Construction and Building Materials 145 (2017) 435–444

Contents lists available at ScienceDirect

Construction and Building Materials

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

Electrochemical chloride extraction: Efficiency and impact on concrete

containing 1% of NaCl
Luan. Rodrigo de Almeida Souza a, Marcelo Henrique Farias de Medeiros a, Eduardo Pereira b,⇑,
Ana Paula Brandão Capraro a
a
Department of Civil Engineering, Federal University of Paraná, Centro Politécnico, Jardim das Américas, CEP: 81531-980 Curitiba, PR, Brazil
b
Department of Civil Engineering, State University of Ponta Grossa, Avenida Carlos Cavalcanti, 4748, Uvaranas, CEP 84030900 – Ponta Grossa, PR, Brazil

h i g h l i g h t s

Concrete with higher w/c ratio has a tendency of having a higher efficiency for removal of chlorides.

In concretes with lower w/c ratio, the thickness of the reinforcement cover has a great influence on the corrosion potential.

Total porosity increases by more than 30% as a result of applying the chloride extraction.

Percentage growth of porosity due to chlorides extraction depends on the w/c ratio.

Concrete water absorption increased 13% and 12% for concrete with w/c ratios of 0.43 and 0.59, respectively.

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

Article history: This paper aims to evaluate the efficiency of the electrochemical chloride extraction technique and its
Received 28 June 2016 impact in concrete contaminated with 1% of chloride by monitoring chloride content, specific porosity,
Received in revised form 29 March 2017 water absorption and corrosion potential, verifying the influence of variables reinforcement cover (1.0
Accepted 3 April 2017
and 3.0 cm) and water/cement ratio (0.43 and 0.59). Results show that the concrete needs additional pro-
tection after chlorides extraction in order to prevent recontamination, since the porosity of the concrete
increases with the application of the electrochemical chloride extraction. Average pore diameter
Keywords:
increased 78% and 42% for concretes with w/c of 0.59 and 0.43, allowing the ingress of aggressive agents.
Chloride extraction
Corrosion
Ó 2017 Elsevier Ltd. All rights reserved.
Porosity
Chlorides
Corrosion potential

1. Introduction which are contaminating the structure, without the need of remov-
ing the concrete [5]. Compared to traditional techniques it has
The corrosion reinforcement, caused by the presence of chloride advantages such as little effect on the environment along with high
ions, is a major cause of the degradation of reinforced concrete efficiency [6].
structures [1–3]. The presence of chloride ions in the concrete The chlorides extraction technique consists of applying an elec-
structure affects the steel passivity layer and, in the presence of tric field between the reinforcement inside the concrete and an
water and oxygen, corrosion occurs. Prevention is still the cheapest electrode constituted by a metal mesh fixed externally to the sur-
measure to fight this deterioration mechanism, however once the face of the material. With the application of electric current nega-
structure is contaminated it is necessary to repair it in order to tively, charged ions such as chloride ions, are attracted to the
avoid additional future spending. A traditional repair technique anode placed on the external surface of the concrete. Also, cations
involves the removal of the affected area replacing the weakened (Na+) migrate to the reinforcement and there is production of
reinforcement and inserting a new repair material [4]. Alterna- hydroxyl ions (OH-) on the reinforcement surface, as a conse-
tively, there is the Electrochemical Chloride Extraction method quence of cathodic reactions [6,7]. Fig. 1 shows a schematic dia-
(ECE). This method consists in removing part of the chloride ions, gram of the chloride extraction technique.
The electrochemical extraction of chlorides and the electric
potential difference between the anode and the cathode cause
⇑ Corresponding author. the following reactions:
E-mail address: [email protected] (E. Pereira).

http://dx.doi.org/10.1016/j.conbuildmat.2017.04.010
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
436 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444

Fig. 1. Schematic design of chloride extraction technique (Adapted from [8]).

1 tory results. Abdelaziz et al. [24] evaluated intervals of 2, 4 and

2OH ¼ O2 þ H2 O þ 2e
2 8 weeks, with best efficiency found at 8 weeks. The author points
out, however, that the reduction rate of chloride content decreases
2H2 O ¼ O2 þ 4Hþ þ 4e along the weeks. Regarding this aspect, Elsener and Angst [12]
found that periods of rest may increase the efficiency of the tech-

