Int. J. Electrochem. Sci.

, 10 (2015) 5309 - 5326

International Journal of
ELECTROCHEMICAL
SCIENCE
www.electrochemsci.org

Critical Chloride Concentration of Rebar Corrosion in Fly Ash

Concrete

Congtao Sun1, Shiqun Liu2,*, Jiangang Niu3, Weichen Xu1

1
Institute of Oceanology Chinese Academy of Science, Qingdao 266071, P.R. China
2
Architectural Design and Research Institute of Qingdao University, Qingdao 266033, P.R. China
3
Inner Mongolia University of Science and Technology, Baotou 014010, P.R. China
*
E-mail: liushiqunqingdao@163.com

Received: 8 December 2014 / Accepted: 17 March 2015 / Published: 27 May 2015

Electrochemical impedance spectrum has been used in this work to study the critical chloride
concentration for the rebar corrosion in fly ash concrete. The analysis and criteria for the critical
corrosion point of rebar have been interpreted, and the surface morphology of the corroded rebar has
been studied using SEM. The influences of the wet-dry cycling period and the fly ash content on the
impedance spectra and the critical chloride concentration have been discussed respectively. It shows
that the critical corrosion point can be precisely manifested by impedance spectra; the resistance of the
concrete layer increases with increasing period of wet-dry cycling and increasing content of fly ash,
which can be observed as translation of the spectra to the right on the complex plane; the critical
chloride concentration tends to increase with increasing period of wet-dry cycling; the critical chloride
concentration for rebar corrosion can be increased by a small dosage of fly ash in the concrete.

Keywords: Critical chloride concentration, rebar corrosion, electrochemical impedance spectrum, fly
ash concrete

1. INTRODUCTION

Critical chloride concentration is significant for the prediction of durability of concrete

structures, and is also essential for the design, test and maintenance of structures. Critical chloride
concentration can be defined in two ways. One definition is from the aspect of scientific research: the
highest chloride concentration in the surrounding electrolyte in the pores of concrete, which would not
induce depassivation. The other definition is from the aspect of practical engineering: the chloride
concentration in the surrounding electrolyte in the pores of concrete, which just induces visible or
acceptable corrosion. However, due to the difficulty of quantifying the visible or acceptable corrosion
Int. J. Electrochem. Sci., Vol. 10, 2015 5310

in the latter definition, the results differ among different researchers. Therefore, the former definition
of the critical chloride concentration has been used more widely and frequently[1,2].
There has been a lot of work about the study on critical chloride concentration[2-8], however,
the experimental data are scattered, which may be ascribed to the various measurement methods on the
corrosion of rebar[6]. Furthermore, different criteria are applied for different methods, which directly
affect the values of the critical chloride concentration. In addition, the method to obtain the sample
from the surrounding rebar environment at the onset of corrosion is also very important. Therefore, the
criteria for the onset of rebar corrosion and the proper method to obtain specimens from the
surrounding rebar environment can both help minimize the scattering of the critical chloride
concentration value. Meanwhile, they also make it possible for the practical application of the
experimental data.
Ac impedance method has been used in this work to monitor and manifest the onset of the rebar
corrosion in the concrete. The influence of the wet-dry cycling period and the fly ash content in the
concrete on the impedance spectra and critical chloride concentration have been analyzed and
discussed. This work should provide reference on manifesting the onset of rebar corrosion, the design
for durability and service life prediction of structures.

2. EXPERIMENTAL METHOD

2.1 Raw materials and mix proportion of concrete

P.O42.5 ordinary Portland cement; the local river sand, fineness modulus of 2.9, medium sand;
basalt, size 5~20 mm; original state of fly ash level Ⅱ; high efficiency water reducing agent, water
reducing rate is 25% ~ 30%; tap water. Concrete mix proportion as shown in Table 1.

Table 1. Mix proportion of concrete

Specimen Material Consumption（kg/m³） Water reducer

Cement Fly ash Water Sand Crushed
stone
A 400 0 160 662.4 1177.6 1.0%
B 360 40 160 662.4 1177.6 1.0%
C 320 80 160 662.4 1177.6 1.0%

2.2 Preparation of specimens

The HRB335 grade Φ10 rebar was grounded, and the rust and corrosion products on the
surface were removed. One side of the rebar was connected with a 500 mm long electric cable and
degreased with acetone. Then Φ16 heat-shrinkable tube and hot melt glue were used to seal the end of
the rebar, so that the exposed part was 20 mm long as a rebar electrode, which was fixed within a
Int. J. Electrochem. Sci., Vol. 10, 2015 5311

100×100×300 mm steel mould using two pieces of plastic end plate, and the concrete protective layer
was 15 mm thick. After molding for 24 h, specimens were demoulded and kept curing for 56 days.
Both ends and two sides of the specimen were sealed with epoxy resin, with only the top and bottom
sides exposed in chloride electrolyte.

