Construction and Building Materials 144 (2017) 406–411

Contents lists available at ScienceDirect

Construction and Building Materials

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

Effects of self-healing cracks in bacterial concrete on the transmission

of chloride during electromigration
Hao Ling ⇑, Chunxiang Qian
College of Materials Science and Engineering, Southeast University, Nanjing 211189, China
Research Institute of Green Construction Materials, Southeast University, Nanjing 211189, China

h i g h l i g h t s

 The effects on transmission of chloride were test through multiple methods.

 The method of electromigration was used to accelerate the transmitting of chloride.
 The microbial self-healing cracks could impede the transmission of chloride.

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

Article history: The microbial self-healing concrete has been a new technology to repair cracks, the former study showed
Received 25 November 2016 that the precipitated CaCO3 could fill the cracks and reduce the permeability coefficient of cracks. On this
Received in revised form 26 February 2017 basis, the impacts on resisting transmission of chloride were studied to evaluate the protective effects of
Accepted 27 February 2017
microbial self-healing cracks, through multiple characterization methods such as electrochemical test,
visual examination of cracks surface, weight-loss ratio of reinforcements and chloride ion content.
Besides the method of electromigration was used to accelerate the transmission of chloride. The results
Keywords:
show that the microbial self-healing cracks can indeed impede the transmission of chloride in cracks and
Concrete
Self-healing
have protective effects for reinforcements in the concrete.
Chloride transmission Ó 2017 Elsevier Ltd. All rights reserved.
Electromigration
Reinforcement

1. Introduction
Concrete materials have the advantages of high compressive
strength, better fireproofing and durability, easiness in obtaining
materials [1], which have been widely used in many practical
projects like water conservancy and hydropower, traffic, industrial
and civil construction. They have become one of the most
extensively used civil engineering materials. But during the prepa-
ration and usage of concrete, the external loads and other factors
can result in loosing and spalling on the surface, seriously making
cracks in the concrete [2]. The cracks will reduce the capacities of
anti-permeability, anti-chloride-corrosion and anti-carbonization
greatly, which can make the corrosion of interior reinforcements
much easier and lower the carrying capacity and durability of
structure. If the repair of concrete cracks isn’t completed in time,
⇑ Corresponding author at: School of Materials Science and Engineering, South-
east University, Jiulonghu Campus, Nanjing 211189, China.
E-mail address: seulinghao@163.com (H. Ling).
it will affect the normal use of concrete structure, resulting in total
destruction and even collapse.
The traditional repairing methods are passive in most cases.
They have complex technology, high cost and even some destruc-
tive effects on the environment, which can’t satisfy the require-
ment of modern intelligence and multifunction for concrete
materials [3]. In 1995, Gollapudi et al. first proposed the method
of using biology progress to repair cracks and got a bran-new solu-
tion for the problem [4]. Microbial self-healing methods mainly
utilize the metabolism of bacteria to induce mineral precipitation.
They have the advantages of simple technology, low cost and free
of contamination, which have become the research hotspot [5–9].
Bacteria induced mineral precipitation has been proposed as an
alternative and environmental technique to improve the capacity
of concrete crack-healing in recent years [10–12]. For the self-
healing effects of microbial self-healing methods, Wiktor et al.
applied contrast test to the self-healing specimens and the control.
The results indicated microbial self-healing concrete had more
mineral precipitation and could repair wider cracks (470 lm),
while only 210 lm width of cracks was in the control [13]. Luo

http://dx.doi.org/10.1016/j.conbuildmat.2017.02.160
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
 H. Ling, C. Qian / Construction and Building Materials 144 (2017) 406–411 407

et al. prefabricated 0.3 mm cracks in the specimens. They found

that the microorganism repaired plenty of cracks in 5 days and
could fill the surface of cracks completely in 20 days [14]. In their
other researches, the bacterium loaded by pottery sand could
repair maximum width of 1 mm. They also found that maximum
CaCO3 were precipitated in the opening of cracks and mainly con-
centrated on the area of 1.5 mm depth from the surface of cracks
[15]. In addition, the changes of permeability coefficient before
and after the healing of cracks can characterize the remediation
effect. Wang et al. verified that microbial self-healing concrete
was able to reduce the permeability coefficient of cracks, especially Fig. 1. The schematic of making cracks by steel sheets.