2Cl ¼ Cl2 þ 2e nique. The authors attribute this behavior to the fact that in the
At the same time, hydroxyl ions are produced around the rein- first step the free chloride ions were removed, while the ions com-
forcement by electrolysis. bined with compounds of the hydrated paste dissolve more slowly.
According to Andrade [4], the rest period is justified since the alka-
1 lization produced during extraction induces a decrease in the
O2 þ H2 O þ 2e ¼ 2OH
2 migration flux of chloride ions.
Huang et al. [25] emphasize the importance of studying the pro-
2H2 O þ 2e ¼ H2 þ 2OH gresses of electrochemical chloride extraction due to the possible
For the technical operation, the presence of an electrolyte effect on the microstructure and performance of the material.
allowing the ion migration process is fundamental. The most com- The application of the chloride extraction technique can have con-
monly solutions used as external electrolytes are saturated calcium sequences on the porosity of the concrete, on the activation of
hydroxide, sodium hydroxide and treated water. From those solu- alkali-aggregate reactions, on the reduction of the steel-concrete
tions, water is the most widely used. According to Polder [9] and adhesion and on steel embrittlement.
Elsener, Angst [10], its widespread use is due to the absence of According to Siegwart et al. [11], the extraction of chlorides
salts, therefore chloride ions move more easily in the concrete alters the size and quantity of pores in the concrete, with effects
and are effortlessly extracted. on their resistance and permeability. Chlorides can be combined
Regarding current intensity, the usual values are between with hydrated cement compounds in order to form chloroalumi-
1 A/m2 and 5 A/m2. Despite this, the majority of the studies focused nate and chloro ferrite hydrates, contributing to changes in poros-
on densities until 1 A/m2 [11–13]. These authors believe that ity [20,26]. The main effect is reducing the number of macropores
higher intensities can generate cracking in the concrete as a func- and increasing the number of micropores [20].
tion of the chloride extraction velocity. In this sense, Elsener, Angst An increase in OH- concentration in the interstitial dissolution
[12] and Broomfield [14] determine that the current intensity around the steel bar has a positive effect in terms of protection
should not exceed 2 A/m2. against corrosion. However, if a concrete contains reactive
The application period of the technique varies between 6 and aggregate particles this can catalyze reactions like alkali-aggregate.
10 weeks, period in which a significant reduction of chloride con- Bennett et al. [28] performed tests on concrete containing reactive
centration is attributed to values below 0.40% of total chlorides, aggregates and verified increase in material expansion. Orellan,
considered to be a critical value [15–18]. If the content of chloride Escadeillas, Arliguie [19] verified formations of sodium rich phases
ions is reduced below the critical threshold, increased hydroxyls in the region near the steel reinforcement. However, the authors
production on the surface of the reinforcement and the simultane- state that these compounds were only verified under accelerated
ous reduction of the ions chloride content create enabling favor- conditions, and it is not possible to assert that the increase in
able environment for the reinforcement repassivation [7]. hydroxyl ions content can be considered as a catalyst of the reac-
According to results from Orellan and Escadeillas [19], chloride tivity with the aggregate.
content after treatment was reduced by around 40% within The objective of this paper is to evaluate the efficiency of the
7 weeks and, at the same time, significant amounts of alkali ions electrochemical chlorides extraction technique and its effect in
were observed around the steel. Efficiency percentages can reach concrete, by monitoring chloride content and corrosion potential,
up to 70% according to experimental works [5,20,21]. checking the influence of variables reinforcement concrete cover
Siegwart et al. [11], Liu and Shi [22], and Yeih; Chang and Hung and water/cement ratio. The measurement of absorption and con-
[23] performed the treatment during 6 weeks, obtaining satisfac- crete porosity, through porosimetry tests by mercury intrusion and
 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444 437

