both local-
ized and general.
Electrochemical techniques generally provided the
lowest corrosion rates for CEM I, with the corrosion
rates for CEM IV in non-carbonated conditions being
somehow comparable to CEM I.
The highest corrosion rates overall were measured for
CEM III. It should be mentioned, however, that the cor-
rosion damage in both the carbonated and non-carbon-
ated states with chlorides were roughly general.
Spectroscopic analyses of the corrosion products
confirmed the electrochemical results. The Raman analy-
sis indicated a protective corrosion layer in the non-car-
bonated chloride environment and non-protective corro-
sion products in the more corrosive carbonated chloride
environment.
Acknowledgment
The financial support received from the Slovenian
Research Agency (SRA) during 2015–2017, within the
scope of the Young Scientist program, is hereby grate-
fully acknowledged.
5 REFERENCES
1
P. K. Mehta, P. J. M. Monteiro, Concrete: microstructure, properties,
and materials, McGraw-Hill, New York 2006, 660
2
J. Bijen, Benefits of slag and fly ash, Constr. Build. Mater., 10 (1996)
5, 309–314, doi:10.1016/0950-0618(95)00014-3
3
A. Vollpracht, B. Lothenbach, R. Snellings, J. Haufe, The pore solu-
tion of blended cements: a review, Mater. Struct., 49 (2016)8 ,
3341–3367, doi:10.1617/s11527-015-0724-1
4
Q. Yuan, C. Shi, G. De Schutter, K. Audenaert, D. Deng, Chloride
binding of cement-based materials subjected to external chloride en-
vironment – A review, Constr. Build. Mater., 23 (2009) 1, 1–13,
doi:10.1016/j.conbuildmat.2008.02.004
5
C. L. Page, N. R. Short, W. R. Holden, The influence of different ce-
ments on chloride-induced corrosion of reinforcing steel, Cem.
Concr. Res., 16 (1986) 1, 79–86, doi:10.1016/0008-8846(86)90071-2
6
M. Maslehuddin, H. Saricimen, A. I Al-Mana, Effect of fly ash addi-
tion on the corrosion resisting characteristics of concrete, ACI Mater.
J., 84 (1987), 42-50
7
P. S. Mangat, B. T. Molloy, Influence of PFA, slag and microsilica on
chloride induced corrosion of reinforcement in concrete, Cem.
Concr. Res., 21 (1991) 5, 819–834, doi:10.1016/0008-8846(91)
90177-J
8
P. S. Mangat, J. M. Khatib, B. T. Molloy, Microstructure, chloride
diffusion and reinforcement corrosion in blended cement paste and
concrete, Cem. Concr. Compos., 16 (1994) 2, 73–81, doi:10.1016/
0958-9465(94)90002-7
9
R. B. Polder, W. H. A. Peelen, Characterisation of chloride transport
and reinforcement corrosion in concrete under cyclic wetting and
drying by electrical resistivity, Cem. Concr. Compos., 24 (2002)5,
427–435, doi:10.1016/S0958-9465(01)00074-9
10
V. Saraswathy, S. Muralidharan, K. Thangavel, S. Srinivasan, Influ-
ence of activated fly ash on corrosion-resistance and strength of con-
crete, Cem. Concr. Compos., 25 (2003) 7, 673–680, doi:10.1016/
S0958-9465(02)00068-9
11
E. Güneyisi, T. Özturan, M. Gesoðlu, A study on reinforcement cor-
rosion and related properties of plain and blended cement concretes
under different curing conditions, Cem. Concr. Compos., 27 (2005)
4, 449–461, doi:10.1016/j.cemconcomp.2004.05.006
12
M. A. Ibrahim, A. E.-D. M. Sharkawi, M. M. El-Attar, O. A.
Hodhod, Assessing the corrosion performance for concrete mixtures
made of blended cements, Constr. Build. Mater., 168 (2018), 21–30,
doi:10.1016/j.conbuildmat.2018.02.089
13
N. Chousidis, I. Ioannou, E. Rakanta, C. Koutsodontis, G. Batis, Ef-
fect of fly ash chemical composition on the reinforcement corrosion,
thermal diffusion and strength of blended cement concretes, Constr.