in the initial 24 h [16]. In the research of Tittelboom et al., it also
showed that after using microbial self-healing agent to repair
cracks, the anti-permeability and durability of concrete specimens
were improved greatly [4]. It demonstrates that microbial self-
healing technology can make repairing effects on the concrete
cracks to some degree, but how much the real value of repairing
effects and the effect on resisting the transmission of chloride, still
need a further research.
In this study, we mainly used the previous microbial self-
healing technology of our research group to test the influence of
self-healing cracks on the transmission of chloride, then assessed
the actual contribution for protecting reinforcements. In order to
accelerate the transmission of chloride, we adopt the method of
electromigration. Besides, the reinforcements were imbedded in
the specimens to further study the protective performance under
the effects of self-healing cracks.
2. Materials and methods
2.1. Microbial self-healing agent
Microbial self-healing agent consisted of two parts: bacteria powder and sub-
strate. Bacteria power was Paenibacillus mucilaginosus. Cultivation of Paenibacillus
mucilaginosus was conducted in sucrose culture (10 g of sucrose and 3 g of sodium
hydrogen phosphate dissolved in deionized water to 1 L, and the pH value was
adjusted to about 7.0) at 35 °C for 24 h. Bacteria powder was made from the above
bacteria liquid by freeze drying process. The appropriate temperature of Paenibacil-
lus mucilaginosus growth is 10–40 °C, while the pH value is 7.0–9.0.
3 days. After that, the self-healing specimens were put into the 30 °C water and kept
21 days to ensure that the cracks would be healed effectively in the self-healing
environment and the control specimens were in the same environment for the con-
trast. Those specimens representing the cracks before being healed were put in the
standard curing room for 24 days at the same time.
2.4. Electromigration accelerated transmission of chloride
The method of using electromigration to accelerate the transmission of chloride
is connecting a DC electrical source between the chloride solution and the reinforc-
ing steel bar to produce an electric field among them. The chloride will invade into
the specimen more quickly under the effect of the electric field, achieving the pur-
pose of accelerating the transmission of chloride. We redesigned the electromigra-
tion device from other study [17]. The working device as shown in Fig. 2.
In the device, the reinforced specimens were immersed in the 3.5% NaCl solu-
tion (to simulate the seawater environment). The wires in the reinforced specimens
were linked with the positive pole of the DC source. The stainless steel mesh with
6 mm square hole was placed above the specimens and connected with the nega-
tive pole of DC source by the wire. During the process of electromigration acceler-
ation, the exterior voltage applied in the specimens would make abundant of heat.
The degree of fever depends on the applied voltage and the resistance of cement
materials. Those heat is harmful that can cause the transform of the microstructure
in the cement-based materials and make the specimens damaged in a certain
degree [18]. Therefore, the voltage of DC source used for electromigration acceler-
ation can’t be too high. Considering the problem of voltage and the test time, we
chose 15 V as the voltage of DC source in the study. Because the water evaporation
during the test may change the concentration of NaCl solution and lower the liquid
level, we should always check the amount of solution during the process of
electromigration. Keep the solution level above the stainless steel mesh to assure
an effective electric circuit between NaCl solution and the steel bar in specimens.
2.5. Multiple characterization methods to evaluate the effect on chloride transmission

2.2. Preparation of cement mortar specimens During the process of electromigration accelerating the transmission of chlo-
ride, the chloride reaching the surface of steel bar would induce the corrosion of
There were two kinds of cement mortar specimens, the microbial self-healing reinforcements constantly. Therefore we can monitor the condition of steel bar cor-
and the control. The microbial self-healing specimens were prepared by mixing rosion to assess the degree of chloride transmission. Electrochemical test method is
ordinary Portland cement, sand, self-healing agent and water. The mixing propor-
tion as shown in Table 1. The control specimens had no self-healing agent so with-
out subjection to bio-deposition. The dimensions of specimens were all
50 mm  50 mm  200 mm. There was a 10 mm diameter reinforcing steel bar
with acid dipping treatment imbedded in the middle, the ends of the steel bar
and the location of linking wires were sealed with epoxy resin.