microtomography (Micro-CT), allowed the evaluation of concrete ies. Density was chosen to be a value, which ensured, as in previous
porosity subjected to chlorides extraction. studies like Ihekwaba, Hope [30] and Broomfield [14], a good
removal of chloride and no significant side effects.
2. Experimental procedure During ECE it was necessary to adjust the voltage in order to
maintain the current because, over time, the system became more
The independent variables of this study are: 1 – Thickness of the resistive and the potentiometer indicated a gradual decrease of
reinforcement concrete cover (2 levels: 1 cm and 3 cm); 2 – Water/ current. The electrolyte was weekly replaced in order to avoid
cement ratio of concrete (2 levels: 0.43 and 0.59) for the same con- the saturation of the solution by ions removed from concrete and
sistency of fresh concrete (100 ± 10 mm). by the product that resulted from the progressive degradation of
Response variables are: 1 – corrosion potential; remaining chlo- the stainless steel. The treatment lasted 61 days: 31 days with con-
ride content; water absorption; and concrete porosity (measured nected current, a seven-day rest period (none connected current)
by porosimetry techniques, by mercury intrusion and by and 24 additional days with connected current, following the
microtomography). model of Pérez et al. [31]. In order to regulate the current a manu-
ally tuning voltage source and a digital multimeter for individual
2.1. Mix proportion and the molding of concrete test samples current measuring were used in all treatment cells. The system cre-
ated to extract chlorides from the six concrete samples, in parallel,
Mix proportion, by weight, was 1:3 and 1:5 (cement: aggre- is illustrated in Fig. 4.
gates), in other words, one high and one low cement consumption During the treatment period, the current was daily measured
concretes, respectively, representing a variation range which and adjusted. Over time, it was noted that the mesh was slowly
includes concrete with great employability and different behaviors. degrading. This fact was already expected since, even with a high
Materials used in the concrete were cement CP II-F-32 (Portland noble metal, it was subjected to an electric current and behaving
cement with added limestone filer and characteristic strength of like an anode migration cell. The product of the mesh degradation
32 MPa), natural quartz as fine aggregate, granite as coarse aggre- created a thick layer of suspended material. This material was
gate with maximum diameter of 25 mm, mixing water from the removed weekly and the electrolyte replaced. This phenomenon
public supply and NaCl (1% Cl regarding cement mass). The was also reported in the study by Monteiro [7] and Monteiro
description of the concrete is shown in Table 1 and the character- et al. [32], which also used stainless steel as an external anode.
istics of the used concrete are defined in Table 2. Concrete descrip- The comparison between the material adhered to the concrete in
tion and its characteristics are shown in Tables 1 and 2, both studies is shown in Fig. 5. At the end of the treatment, the
respectively: mesh had lost strength and rupture could be observed, mainly in
Six prismatic concrete test samples were used, of equal dimen- the contact points between concrete and mesh.
sions 15  15  45 cm, with 4 steel bars of 10 mm in diameter,
simulating a portion of a reinforced beam. The schematic design 2.3. Techniques and characterization tests
of the samples used in the study is shown in Fig. 2.
When subjected to chlorides extraction, samples had three The method for determining the corrosion potential is in accor-
years old and, consequently, three years interacting with chloride dance with ASTM C876 [33]. A high impedance digital multimeter
ions and exposed to the external environment (sun and rain). Sam- was used, connected to the reinforcement to be evaluated and to a
ples were designed by Rocha [29] and, for more details about reference copper/copper sulfate electrode in contact with the con-
molding and proportions; those can be found on the references. crete at the reading point. A water saturated sponge was placed on
the contact surface between the reference electrode and the con-
2.2. Chloride extraction system crete, as illustrated in Fig. 6A. Measurements were made in six con-
trol points with water-saturated specimens, as shown in Fig. 6B.
The electrolyte used was water from the public supply. As an Water absorption by immersion was conducted in accordance
external anode, a stainless steel mesh with aperture equal to to NBR 9778 [34]. Two absorption measurements were conducted:
5  5 mm was used. It was completely immersed in the electrolyte the first with approximately three years of exposure to the envi-
disposed in a plastic box. The concrete test sample was placed over ronment prior to ECE and the second after chlorides extraction in
the mesh ensuring the pressing contact between electrolyte, anode order to assess the final status of the concrete.
and concrete as illustrated in Fig. 3A and in Fig. 3B. Total chloride content (soluble in acid) was determined accord-
The extraction was executed simultaneously in the concrete ing to ASTM C1152 [35]. For this analysis, concrete samples were
test samples on a single system connected in parallel. The electric extracted with a drill at depths of 1 cm and 3 cm. The extracted
current density applied was calculated as a function of the exposed material was manually pulverized and sifted at the strainer with
surface area of the concrete in order to ensure an electric current a 710 lm opening. The titration procedure consisted in the solubi-
density of around 1 A/m2 of concrete for all test samples in the ser- lization of 10 g of powder sample in 75 ml solution of water and

Table 1
Physical, mechanical and chemical characteristics of cement CP II-F-32.

Chemical analysis
CaO (%) SiO2 (%) Al2O3 (%) Fe2O3 (%) MgO (%) Na2O (%) SO3 (%) Ignition loss (%) Insoluble Residue (%)
60.68 18.83 4.43 2.64 4.98 0.61 2.83 5.05 1.78

Physical and mechanical characteristics

Setting time Normal consistency (%) Blaine’s specific surface area Residue on Compressive strength
(cm2/g)
Initial (h) Final (h) # 200 (%) # 325 (%) 3 days (MPa) 7 days (MPa) 28 days (MPa)
03:35 04:00 26.4 3140 3.00 14.70 26.90 32.50 41.30
438 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444

Table 2
Concrete mix proportioning.

m* Cement: sand: gravel: w/c Slump (mm) Cement content (kg/m3) Specific gravity (kg/m3) Compressive strength (MPa) – 28 days
3 1 1.08 1.92 0.43 105 540 2390 50
5 1 2.12 2.88 0.59 95 346 2280 31
*
Fine aggregate + coarse aggregate.

Fig. 2. Schematic representation of the concrete test samples [29].