2.3 Wet-dry cycling

For the wet process, the concrete specimen was immersed in 3.5% NaCl. For the dry process,
the specimen was dried in an oven in 50℃. The wet-dry cycling was carried out by controlling the time
of immersing and drying. The details of the process are shown in Table 2.

Table 2. Wet-dry cycling test system

Specimen Serial Number Cycle Time(h) Dry-wet Cycling System

A A 72 3:1(54:18)
B B1 72 3:1(54:18)
B2 96 3:1(72:24)
B3 144 3:1(108:36)
C C 72 3:1(54:18)

2.4 Monitoring the onset of rebar corrosion

Electrochemical impedance spectrum (EIS) was applied for the monitoring of the rebar
corrosion at room temperature. The electrochemical workstation was a PARSTAT®-2273 Advanced
Electrochemical System, and a three electrode system was used. At open circuit potential, the
amplitude was 10mv, and the frequency was between 0.01 Hz and 105 Hz, with 9 points recorded in
each decade. The schematic diagram is shown in Fig.1.

Wire PARSTAT-2273
Electrochemical
Workstation
Auxiliary Reference
electrode electrode

3.5%NaCl solution
Work electrode Computer
Concrete test cube

Figure 1. Schematic diagram of the testing of AC impedance

Int. J. Electrochem. Sci., Vol. 10, 2015 5312

The procedure is as follows. The specimen was immersed in 3.5% NaCl, and OCP was
measured to test the stability of the system. If the change of the OCP was less than ±2 mV within 5
min, the system can be regarded as in a steady-state, indicating that EIS measurement can be carried
out. The wet-dry cycling continued until corrosion of rebar took place. At the onset of corrosion, linear
polarization (LPR) and Tafel polarization were applied which may confirm the onset of corrosion. The
LPR was carried out between -20 mV and +20 mV with respect to OCP, and potential scan rate was
0.1 mV/s. Tafel polarization was carried out between -250 mV and +250 mV with respect to OCP, and
the potential scan rate was 0.5 mV/s.

2.5 Measurement of the critical chloride concentration

At the onset of rebar corrosion, the specimen was splitted along the rebar by compression-
testing machine and the corrosion on the surface can be observed; then one half of the specimen was
cut off using a cutting machine to remove the boundary effect, only leaving the middle part of the
concrete (ca. 150 mm long). Then this part of the concrete was placed in pulverizer, and powder
samples were obtained layer by layer along the transport direction of chloride ions. The concentration
of free chloride ions in the powder sample was measured using a chloride ion selective electrode. The
concentration of chloride ions at a depth of 15 mm in the reinforcement cover is regarded as the critical
chloride concentration.

3. RESULTS AND DISCUSSION

3.1 The analysis and criteria of the critical corrosion point of rebar

This work applies EIS method to monitor the corrosion of rebar in concrete, which includes
qualitative and quantitative analysis. Qualitative analysis is usually applied on slight corrosion
tendency, while quantitative analysis is used for noticeable corrosion tendency.
(1) Qualitative analysis of EIS
Cheng-hong Yang, Mei-lun Shi[9] gives a typical Nyquist graph about the passivated
reinforcement in the reinforced concrete through analyzing the Nyquist graph theory, as shown in
Figure 2. Can be seen from the figure in the Nyquist graph is made up of two time constant which
shows a small semicircle and an oblique line (which is actually a small arc section of a big
circle ).When testing the actual reinforced concrete specimen , the frequency (ω) changing from high
frequency to low frequency (ω ranged from ∞ to 0 ), and reflecting the response of the disturbance
signal on the Nyquist graph which was caused by the respective parts of the reinforced concrete
specimens, all the parts of the graphics stand for different kinetic parameters of the reinforced concrete
system, including the high frequency part reflects the characteristics of the concrete and steel/concrete
interface, and the low frequency part represents the characteristics of the passivation membrane on the
surface of the steel. Therefore, can be respectively obtained the relevant information of the reinforced
concrete and passivation membrane on the interface of steel/concrete by the curve change of high
Int. J. Electrochem. Sci., Vol. 10, 2015 5313

frequency part or low frequency part. When Specific judgment Nyquist diagram can be observed in the
low frequency curve, if it’s a larger slope diagonals, denoting the passivation membrane on the surface
of the steel in concrete is complete;If the slope of low frequency curve in figure is obviously reduced,
the radius of the second semicircle present a limited number, the Nyquist shows distinctly two time
constant, which explains the passivation membraneon the surface of the steel began to destroy.