2.3. Creation of artificial cracks and curing condition

In order to simulate the real situation of concrete cracks and reduce the influ-
ence due to the difference of cracks, we used the method of presetting the steel
sheets in the mortar while making the specimens. The width of steel sheets was
40 mm, thickness was 0.2 mm and they must inserted vertically in the mortar with
the depth of 20 mm and separation distance of 40 mm, as shown in Fig. 1. Take off
the steel sheets at the final setting time of mortar, remove the form after one day
and then put the specimens into standard curing room (20 ± 2 °C, RH > 95%) for

Table 1
Composition of self-healing specimens.

Material kg/m3
Ordinary Portland Cement 600
Water 300
Sand 1800
Substrate 12
Bacteria powder 1.2
w/c 0.5
Fig. 2. The device of electromigration accelerated transmission of chloride.
408 H. Ling, C. Qian / Construction and Building Materials 144 (2017) 406–411
nondestructive analysis that can detect the corrosion degree. The final corrosion
rate can be got by testing the weight-loss ratio of reinforcements that were taken
out from decomposed specimen. Furthermore, observing the condition of corrosion
products spilled from the cracks can also estimate the corrosion degree to some
extent.
In this experiment, the measurement system used for electrochemical test
method was three-electrode system, including 10 mm reinforcement for working
electrode, SCE (saturated calomel electrode) for reference electrode and the stain-
less steel sheet for auxiliary electrode. The adoptive measure method was linear
polarization method, derived from the Stern and Geary in 1957 and it’s a simple,
quick and effective non-destructive test methods. The theory is connecting the cor-
rosion ratio with polarization curve at the point of corrosion potential. When the
overpotential is very little (g < 10 mV), it will have a linear relation with polariza-
tion current [19].
In the actual data processing, the numerical value of corrosion current Icorr and
self-corrosion potential Ecorr can be obtained from the electrochemical software
VersaStudio by RP fitting. Ecorr is the stable potential that corrosion system can’t
be applied by the additional polarization. The more negative value of Ecorr, the more
corrosion tendency. On the contrary, more positive value of Ecorr can make less ten-
dency of corrosion and better corrosion resistance [20]. Besides, the changes of self-
corrosion potential can reflect the degree of reinforcement corrosion. Some other
studies have pointed that the more corrosion current and negative value of self-
corrosion potential indicate the more serious degree of reinforcement corrosion
[21]. The single type of data can’t reveal the reinforcement corrosion accurately
because the results of electrochemical test always possess some volatility and devi-
ation for the real reinforcement corrosion state. As a result, this study combined
corrosion current data with self-corrosion potential to reflect the corrosion state
of reinforcement more accurately.
Testing weight-loss ratio can measure corrosion of reinforcements accurately.
The reinforcements should be dealt with acid pickling to wash off the corrosion
products on the surface. During the acid pickling, put two steel bars that have the
same dimension and non-stain to work as calibration of blank value. When acid
pickling was accomplished, dry and weigh the reinforcement.
Otherwise the chloride transmission degree can be characterized by measuring
the chloride content in the specimens. When drilling hole to get powder used for
measuring the chloride content, the location of those drilling hole were the middle
of the specimens surface and the distance from the right side were 60 mm, 100 mm
and 140 mm respectively, which could avoid the chloride immersing from the both
ends exerting an influence on the test results. And the depths of drilling hole are
5 mm, 10 mm and 15 mm respectively, as shown in Fig. 3. The powder in the three
drilling holes should be put together to assure the quantity of powder in each depth
were enough.
3. Results and discussion
3.1. Electrochemical test results
During the process of electromigration accelerating transmis-
sion of chloride, the changes of reinforcement corrosion current
and self-corrosion potential, along with the differences of before
and after cracks healing are shown in Fig. 4. We can find that the
corrosion current of the self-healing specimens and the control
specimens are both lower when the cracks were healed. Especially,
the self-healing specimens has a lower level than the control. From