Fig. 3. Chlorides extraction cell: Container with stainless steel mesh (A); System installed with the concrete test sample to be treated (B).

titration was triggered by a silver billet electrode coated with silver

chloride and by a reference electrode (Ag/AgCl).
In order to measuring, subjectively, the value of concrete poros-
ity, analyzes were performed by microtomography of the concrete
samples after they were submitted to chlorides extraction. Then,
samples were extracted and separated in 3–4 cm thick slices. The
top (portion from where the chlorides were extracted) and lower
(distal portion of the extraction face) parts were reserved for anal-
ysis and the central portion of the samples was discarded. Each
sample was reduced and brought to a micro-CT scanner model
SkyScan 1172 high-resolution micro-CT.
Four samples were scanned. Two samples with a mix propor-
tion of 1:5 (w/c = 0.59), one extracted from the face with chlorides
and another from the opposite face, and two samples with a mix
proportion of 1:3 (w/c = 0.43), one from the face with chlorides
and the other from the opposite side. Analyzed parameters were
total porosity, open porosity, closed porosity and average pore
diameter.
Fig. 4. Schematic of the system connected in parallel.
In addition, analyses of concrete porosity were executed by
means of the porosimetry technique of mercury intrusion. Samples
for this analysis were obtained similarly to the ones of the micro-
nitric acid at 1:1 ratio (50% concentration). After dissolved, it was tomography. The used equipment was a Micromeritics Autopore IV
heated to boil, filtered and placed on a magnetic stirrer, where 9500 porosimeter.
 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444 439

Fig. 5. Material formed by the degradation of the stainless steel mesh in this research (a) and in Monteiro’s [7] (b).

Fig. 6. Corrosion potential readings. Execution of the test (A) and control points (B).

According to Mattana et al. [36], the test by mercury intrusion mix, chloride removal efficiency was 80% and 84%, for cover thick-
porosimetry is limited to a pore size of about 400 lm (0.4 mm), nesses of 1 cm and 3 cm, respectively. Thus, it can be verified that
differently from microtomography that can evaluate pores of over the higher the water-cement ratio is, the higher the efficiency of
1800 lm (1.8 mm), but it is limited regarding the observation of the technique. Similar results were obtained by Shan et al. [6]
pores smaller than 10 lm (0.01 mm). This indicates that testing and Elsener; Angst [12].
should be complementary and not comparative. Santos and Lima [37] and Medeiros Junior et al. [38] present the
state of the art in resistivity measurement and explain that con-
crete with higher w/c ratio tends to have lower resistivity. Lopez
3. Results and discussion et al. [8] argue that extraction rate is directly proportional to the
applied current flow, and it is greater when the resistivity of the
3.1. Chlorides content concrete is smaller. In this way, results confirm the increased sus-
ceptibility to extracting chlorides in concrete with higher water/
It is important to emphasize that the treatment was applied on cement ratio, which can be attributed to greater porosity and con-
one side with a thicknesses of two concrete covers. Therefore, the sequent easiness of chloride movement to the outside of the
treated side had two bars: one with a 1.0 cm cover and another concrete.
with a 3.0 cm concrete cover. Samples to determine total chloride Castellote et al. [39] found that the highest current densities
content were taken at the center of the steel bar length and at occur in regions where the concrete cover is smaller, because the
1.0 cm in the lateral of each bar. Thickness of the concrete cover electrical resistance between the reinforcement and the anode is
of 1.0 cm thickness was represented by the extracted sample of lower. Regarding the study by Lopez et al. [8], it was expected that
0.8–1.2 cm. On the other hand, the thickness of the concrete cover in regions with lower concrete cover thicknesses the removal effi-
of 3.0 cm was represented by the sample extracted between 2.8 ciency was going to be higher. However, this theory was not proved
and 3.2 cm. by the results presented in this research since the final chloride
Fig. 7 shows the results for each reinforcement cover, before concentration found (after ECE) showed no significant differences
and after chlorides extraction. Samples from two depths of 1 cm in the measured concrete cover. Verified contents may be associ-
and 3 cm were analyzed, representing the reinforcement concrete ated to the amount of time used applying the technique. It is possi-
cover for two different mix proportions (w/c = 0.43 and 0.59). Per- ble that for smaller cover thicknesses a shorter time is required for
centage of initial chloride was determined in spot test (without chlorides extraction when compared to concretes with larger con-
layer samplings), since chloride was added to the mixing water crete cover thicknesses. Thus, the concrete of this research may
during the molding of concrete samples, therefore, having theoret- have reached its limit of chloride extraction in a shorter time than
ically the same amount of chloride at any depth. the one adopted in the treatment, as mentioned by Fajardo;
It can be noted that there was a significant decreasing in chlo- Escadeillas; Arliguie [40]. The authors verified that for thicknesses
ride content for both mix proportions. The 1:3 mix which has a of 20 mm, contents of chlorides are reduced in a short time, while
lower water/cement (w/c = 0.43) had less significant results, with for thicknesses of 50 mm the extraction speed is smaller.
removal efficiency of 62% and 55% for cover thicknesses of 1 cm Shan et al. [6] comment that ECE is not able to completely
and 3 cm, respectively. In the case of 1:5 (w/c = 0.59) concrete remove chloride ions. Total chloride extraction from concrete is
440 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444