Figure 2. The typical Nyquist figure of reinforced concrete [9]

The electrochemical impedance spectra of the specimen in each group are shown in Fig.3 as a
function of time (only one specimen is shown for each group). It can be seen that there are two
capacitive loops in the Nyquist plots, corresponding to two time constants. In the beginning of the
experiment, the capacitive loop in the low frequency region is a rising “straight line”, which means the
impedance and capacitance of the passive film are relatively large, and the rebar in the concrete is
passive[10]. During the process of the experiment, the rising angle of the “straight line” decreases and
becomes a small arc in the end. The change on the spectra indicates the onset of corrosion, i.e. the
decrease of resistance and capacitance of the passive film. The qualitative analysis on the EIS can be
used as a preliminary assessment on the corrosion of rebar, while more precise measurement
(quantitative analysis) has been carried out on the impedance spectra using equivalent circuit.

2500 2500

B1-3(0d) A-1(0d)
B1-3(36d) 2000 A-1(36d)
2000
B1-3(48d) A-1(48d)
B1-3(60d) A-1(60d)
B1-3(72d)
1500 1500
B1-3(84d)
Zim(ohms)
Zim(ohms)

1000 1000

500 500

0 0

0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
Zre(ohms) Zre(ohms)

（a） EIS of specimen B1-3 （b） EIS of specimen A-1

Int. J. Electrochem. Sci., Vol. 10, 2015 5314
2200

2000 C-2(0d)
1800 C-2(36d)
C-2(48d)
1600 C-2(60d)
1400 C-2(72d)
C-2(84d)
1200

Zim(ohms)
1000

800

600

400

200

-200
-200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Zre(ohms)

（c） EIS of specimen C-2

3000 B2-3(0d)
B2-3(36d)
2500 B2-3(48d)
B2-3(60d)
B2-3(72d)
2000
B2-3(84d)
B2-3(88d)
Zim(ohms)

1500

1000

500

0 500 1000 1500 2000 2500 3000

Zre(ohms)

（d） EIS of specimen B2-3

1800
B3-2(0d)
1600
B3-2(36d)
1400 B3-2(48d)
B3-2(60d)
1200 B3-2(72d)
B3-2(84d)
1000
Zim(ohms)

B3-2(90d)
800

600

400

200

-200
0 200 400 600 800 1000 1200 1400 1600 1800
Zre(ohms)

（e） EIS of specimen B3-2

Figure 3. EIS results with time

(2) Quantitative analysis of EIS

Quantitative analysis of EIS was immediately carried out when the qualitative analysis showed
large corrosion tendency. Quantitative analysis started after 48d in the test. Firstly, ZSimpWin
software should be used to fit the equivalent circuit (equivalent circuit was shown in Fig.4), obtaining
the value of impedance and capacitance, analysing the passive film. Line polarization measurements
and Tafel polarization measurements was used to manifest the corrosion of rebar at the onset of
corrosion (which was shown in Fig.5). The results of quantitative analysis of the critical corrosion
point of rebar are shown in Table 3, 4.
Int. J. Electrochem. Sci., Vol. 10, 2015 5315

R2 R4
R1

Q Q
Q1 Q3

Note: R1 represents the resistance value of the solution; R2 represents the resistance value of concrete;
R4 represents the electric double layer steel/concrete pore fluid interface electron transfer
resistance

Figure 4. Equivalent circuit of corrosion of the rebar in concrete at the onset of corrosion
-0.1
-0.39 B1-3-84d(1)
B1-3-84d(1)
-0.2

-0.40
-0.3

-0.41 -0.4

E(V)
E(V)

-0.5
-0.42

-0.6

-0.43
-0.7

-0.44 -0.8
-0.00004 -0.00003 -0.00002 -0.00001 0.00000 0.00001 0.00002 0.00003 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0

I(A) log(Icorr)

-0.41 -0.1

A-1-60d(1)
A-1-60d(1) -0.2
-0.42

-0.3
-0.43
-0.4
E(V)

E(V)

-0.44
-0.5

-0.45 -0.6

-0.7
-0.46

-0.8
-0.00005 -0.00004 -0.00003 -0.00002 -0.00001 0.00000 0.00001 0.00002 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0
I(A) log(Icorr)

-0.1

C-2-84d(1)
-0.38 C-2-84d(1) -0.2

-0.39 -0.3

-0.4
E(V)

-0.40
E(V)

-0.5
-0.41

-0.6

-0.42

-0.7
-0.000025
-0.000020
-0.000015
-0.000010
-0.0000050.0000000.0000050.0000100.0000150.000020 -7.5 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5

I(A) log(Icorr)

-0.35
-0.1 B2-3-88d(1)
B2-3-88d(1)

-0.36 -0.2

-0.3
-0.37
E(V)
E(V)

-0.4

-0.38
-0.5

-0.39
-0.6

-0.40 -0.7
-0.000025-0.000020-0.000015-0.000010-0.0000050.0000000.0000050.0000100.000015 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0