the changes of self-corrosion potential, there are more obvious
results. During the whole electromigration process, the negative
value of self-corrosion potential in the control specimens is higher
than that of the self-healing specimens, although both of them are
lower than those results of cracks without being healed. In the sit-
uation of BH (before healing), there aren’t too much differences.
Those results indicate that the corrosion current and the self-
corrosion potential can be reduced after the cracks were healed.
The cracks in the self-healing specimens may be blocked by the
Fig. 3. The drilling holes in the specimen.
deposited CaCO3 from microbial mineralization, which hindered
the chloride from invading into the reinforcement surface through
the cracks and slowed down the process of reinforcement corro-
sion. And the control specimens containing the cement also had
the ability to heal the crack in the early period, which could reduce
the corrosion. However, it is not as good as the self-healing
specimens.
3.2. Visual examination of cracks surface
During the process of electromigration, it’s easy to find that the
corrosion products exceeded form the cracks because of the rein-
forcement corrosion. So, it’s a way to evaluate the degree of rein-
forcement corrosion. In Fig. 5, the change of cracks surface
stained by rust become more and more serious in the control spec-
imens, so the rust stain are increasing constantly. For the self-
healing specimens, the degree of cracks surface corrosion is always
slight during the electromigration time. Only after 14 h, we can
find obvious corrosion exceeding from the cracks. It indicates that
the healing of cracks in the self-healing specimens could slow
down the transmission of chloride during the process of electromi-
gration, mainly because the self-healing products filled the cracks,
hindered the transmission of chloride in the specimens and pro-
tected the reinforcement.
3.3. Weight-loss ratio of reinforcement
After electromigration was finished, the corrosion situation at
the ends of reinforcement was very serious. It may because protec-
tive effect of the ends was the weakest. The corrosion products
heaped up instantly during the test and destroyed the protective
layer of the ends, so the erosion of chloride became easy and fast.
The corrosion of the ends had nothing to do with the cracks, so we
should avoid the corrosion affecting on the reinforcement weight-
loss ratio. In this study, we took out the reinforcement in the spec-
imens and cut out the middle 120 mm section for the weight-loss
ratio test, as shown in Fig. 6. It can exclude the effect of the corro-
sion at the ends and the effect of no corrosion because of the epoxy
resin protective layer at the touch point of the wire. The chosen
120 mm sections of reinforcement were mainly focus on the area
of cracks. It could reflect the influence of healed cracks on the pro-
tective effect of reinforcement.
Although the initial quality of reinforcement in the chosen sec-
tion is unknown, we can get that by calculational methods, mainly
depending on two parameters of length and linear density. The lin-
ear density can be calculated by the three standard reinforcements.
The final results and other parameters as shown in Table 2. The
average linear density is 5.72 g/cm.
Take out the reinforcement in the specimen when the electro-
migration was finished and choose the middle section. The final
results of average corrosion ratio of reinforcements are shown in
Fig. 7. It can be found that in the state of reinforcement corrosion
after 18 h of electromigration, the average weight-loss ratios of
self-healing specimens and the control are both lower than that
of before healing. The self-healing is lower than the control speci-
mens by 0.55(%). This result is consistent with the former results of
corrosion current and self-corrosion potential in the electrochem-
ical test and visual examination of cracks surface. It indicates the
cracks healed by the microbial self-healing agent can better hinder
the transmission of chloride in the state of electromigration, slow
down the corrosion of inner reinforcement and have preventive
effects. We can also find that the weight loss of reinforcements
in the self-healing specimens is lower than that of the control
before the healing. The reason may be that the self-healing speci-
mens also made the healing in the curing room to some degree
and it’s difficult to avoid it.
 H. Ling, C. Qian / Construction and Building Materials 144 (2017) 406–411 409