Fig. 7. Chloride content before and after ECE for both mix proportions and for both concrete covers.

limited since only free chloride is available for extraction. Some of Regarding international standards, the critical total chloride
the chlorides are mobilized in the cement matrix linked to limit for reinforced concrete is 0.40% for European standards
hydrated products [27,41]. [15–17] and 0.35% for British standards [18]. Thus, for all tested

Fig. 8. Evolution of corrosion potential for concrete covers with thicknesses of 1 cm and 3 cm.
 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444 441

samples, the final contents of chlorides were consistent with these sion potential at the end of chloride extraction when compared
standards, in other words, chloride extraction resulted in contam- with data of the cover of 3 cm of thickness. Fajardo; Escadeillas;
ination levels below the critical limit specified in those documents. Arliguie [40] verified that regardless the thickness of the concrete
cover (2 and 5 cm), after ECE the corrosion potential remains on
3.2. Corrosion potential the uncertainty zone. On the other hand, for the higher w/c ratio

Corrosion potential (Ecorr) was read before, during and after

chlorides extraction. Results presented in Fig. 8 show gradual
increasing of corrosion potential. At the starting point (day 0, still
contaminated by chlorides), test indicates a 90% probability of cor-
rosion, according to ASTM C 876 [33]. After extraction, the poten-
tial decreased to values close to the zone of uncertainty.
According to results and statistical analysis (ANOVA), it is pos-
sible to conclude, with 95% certainty, that there is a relation
between corrosion potential measurements and water/cement
ratio during the extraction period.
These results are similar to the ones found by Elsener and Angst
[10] and Rocha [29]. They found the same relation in concrete con-
taminated by chlorides that was not subjected to extraction of
chlorides.
In concrete with lower w/c ratio (0.43), the thickness of the con-
crete cover influences corrosion potential. It was discovered that
the thinner concrete cover resulted in less electronegative corro- Fig. 10. Absorption comparison between the studied mix proportions.

Fig. 9. Evolution of corrosion potential by comparing different w/c ratio.

442 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444

(0.59), the thickness of the concrete cover had no significant influ- It’s important to highlight the difference found on readings
ence, resulting in the same level of corrosion potential at the end of from days 31 and 38. The reading of day 31 is the beginning of
chloride extraction. Fig. 9 shows results comparing the differences the rest period and day 38 is the moment before the restart of chlo-
between water/cement. ride extraction. The observed difference may be explained by the
It is also observed that, during the course of time, potential polarization caused by the current, which acts on the reinforce-
growth rate reduces. This is due to the fact that while chlorides ment while being subjected to the chlorides extraction system, as
are being extracted, the electrical resistivity of the environment also stated by Sánchez et al. [13].
increases (it was necessary to increase the voltage in order to
maintain the set current density), making difficult to remove most
of the ions and also influencing the reading of corrosion potential. 3.3. Water absorption
This effect was also verified by Carmona; Garces and Climent [5].
The effect is attributed to the increasing of concrete resistivity Absorption test represents the water penetration into the
and changes in the pore network of the material during the electric concrete under immersion conditions. In this case, the water per-
field application. The change in pore size influences the migration colates the concrete not only by capillarity, but through all open
properties of ions below a certain diameter, and thus influences the pores, and it may be associated with the total porosity within the
electrical resistance of the system [11]. concrete. Results of immersion absorption on the concrete

Fig. 11. Porosity by microtomography, before and after chlorides extraction.

Fig. 12. Porosity by mercury intrusion porosimetry, before and after chlorides extraction.
 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444 443