I(A) log(Icorr)
Int. J. Electrochem. Sci., Vol. 10, 2015 5316
-0.33

B3-2-90d(1) -0.1 B3-2-90d(1)

-0.34
-0.2

-0.35 -0.3

E(V)
E(V)
-0.36 -0.4

-0.5
-0.37

-0.6

-0.38
-0.00004 -0.00003 -0.00002 -0.00001 0.00000 0.00001 0.00002 0.00003 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0
I(A) log(Icorr)

（a）LPR polarization curves （b） Tafel polarization curves

Figure 5. LPR and Tafel polarization curve of the specimens

Table 3. Data analysis of EIS equivalent circuit

R1 R2 R4
specimen Chisq
(Ω/cm2) (Ω/cm2) (Ω/cm2)
B1-1-87 1.02E-07 1482 266.8 0.03299
B1-2-87 7.31E-11 1210 366.1 0.02308
B1-3-84 1.00E-07 787.4 254.6 0.02566
A-1-60 2.12E-06 597.2 229 0.01162
A -2-54 1.53E-08 345.3 195.8 0.03366
A -3-42 1.48E-05 558.1 218 0.01002
C-1-84 9.98E-06 1262 223.2 0.0287
C -2-84 1.00E-07 1220.8 264.1 0.0208
C -3-84 5.63E-05 392.9 320.6 0.004521
B2-1-88 2.04E-05 1378 296 0.01305
B2-2-88 0.0001962 1416 233.5 0.148
B2-3-88 7.49E-05 1552 318.4 0.02251
B3-1-90 2.90E-07 1910 274 0.0362
B3-2-90 1.00E-07 968.1 266.4 0.01261
B3-3-90 1.00E-07 313.6 292.4 0.02606
Note: number B1-1-87 represents the fitting data of No.1 in group B1 after 87d, and the value of Chisq
represents degree of fitting, the lower it is the better the degree of fitting is.

Table 4. Data analysis of LPR and Tafel

LPR Tafel
specimen 2
Rp（Ω/cm ） 2
Ecorr(mV) icorr(μA/cm ) Ecorr（mV） icorr(μA/cm2)
B1-1-87 1782.00 -388.36 12.20 -469.32 2.22
B1-2-87 1502.00 -391.79 14.47 -417.85 1.95
B1-3-84 1092.00 -418.38 19.90 -435.61 2.04
A-1-60 895.20 -402.66 11.13 -426.64 0.83
A-2-54 614.51 -408.68 35.38 -428.12 4.49
A-3-42 774.66 -406.40 28.06 -423.15 2.70
Int. J. Electrochem. Sci., Vol. 10, 2015 5317

C-1-84 1441.00 -397.69 15.09 -416.47 3.48

C-2-84 1485.00 -405.00 14.63 -425.58 3.14
C-3-84 694.53 -446.66 31.30 -469.42 6.18
B2-1-88 1693.12 -387.12 21.89 -407.77 2.62
B2-2-88 1793.54 -440.66 24.33 -430.98 4.48
B2-3-88 1911.20 -380.59 10.69 -407.09 1.01
B3-1-90 2203.00 -403.88 35.03 -430.98 4.48
B3-2-90 1002.00 -356.94 21.68 -383.98 2.13
B3-3-90 494.68 -418.54 43.95 -442.47 5.61

Equivalent circuit fitting data from Nyquist graph could carry out a quantitative analysis about
the degree of rebar corrosion in reinforced concrete, and meanwhile get some other parameters about
the reinforced concrete specimen, which could more accurately estimate the degree of rebar corrosion
in order to quantify the beginning of corrosion to get more accurate critical chloride ion density.
Reference [9] came up with the typical reinforced concrete ac impedance spectra (Fig. 2) through
theoretical analysis. The meanings of symbols in the Figure are as follows: (1) Rs is the limit value
when ω→∞, on behalf of the resistance value of concrete pore solution, which is inversely
proportional to the total porosity [11]; (2) Rct represent the diameter of the high frequency semicircle,
on behalf of OH - reaction impedance in the pore solution. According to the parallel capillary bundle
model of microstructure of concrete, the interior of the concrete is a network space composed by
porosity and capillary. Due to the same reaction kinetics constants from the same block, Rct reflects
that the contact reaction area of steel electrode and concrete surface is inversely proportional to the
total surface area of all capillaries [12]; (3) Cdl can be obtained by the radius of the high frequency
semicircle, represents the capacitance between electrodes and concrete surface, which is proportional
to total surface area of all capillaries.
In addition, as shown in the comparison between table 3 and table 4, the fitting data is obtained
from the data analysis of EIS and LPR. R4 that obtained from EIS shows that, due to the eliminated
influence of Ohm drop, the transfer resistance of electric double layer from the interface area of
reinforced concrete is a real polarization resistance. Moreover, the result of polarization Rp from LPR
(linear polarization) actually includes two parts - R2 and R4, which could be proved by the result of
(R2 + R4)/Rp close to 1 from the table. To sum up, R2 is the resistance value of concrete layer and R4
is the real polarization resistance value - Rp. Thus in fitting the equivalent circuit analysis (as shown in
table 3), the resistance value of R4 (which is the polarization resistance Rp [13]) is basically all around
250 Ω/cm2, according to the estimate criteria from Reference[10], steel passivation membrane has
basically entered the stage of corrosion.
Then the linear polarization curve of the specimens and Tafel polarization curve (as shown in
Figure 5 (a),(b) was fitting analysised, the fitting data as shown in table 4. The LPR fitting data can be
found in the table, the Rp basically all around 1000Ω/cm2 (note that this includes the resistance of
concrete layer), corrosion potential Ecorr is about - 400 mv, corrosion current density icorr is about 20
mu A/cm2; Furthermore, Tafel fitting data can be found in the table at the same time, the corrosion
potential Ecorr is about - 400 mv, corrosion current density icorr approximately 3μA/cm2.
Furthermore, Table 4 indicates that the fitting data Ecorr of LPR and Tafel test are close, while the icorr
Int. J. Electrochem. Sci., Vol. 10, 2015 5318