Build. Mater., 126 (2016), 86–97, doi:10.1016/j.conbuildmat.2016.
09.024
14
A. N. Borade, B. Kondraivendhan, Effect of metakaolin and slag
blended cement on corrosion behaviour of concrete, IOP Conf. Ser.
Mater. Sci. Eng., 216 (2017) 1, 1-6, doi:10.1088/1757-899X/216/
1/012022
15
R. Liu, L. Jiang, J. Xu, C. Xiong, Z. Song, Influence of carbonation
on chloride-induced reinforcement corrosion in simulated concrete
pore solutions, Constr. Build. Mater., 56 (2014), 16–20,
doi:10.1016/j.conbuildmat.2014.01.030
16
A. Poursaee, C. M. Hansson, Reinforcing steel passivation in mortar
and pore solution, Cem. Concr. Res., 37 (2007) 7, 1127–1133,
doi:10.1016/j.cemconres.2007.04.005
17
W. Chen, R.-G. Du, C.-Q. Ye, Y.-F. Zhu, C.-J. Lin, Study on the cor-
rosion behavior of reinforcing steel in simulated concrete pore solu-
tions using in situ Raman spectroscopy assisted by electrochemical
techniques, Electrochimica Acta., 55 (2010) 20, 5677–5682,
doi:10.1016/j.electacta.2010.05.003
18
J. K. Singh, D. D. N. Singh, The nature of rusts and corrosion char-
acteristics of low alloy and plain carbon steels in three kinds of con-
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686 685

crete pore solution with salinity and different pH, Corros. Sci. 56
(2012) 129–142. doi:10.1016/j.corsci.2011.11.012.
19
P. Ghods, O. B. Isgor, G. McRae, T. Miller, The effect of concrete
pore solution composition on the quality of passive oxide films on
black steel reinforcement, Cem. Concr. Compos. 31 (2009) 2–11.
doi:10.1016/j.cemconcomp.2008.10.003.
20
M. Moreno, W. Morris, M. G. Alvarez, G. S. Duffó, Corrosion of re-
inforcing steel in simulated concrete pore solutions: Effect of carbon-
ation and chloride content, Corros. Sci., 46 (2004), 2681–2699,
doi:10.1016/j.corsci.2004.03.013
21
F. Zhang, J. Pan, C. Lin, Localized corrosion behaviour of reinforce-
ment steel in simulated concrete pore solution, Corros. Sci., 51
(2009) 9, 2130–2138, doi:10.1016/j.corsci.2009.05.044
22
A. ^esen, T. Kosec, A. Legat, V. Bokan-Bosiljkov, Corrosion proper-
ties of different forms of carbon steel in simulated concrete pore wa-
ter, Mater. Tehnol., 48 (2014) 1, 51–57
23
G. Liu, Y. Zhang, Z. Ni, R. Huang, Corrosion behavior of steel sub-
mitted to chloride and sulphate ions in simulated concrete pore solu-
tion, Constr. Build. Mater., 115 (2016), 1–5, doi:10.1016/
j.conbuildmat.2016.03.213
24
A. Poursaee, C. M. Hansson, Potential pitfalls in assessing chlo-
ride-induced corrosion of steel in concrete, Cem. Concr. Res., 39
(2009) 5, 391–400, doi:10.1016/j.cemconres.2009.01.015
25
M. Cyr, P. Rivard, F. Labrecque, A. Daidi}, High-pressure device for
fluid extraction from porous materials: Application to cement-based
materials, J. Am. Ceram. Soc., 91 (2008) 8, 2653–2658,
doi:10.1111/j.1551-2916.2008.02525.x
26
F. Dubois, C. Mendibide, T. Pagnier, F. Perrard, C. Duret, Raman
mapping of corrosion products formed onto spring steels during salt
spray experiments. A correlation between the scale composition and
the corrosion resistance, Corros. Sci., 50 (2008) 12, 3401–3409,
doi:10.1016/j.corsci.2008.09.027
M. HREN et al.: CORROSION PERFORMANCE OF STEEL IN BLENDED CEMENT PORE SOLUTIONS
686 Materiali in tehnologije / Materials and technology 53 (2019) 5, 679–686