samples, before and after chlorides extraction, are shown in Fig. 10, rides extraction and 5.4% in the concrete without extraction.
indicating the increased absorption caused by the ECE. Therefore, in the case of water/cement ratio equal to 0.59, chlorides
Absorption increase is related to the increased amount, size and extraction caused a 33% increase in the total porosity and in the
distribution of pores. Chloride transport usually occurs through case of the water/cement ratio equal to 0.43 the increase was 30%.
capillary pores, which have high connectivity [27]. These factors Using data from the subdivision into macropores (pore diame-
may result in higher permeability and, consequently, enhanced ter greater than 50 nm) and mesoporous (pore diameter between
susceptibility to corrosion. With the extraction of chlorides by 10 nm and 50 nm); the data trend appears to be the same. Chlo-
ECE, the permeability of the concrete can increase by altering the rides extraction has caused an increase in these two porosity sub-
kinetics of chloride penetration in case of a recontamination of this divisions in all cases of this experiment.
treatment. Once the increased water absorption was confirmed a The proven increase in porosity of the concrete caused by the
microtomography and a mercury intrusion porosimetry were exe- chlorides extraction system explains the increment in water
cuted, aiming to understand the changes in porosity and perme- absorption detected by the water absorption test. This fact is
ability of the studied concrete. indicative that structures exposed to this recovery technique must
be protected with protective systems in order to prevent recon-
tamination, which will certainly be facilitated by the higher poros-
3.4. Microtomography and porosimetry by mercury intrusion
ity present. About this, Medeiros, Helene [42,43], Medeiros et al.
[44] present a study where the efficiency of protection systems
Microtomography was executed approximately 6 months after
for concrete structures was tested against chloride attack.
the end of chloride extraction, in order to understand the distribu-
tion of pores. The samples were divided into two groups: 1 – sam-
ples taken from the face from which chlorides were extracted
4. Conclusions
named C-CS, and 2 – samples taken from the opposite face, named
S-EC, without the effect of chloride extraction. Fig. 11 resumes the
This paper leads to the following considerations:
results of 3D analysis of samples in microtomography. Open poros-
ity values represent the percentage of interconnected porosity,
– Concrete produced with w/c ratios of 0.43 and 0.59 had an effi-
which are responsible for the water permeability and absorption
ciency for removing chlorides of 55–62% and 80–84% respec-
of the concrete. Microtomography indicates open porosity when
tively, confirming that concrete with higher w/c ratio tends to
there is a connection between points that are considered as pores
have a higher removal efficiency;
in the binarization process.
– There are no significant differences between the final chlorides
Data shows that open and closed porosity, consequently total
concentrations measured on the reinforcement concrete covers
porosity, are influenced by chloride extraction. The increase in
for both mix proportions;
total porosity was 35% in the case of concrete with water/cement
– During chloride extraction there was a gradual growth in corro-
ratio of 0.43 and 31% for concrete with water/cement ratio of 0.59.
sion potential, going from the high corrosion probability zone to
Porosimetry tests by mercury intrusion followed the same divi-
the uncertainty zone or even to the zone of low corrosion
sion and a resume of results can be seen in Fig. 12, which shows the
probability;
percentage of pores in each sample. Furthermore, Fig. 13 shows the
– ECE causes negative polarization of the reinforcement, which
average diameter of pores in the experimental samples, indicating
means that, after turning off the chloride extraction system, val-
that chloride extraction increases the average pore diameter.
ues tend to rise as discovered during the rest period of the
For w/c ratio 0.59, results of Fig. 12 show that concrete sub-
experiment;
jected to chlorides extraction have a higher porosity (13.0%) when
– For concrete samples produced with lower w/c ratio (0.43), the
compared with the face of concrete where no extraction was exe-
thickness of the reinforcement concrete cover has a great influ-
cuted (average 9.8%). In the case of w/c ratio 0.43, results follow
ence on the corrosion potential, and the lower cover thickness
the same trend with 7.0% total porosity in the concrete with chlo-
resulted in less negative corrosion potential. For concrete pro-
duced with higher w/c ratio (0.59), there was no significant
influence of the thickness of the concrete cover on the final val-
ues of corrosion potential;
– After ECE, the concrete water absorption increased 13% and 12%
for concrete w/c ratios of 0.43 and 0.59, respectively;
– Total porosity, measured both by microtomography and mer-
cury intrusion porosimetry, increases by more than 30% as a
result of applying chloride extraction. Chlorides extraction
increases the open and closed porosities and the percentage of
macropores (pores with diameter greater than 50 nm) and
mesopores (pores with diameter between 10 nm and 50 nm).
The average diameter of pores determined by mercury intrusion
porosimetry was increased due to the chloride extraction;
– Absorption and microtomography tests showed that the per-
centage growth in porosity due to chlorides extraction does
not depend on the w/c ratio; this total growth in porosity
explains the increase in water absorption observed after apply-
ing chloride extraction.
– From these conclusions, it is understood that the in order to
perform the electrochemical extraction of chlorides it is
required to take measures for preventing recontamination,
Fig. 13. Average diameter of pores, by mercury intrusion porosimetry, before and due to both the increased porosity and the absorption of the
after chlorides extraction. treated concrete.
444 L.R. de Almeida Souza et al. / Construction and Building Materials 145 (2017) 435–444