are of great difference, so we can know that the fitting data of LPR are just for the surface corrosion of
the locate corroded points, it is just a locate value. While the fitting data of Tafel are for the surface
corrosion of whole working electrode, so it is an average value. [14] Combined with the literature [10]
[14] which give the judgment standards, through the comprehensive analysis including the ac
impedance spectrum method, linear polarization method and Tafel polarization method, we can
eventually determine the corrosion of rebar that reached a tipping point. In addition, the test that
splitting the reinforced concrete test specimens can be found there are small amounts of corrosion
products on the surface of rebar (at the stage of initial development of corrosion, as shown in Figure 7,
more detailed analysis, please see section 3.2 ), and it’s more advantageous to prove that the above-
mentioned judgment is right.

-340
45 icorr(LPR) Ecorr(LPR)
icorr(Tafel) Ecorr(Tafel)
40 -360

35
-380
30
icorr(uA/cm )

-400
2

25
Ecorr(mV)
20 -420

15
-440
10
-460
5

0 -480
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14 16
Test Specimen Test Specimen

（a） The contrast analysis diagram about icorr （b） The contrast analysis diagram about Ecorr

Figure 6. The contrast analysis diagram about icorr and Ecorr through the LPR and Tafel polarization
method

In Table 3, the resistance values of R4 are about 250Ω/cm2, and it shows that the film had been
destroyed as reported in paper [10]. The resistance values of specimens B2-3-88 and C-3-84 range
from 300 to 400Ω/cm2, however, the LPR and Tafel test results in Table 4 and the observation by
splitting the specimens show that the rebar enter the initial stage of corrosion, as shown in Fig.7.
Compared with the standard in paper [10], middle or low corrosion rate can be obtained by the
polarization resistance Rp, but high corrosion rate can be obtained by the corrosion current density i corr,
and medium high corrosion rate was observed by splitting the specimens. Thus errors exist in the
corrosion tests of rebar, and the judgment of rebar corrosion should not limit to one method, but many
methods should be combined for aggregate analysis. Based on the analysis of the test results, the
Rp=300Ω/cm2 can be assumed to be the judgment standard of the corrosion critical point about
reinforced in concrete.
Through the above test and the analysis, the critical corrosion point of rebar can be precisely
manifested by qualitative and quantitative analysis of EIS, laying a good foundation of the study on
critical chloride concentration.
Int. J. Electrochem. Sci., Vol. 10, 2015 5319

（a）Rebar corrosion of B2-3 （b）Rebar corrosion of C-3

Figure 7. Rebar corrosion (observation by splitting)

3.2 SEM measurement

Scanning electron microscope (SEM) was used to observe the surface of the rebar samples
which included corroded parts and the uncorroded parts. First of all, the uncorroded parts were
observed, as shown in Fig.8. It can be seen that the surface of the rebar is rough, and there are a large
number of cement hydration products (mainly needle stick ettringite) attached to the surface of rebar,
but the distribution of crystallization is not uniform, and the passive film on the surface of rebar is also
not uniform, however it is not destroyed.