[24] G.E. Abdelaziz, A.M.K. Abdelalin, Y.A. Fawzy, Evaluation of the short and long-
References term efficiencies of electro-chemical chloride extraction, Cem. Concr. Res. Ed.
Elsevier (2009).
[1] P. Romano, P.S.D. Brito, L. Rodrigues, Monitoring of the degradation of concrete [25] T. Huang, X. Huang, P. Wu, Review of recent developments of electrochemical
structures in environments containing chloride ions, Constr. Build. Mater. 47 chloride extraction on reinforced concrete in civil engineering, Int. J.
(2013) 827–832. Electrochem. Sci. 9 (2014) 4589–4597.
[2] M.H.F. Medeiros, A. Gobbi, G.C. Réus, P. Helene, Reinforced concrete in marine [26] I. Sánchez, X.R. Nóvoa, G. de Vera, M.A. Climent, Microstructural modifications
environment: effect of wetting and drying cycles, height and positioning in in Portland cement concrete due to forced ionic migration test. Study by
relation to the sea shore, Constr. Build. Mater. 44 (2013) 452–457. impedance spectroscopy, Cem. Concr. Res. 38 (2008) 1015–1025.
[3] R.A. Medeiros Junior, M.G. Lima, P.C. Brito, M.H.F. Medeiros, Chloride [27] K.B. Kim, J.P. Hwang, K.Y. Ann, Influence od cementitious binder on chloride
penetration into concrete in an offshore platform-analysis of exposure removal under electrochemical treatment in concrete, Constr. Build. Mater.
conditions, Ocean Eng. 103 (2015) 78–87. 104 (2016) 191–197.
[4] C. Andrade, Manual for Diagnosis of Damaged Works by Corrosion of [28] J. Bennet, T.J. Schue, Evaluation of NORCURE Process for Electrochemical
Reinforcement, PINI, São Paulo, 1992. Chloride Removal from Steel-Reinforced Concrete Bridges Components.
[5] J. Carmona, P. Garcés, M.A. Climent, Efficiency of a conductive cement-based Strategic Highway Research Program, Report SHRP-C-620, 31, 1993.
anodic system for the application of cathodic protection, cathodic prevention [29] Rocha, F. C. Corrosion potential measurements in reinforced concrete
and electrochemical chloride extraction to control corrosion in reinforced structures: influence of water/cement ratio, temperature, chloride
concrete structures, Corros. Sci. 96 (2015) 102–111. contamination, concrete cover thickness and moisture content, Master’s
[6] H. Shan, J. Xu, Z. Wang, L. Jiang, N. Xu, Electrochemical chloride removal in thesis in Civil Engineering. Federal University of Paraná, Curitiba, Brazil,
reinforced concrete structures: improvement of effectiveness by simultaneous 2012. (in Portuguese).
migration of silicate ion, Constr. Build. Mater. 127 (2016) 344–352. [30] N.M. Ihekwaba, B.B. Hope, Mechanical properties of anodic and cathodic
[7] E.C.B. Monteiro, Electrochemical chloride extraction for rehabilitation of regions of ECE treated concrete, Cem. Concr. Res. 26 (5) (1996) 771–780.
concrete structures with steel reinforcement corrosion (Doctoral thesis), [31] A. Pérez, M.A. Climent, P. Garcés, Electrochemical extraction of chlorides from
Polytechnic School, University of São Paulo, São Paulo, Brazil, 2002 (in reinforced concrete using a conductive cement paste as the anode, Corr. Sci.
Portuguese). Ed. Elsevier (2010).
[8] C.A. Lopez, E.C.B. Monteiro, E.P. Figueiredo, F. Pereira, M.H.F. Medeiros, P. [32] P.J.M. Monteiro, P. Helene, I AOKI, P.E. Barbosa, E.C.B. Monteiro, Influence of
Helene, Repair and Protection of Reinforcing Steel, in: Paulo Helene, Fernanda water-cement ratio and cover thickness on chloride extraction of reinforced
Pereira. (Org.). Rehabilitation and Maintenance of Concrete Structures, first ed., concrete, ACI Mater. J., Farmington Hills 102 (1) (2005) 9–14.
São Paulo, Cargraphics, 2007. (in Portuguese). [33] American Society for Testing and Materials, ASTM C876: Standard Test Method
[9] R.B. Polder, Electrochemical chloride removal of reinforced concrete prisms for Half-Cell Potentials of Uncoated Reinforcing Steel in Concrete, Annuual
containing chloride from sea water exposure, UK Corrosion & Eurocorr’94, book of ASTM Standards, West Conshohocken, 2009.
Boumnemouth, 1994. [34] Brazilian Association of Technical Standards. NBR 9778: Mortar and hardened
[10] B. Elsener, U. Angstu, Mechanism of electrochemical chloride removal, in: concrete – Determination of water absorption, voids and density. 2005.