Ettringit
e

（a）Surface appearance of the uncorroded（b）Surface appearance of the uncorroded

rebar(x2000) rebar(x5000)

Figure 8. SEM of surface appearance of the uncorroded rebar

Secondly, the corroded part of the rebar was observed, as shown in Fig.9. From Fig.9 (a), (b), it
can be seen that the products of the corroded rebar are black, with dense structure (black rust Fe 3O4),
and some cracks existing on the surface of the corroded rebar. The reason may be that the chloride ions
gathered on the surface of the rebar and entered through the weak part of the passive film, and reacted
with the iron substrate. From Fig.9 (c), (d), we can see that there are evenly, loose corrosion products
Int. J. Electrochem. Sci., Vol. 10, 2015 5320

(red rust Fe2O3·xH2O) and obvious cracks on the surface of the rebar. It is mainly because the volume
expansion of the red rust generated through the corrosion process resulted in corroded expansion
cracks.

（a）Surface appearance of black rust (x1000) （b）Surface appearance of black rust (x5000)

（c）Surface appearance of red rust (x1000) （d）Surface appearance of red rust (x5000)

Figure 9. SEM of surface appearance of corroded rebar

Finally, the transition part of the corroded and uncorroded rebar surface was observed, as
shown in Fig.10. It shows that the destroyed area of the passive film have some rust (mainly black rust
Fe3O4, a small number of red rust Fe2O3·xH2O). And the intact surface is still closely integrated with
the cement hydration products, surface passive film is not destroyed. From the point of rebar corrosion
situation, most of the rebar surface is black rust. The main reason is because, compared with Fe2O3,
Fe3O4 has both trivalent iron ions and bivalent ferrous ions, so its corrosion products is not stable, but a
product between passivation and corrosion. If O2 and H2O were cut off this moment, corrosion reaction
will be terminated. Only in the case of O2 diffusion enough, Fe3O4 can continue to generate stable
oxide corrosion products Fe2O3. Through the above analysis, it shows that the rebar corrosion state was
at the initial stage, as the onset of corrosion of rebar, and the concentration of chloride ions at the depth
of reinforcement cover is regarded as the critical chloride concentration this moment.
Int. J. Electrochem. Sci., Vol. 10, 2015 5321

（a）Surface appearance of the transition （b）Surface appearance of the transition

part(x2000) part(x5000)

Note: In fig (6) oval area was red rust. Rectangle area was uncorroded part of the rebar passive film.
Triangle area was black rust.

Figure 10. SEM of surface appearance of transition part of the corroded and uncorroded rebar surface

3.3 The impact analysis on ac impedance spectra of wet-dry cycling period

Ac impedance spectra of concrete specimens under different cycle are shown in Figure 11. It
can be seen that obvious difference exists in the ac impedance spectra of the specimens under different
cycle. With increasing period of wet-dry cycling, translation phenomenon of the spectra appeared on
the complex plane. In addition, when the specimens are at the end of drying and soaking under the
action of the different cycle, testing the saturation of them , as shown in Figure 12, we can see from the
picture, in the case of dry-wet circulation time ratio remains the same (all are G: S = 3:1), with the
increase of dry-wet cycle, the saturation of concrete block is similar at the end of soaking, but the
saturation of them is decreased obviously at the end of drying. Through the above experiment can
explain that the resistance and capacitance (indicators by the high frequency part) of concrete under
different cycle produce differences due to the different degree of drying, namely, with increasing
period of wet-dry cycling, the degree of drying for specimens increased, and the resistance of the
concrete layer increased, which can be observed as translation of the spectra to the right on the
complex plane.

1800
3000
B1-3(0d) B1-3(36d)
1600 B2-3(36d)
B2-3(0d)
2500 B3-2(0d) B3-2(36d)
1400

1200
2000
1000
Zim(ohms)
Zim(ohms)

1500
800

600
1000
400

500 200

0
0
-200
0 500 1000 1500 2000 2500 3000 -200 0 200 400 600 800 1000 1200 1400 1600 1800

Zre(ohms) Zre(ohms)

（a）EIS after 0d （b）EIS after 36d

Int. J. Electrochem. Sci., Vol. 10, 2015 5322

2200
1600
B1-3(48d) B1-3(60d)
2000
B2-3(48d) B2-3(60d)
1400
1800 B3-2(48d) B3-2(60d)
1600 1200
1400
1000

Zim(ohms)
Zim(ohms)

1200
800
1000

800 600
600
400
400
200
200

0 0

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 0 200 400 600 800 1000 1200 1400 1600
Zre(ohms) Zre(ohms)

（c）EIS after 48d （d）EIS after 60d

1200
1400
B1-3(72d) B1-3(84d)
B2-3(72d) 1000 B2-3(84d)
1200 B3-2(72d) B3-2(84d)

1000 800
Zim(ohms)

Zim(ohms)
800
600

600
400
400

200
200

0 0

0 200 400 600 800 1000 1200 1400 0 200 400 600 800 1000 1200
Zre(ohms) Zre(ohms)

（e）EIS after 72d （f）EIS after 84d

Figure 11. EIS of the specimens under different cycle

100

90

80

70

60
Saturation(%)

50

40

30

20
B1(Z=72h)
10 B2(Z=96h)
B3(Z=144h)
0

-3 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54
Time(d)

Figure 12. The change of saturation along with dry-wet cycles

In addition, it can be seen that although the spectra shows some translation, the rebar corrosion
development process of B1, B2, B3 and other groups of specimens are almost the same (the topology
of low frequency part changed almost the same), some exceptions of the specimens B1-3 existed after
36d, which may be caused by the test error.
Int. J. Electrochem. Sci., Vol. 10, 2015 5323

3.4 The impact analysis on critical chloride ion concentration of wet-dry cycling period

When judging the rebar corrosion in concrete took place, its critical chloride concentration
should be tested immediately according to the test process in section 2.5 of this paper. This sampling
method is more reasonable than sampling by knocking mortar around reinforced to grind, because the
chloride ions influencing rebar corrosion distribute in a small range around the rebar, it is difficult to
control the sampling scope, causing test result greatly discrete. In addition, this method is compared
with the embedded type chlorine ion selective electrode, although the method of embedded type
chlorine ion selective electrode can get the targeted data on the point [15], its limitations is bigger, the
influencing factors of test results is more, and it is also not convenient to apply it to the detection of
practical engineering; while the method of filter pressing can get more rational data, but this method
can only get a regional average data, it turns out to be inaccurate in the case that chlorine ion is in
gradient distribution[2]. Comprehensively comparative analysis, The sampling method of this article
can avoid to produce larger human error, the result is more accurate and reliable.
The critical chloride concentration results of specimens under different wet-dry cycling are
shown in Table 5. It can be seen that the critical chloride concentration increases slightly with
increasing period of wet-dry cycling. The main reason is that with the increase of wet-dry cycling, the
degree of drying for concrete increased, leading to the increasing of capillary absorption, which can
absorb a lot of chloride ions in the process of immersing. But the diffusion of O2 is restrained because
of extending time of immersing, so rebar corrosion is delayed onset, and the critical chloride
concentration increases.

Table 5. The critical chloride concentration under the action of different wet-dry cycling

Average
Percentage of Average value
Cl-（mol/L Percentage of value
Specimens binding material of binding material
） concrete (%)
(%)
of concrete
(%)
(%)
B1-1 0.004108 0.364 2.039
B1-2 0.003326 0.295 1.651 0.318 1.780
B1-3 0.003326 0.295 1.651
B2-1 0.004566 0.405 2.266
B2-2 0.003829 0.339 1.900 0.357 2.000
B2-3 0.003697 0.328 1.835
B3-1 0.000001 0.352 1.968
B3-2 0.000001 0.339 1.900 0.347 1.946
B3-3 0.000001 0.352 1.968

3.5 The impact analysis on EIS of fly ash content

EIS of concrete specimens under different content of fly ash are shown in Fig.13. It can be seen
that, with the increase of fly ash content, obvious translation phenomenon of the spectra existed on the
Int. J. Electrochem. Sci., Vol. 10, 2015 5324

complex plane. The main reason is that with the increase of fly ash content, the compactness of
concrete increases, leading to the increase of the resistance and capacitance, which can be observed as
translation of the spectra to the right on the complex plane (resistance increase), and the radius of
capacitive loop increase obviously (capacitance increase), indicating that resistance Rs of concrete pore
solution (that is, the equivalent circuit of the R2) increases with the increase of fly ash content. And it
fits in with the literature [8] which described that the resistance Rs of concrete pore solution is
inversely proportional to the total porosity.

2500
1800
B1-3(0d) B1-3(36d)
A-1(0d) 1600 A-1(36d)
2000
C-2(0d) 1400
C-2(36d)

1200
1500
1000
Zim(ohms)

Zim(ohms)
1000 800

600

500 400

200

0 0

-200
0 500 1000 1500 2000 2500 -200 0 200 400 600 800 1000 1200 1400 1600 1800
Zre(ohms) Zre(ohms)

（a）EIS after 0d （b）EIS after 36d

2200
1600
2000 B1-3(48d) B1-3(60d)
A-1(48d) 1400 A-1(60d)
1800
C-2(48d) C-2(60d)
1600 1200

1400
1000
1200
Zim(ohms)
Zim(ohms)

800
1000

800 600

600
400
400
200
200

0 0

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 0 200 400 600 800 1000 1200 1400 1600

Zre(ohms) Zre(ohms)

（c）EIS after 48d （d）EIS after 60d

700
1000 B1-3(72d) B1-3(84d)
C-2(72d) C-2(84d)
600
800
500
Zim(ohms)
Zim(ohms)

600 400

300
400

200

200
100

0
0
0 200 400 600 800 1000 0 100 200 300 400 500 600 700

Zre(ohms) Zre(ohms)