Corrosion Science. Ed. Elsevier, Jun. 2007. [35] American Society for Testing and Materials, ASTM C1152: Standard Test
[11] M. Siegwart, J.F. Lyness, F.J. McFarland, Change of pore size in concrete due to Method for Acid-Soluble Chloride in Mortar and Concrete, Annual book of
electrochemical chloride extraction and possible implications for the ASTM Standards, West Conshohocken, 2004.
migration of ions, Cem. Concr. Res. 33 (2003) 1211–1221. [36] A.J. Mattana, E. Pereira, M.R.M.M. Costa, Evaluation of Porosity in Mortar Bay
[12] B. Elsener, U. Angst, Mechanism of electrochemical chloride removal, Corros. X-Ray Microtomography (Micro-CT) as an Additional Tool for Mercury
Sci. 49 (2007). 4504-2522. Intrusion Porosimetry, Iberoamerican Jornal of Applied Computing, 2015.
[13] M. Sánchez, M.C. Alonso, Electrochemical chloride removal in reinforced [37] L. Santos, M. Lima, Key Factors Affecting Surface Electrical Resistivity
concrete structures: improvement of effectiveness by simultaneous migration Measurements of Concrete - the Art-State, VI International Congress of
of calcium nitrite, Constr. Build. Mater. 25 (2011) 873–878. Pathology and Rehabilitation of Structures, Córdoba, Argentina, 2010.
[14] J.P. Bromfield, Corrosion on Steel in Concrete: Understanding, Investigation [38] R.A. Medeiros Junior, M.G. Lima, M.H.F. Medeiros, L.V. Real, Investigation of
and Repair, E & FN Spon, 2003. compressive strength and electrical resistivity with different types of cement,
[15] Comité Euro-international du beton. Durable concrete structures: design ALCONPAT Journal 4 (2014) 116–132 (in Portuguese).
guide. Bulletin d’information, n. 152. Lausanne, 1992. [39] M. Castellote, C. Andrade, C. Alonso, Electrochemical removal of chlorides.
[16] European standards. DD ENV 206 Concrete: Performance, production placing Modelling of the extraction, resulting profiles and determination of the
and compliance criteria, 1992. efficient time of the treatment, Cem. Concr. Res. 30 (4) (2000) 615–621.
[17] Committee of concrete, EHE 2008, Struct. Concr. (2008). [40] G. Fajardo, G. Escadeillas, G. Arliguie, Electrochemical chloride extraction (ECE)
[18] British Standard. BS 8110–1 Structural Use of Concrete. Code of Practice for from steel-reinforced concrete specimens contaminated by ‘‘artificial” sea-
Design and Construction, 1997. water, Corros. Sci. 48 (2006) 110–125.
[19] J.C. Orellan, G. Escadeillas, G. Arliguie, Electrochemical chloride extraction: [41] J. Hoppe Filho, M.H.F. Medeiros, E. Pereira, P. Helene, G.C. Isaia, High-volume
efficiency and side effects, Cem. Concr. Res. 34 (2004) 227–234. fly ash concrete with and without hydrated lime: chloride diffusion coefficient
[20] M. Sánchez, P. Garcés, M.A. Climent, Electrochemical extraction of chlorides from accelerated test, J. Mater. Civ. Eng. 25 (2013) 411–418.
from reinforced concrete: variables affecting treatment efficiency, Materiales [42] M.H.F. Medeiros, P. Helene, Efficacy of surface hydrophobic agents in reducing
de constrcción 56 (284) (2006) 17–26. water and chloride ion penetration in concrete, Mater. Struct. 41 (2008) 59–
[21] T.H. Nguyen, T.A. Nguyen, V.K. Le, T.M.T. Dinh, H. Thai, X. Shi, T.H. Nguyen, A 71.
study of the function mechanism of a corrosion inhibitor package for [43] M.H.F. Medeiros, P. Helene, Surface treatment of reinforced concrete in marine
martensitic stainless steel tubulars, Anti-Corros. Methods Mater. 63 (5) environment: influence on chloride diffusion coefficient and capillary water
(2016) 398–404. absorption, Constr. Build. Mater. 23 (2009) 1476–1484.
[22] Y. Liu, X. Shi, Electrochemical chloride extraction and electrochemical [44] M.H.F. Medeiros, E. Pereira, V.A. Quarcioni, P. Helene, Surface treatment
injection of corrosion inhibitor in concrete: State of the knowledge, Corros. systems for concrete in marine environment: Effect of concrete cover
Rev. (2009). thickness, REM – Int. Eng. J. 69 (2016) 287–292.
[23] W. Yeih, J. Jhy, C.C. Hung, Selecting an adequate procedure for the
electrochemical chloride removal, Cem. Concr. Res. Ed. Elsevier (2005).