（e）EIS after 72d （f）EIS after 84d

Note: The specimen A-1 rebar had been corroded since 72 days ago, so the EIS of specimen A-1 was
not given in fig. (e), fig. (f)

Figure 13. EIS of the specimens under different dosage of fly ash
Int. J. Electrochem. Sci., Vol. 10, 2015 5325

3.6 The impact analysis on the critical chloride ion concentration of fly ash content

When judging the rebar corrosion in concrete took place, the same method was used as before
to test the critical chloride concentration of rebar corrosion, the results as shown in Table 6. It can be
found that compared with group A, the critical chloride concentration of group B1 increase. It is
mainly because the compactness of specimens of group B1 with 10% fly ash, leads to the reduce of the
transmission speed of the Cl- , O2 and H2O in the concrete, but pH value reduces a little, so the critical
chloride concentration increases; And compared with group A, the compactness of group C increases,
but the pH value of pore solution in concrete reduces because of the mix of fly ash, causing less
chloride ion can break original chemical equilibrium, so the critical chloride concentration decreases.

Table 6. The critical chloride ion concentration under different fly ash content

Average
Percentage of Average value of
Cl-（mol/L Percentage of value
Specimens binding material binding material
） concrete (%)
(%)
of concrete
(%)
(%)
A-1 0.004108 0.364 2.039
A-2 0.003569 0.316 1.771 0.298 1.671
A-3 0.002423 0.215 1.203
B1-1 0.004108 0.364 2.039
B1-2 0.003326 0.295 1.651 0.318 1.780
B1-3 0.003326 0.295 1.651
C-1 0.003445 0.305 1.710
C-2 0.003211 0.285 1.594 0.282 1.579
C-3 0.002889 0.256 1.434

4. CONCLUSIONS

Ac impedance method can precisely manifest the critical corrosion point of rebar, the
Rp=300Ω/cm2 can be assumed to be the judgment standard of the corrosion critical point about
reinforced in concrete. The critical chloride concentration of concrete specimens increase with
increasing period of wet-dry cycling approximately. Small content of fly ash can increase the critical
chloride concentration of rebar corrosion in concrete, but when adding the content of fly ash, even if
the compactness of concrete is improved, the critical chloride concentration in concrete pore solution
will also reduce with the reduce of the pH value. With increasing period of wet-dry cycling, the degree
of drying of the specimens increases, and the resistance of the concrete layer increases, characterizing
the translation of spectra to the right on the complex plane; With the increase of fly ash content, the
compactness, the resistance and the capacitance of concrete increase, characterizing the translation of
the spectra to the right on the complex plane.
Int. J. Electrochem. Sci., Vol. 10, 2015 5326

ACKNOWLEDGEMENTS
This study was supported by the National Natural Science Foundation of China (Grant No. 51108442),
China Postdoctoral Science Foundation (Grant No.2012M521380), Shandong Postdoctoral Innovation
Foundation (Grant No. 201203106).

References

1. H. Böhni, Corrosion in Reinforced Concrete Structures, Wood head Publishing Limited,

Cambridge(2005)
2. U. Angst, B. Elsener, C. K. Larsen, Ø. Vennesland, Cem. Concr. Res., 39(2009)1122-1138.
3. K. Ann and H. Song, Corros. Sci., 49(2007) 4113-4133
4. M. Kouřil, P. Novák and M. Bojko, Cem. Concr. Res., 40(2010)431-436
5. J.X. Xu, L.H. Jiang, W.L. Wang and Y. Jiang, Constr. Build. Mater., 25(2011)663-669
6. J.X. Xu, L.H. Jiang and J.X. Wang, Constr. Build. Mater., 23(2009)1902-1908
7. U.M. Angst, A. Rønnquist, B. Elsener, C. K. Larsen and Ø. Vennesland, Corros. Sci.,
53(2011)177-187
8. Y. Li.Nanjing:Nanjing water conservancy scientific research institution,(2003).
9. Zh.H. Yang, M.L. Shi, J. Build. Mater., 2 (1999) 81-84.
10. S. G. Millard, D. Law, J. H. Bungey, NDT &E Int., 34(2001)409-417.
11. D.H. Hong, Beijing: China Railway Publishing House, 1998.
12. M.L. Shi, Zh.Y. Chen, J. Chin. Ceram. Soc., 25(1997) 241.
13. Ch. Xu , Zh.Y. Li and W.L. Jin, Corro. Sci. Prot. Tech., 5 (2011)393-398.
14. X.L. Zhang, Zh.H. Jiang, Zh.P. Yao, Y. Song and Zh.D. Wu, Corros. Sci.,51(2009)581-587.
15. U. M. Angst, B. Elsener, C. K. Larsen, Ø. Vennesland, Corros. Sci., 53(2011)1451-1464.

© 2015 The Authors. Published by ESG (www.electrochemsci.org). This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/4.0/).