Fig. 4. The changes of corrosion current and self-corrosion potential with electromigration time (BH: before healing; AH: after healing).

Fig. 5. Visual examination of cracks surface during the electromigration time.

Fig. 6. The chosen section of reinforcement in the weight-loss ratio test.

Table 2
Linear density and other parameters of standard reinforcement.

Number ① ② ③
Length (cm) 11.88 11.92 11.80
Quality (g) 67.945 67.938 67.781
Linear density (g/cm) 5.72 5.70 5.74
Fig. 7. The average corrosion ratio of reinforcements.
410 H. Ling, C. Qian / Construction and Building Materials 144 (2017) 406–411

3.4. The test of chloride content

The test results of average chloride content in different depths

at the end of 18 h electromigration are shown in Fig. 8. It can be
found that all the specimens after healing have lower chloride con-
tent than those without healing. Chloride content in the self-
healing specimens is also lower than the control before healing.
For the situation that cracks had been healed, in the depth of
5 mm, the average chloride content of self–healing specimens is
a little higher than that of the control, which may result from the
self-healing agent added into the matrix changing the compact-
ness. But with the depth increasing, the chloride contents of
10 mm and 15 mm are both lower in the self-healing specimens.
The changes may relate to the healing of cracks, but still need fur-
ther analyses. So in this study, we simply simulated the chloride
transmission before and after the healing of cracks.
For the chloride transmission, we mainly considered one- Fig. 9. The simulation of chloride transmission in the crack: (a) self-healing
specimens; (b) control specimens.
dimensional oriented transmission from exterior to interior in
the cracks when simulating, the method was finite element simu-
lation by ANSYS and the other major parameters were set as fol-
4. Conclusions
lows. The initial concentration of chloride C0 = 0, exterior
concentration Ce = 35 kg/m3 (means the chloride solution of 3.5%
During the process of electromigration accelerating transmis-
mass fraction that simulated the seawater environment). The diffu-
sion of chloride, the results of corrosion current and self-
sion coefficient of chloride in the concrete D0 = 5.5  10 12 m2/s
corrosion potential by electrochemical test show that the corrosion
was got from the RCM test. In the research of Ismail et al.,
degree of reinforcements is lower when the cracks were healed by
0.2 mm crack was not a limiting factor controlling the diffusion
self-healing agent.
process perpendicular to the crack wall and had no influence on
The visual examination shows less corrosion products overflow-
the chloride transmission [22]. So the diffusion coefficient of chlo-
ing from the cracks and the lower weight-loss ratio of reinforce-
ride in the cracks Dcr = DS = 2.58  10 9 m2/s, that was the same as
ments that were took out from the self-healing specimens by
the diffusion coefficient in solution. For the area of healed cracks,
destructive test method. The mainly reason is the healed cracks
the compactness of healing substance was lower than the concrete,
can hinder the chloride form invading into the inner of specimens
so the diffusion coefficient of chloride was higher, set as
by the cracks, thus slow down the process of reinforcements
Dh = 5.5  10 11 m2/s. The healing depth of cracks could reach
corrosion.
about 1.5 mm in the previous study [15]. Total simulation time
In addition, the test results of chloride content in different
was 10 days and the final results as shown in Fig. 9.
depths verify that chloride content in the self-healing specimens
In Fig. 9, the chloride mainly diffuses from the crack and has a
is lower after the cracks were healed. It can indicate the lower
peak distribution. When the crack was healed in the self-healing
transmission degree of chloride accurately.
specimens, the diffusion depth is shallower than that of the control
Through those multiple kinds of characterization methods, we
and the chloride concentration is lower in the area of crack. So, we
can get that the microbial self-healing applied on the concrete
can conclude that the lower results of chloride concentration of
cracks can impede the transmission of chloride in the cracks, which
self-healing specimens in 10 mm and 15 mm as shown in Fig. 8
has a directly protective effect on reinforcements. This study also
is related to the healing of cracks. The healing cracks could hinder
shows good application values of microbial self-healing technique
the chloride transmission in the cracks, thus decreasing the inner
used in the practical construction and provides a new approach to
chloride concentration in the specimens.
reinforce the durability of structure.

Acknowledgments

Authors would like to appreciate the financial support from

National Nature Science Foundation of China (Grant No.
51178104) and Scientific Research Foundation of Graduate School
of Southeast University.

References

[1] K. Flaga, Advances in materials applied in civil engineering, J. Mater. Process.

Technol. 106 (1) (2000) 173–183.
[2] R. Wang, C. Qian, Restoration of defects on the surface of cement-based
materials by microbiologically precipitated CaCO3, J. Chin. Ceram. Soc. 36 (4)
(2008) 457.
[3] H. Guo, Y. Zhang, et al., Crack repair and surface deposition of cement-based
materials by MICP technology, J. Ind. Constr. 45 (7) (2015) 36–37 (in chinese).
[4] U.K. Gollapudi, C.L. Knutson, S.S. Bang, et al., A new method for controlling
leaching through permeable channels, Chemosphere 30 (4) (1995) 695–705.
[5] W. De Muynck, K. Cox, N. De Belie, et al., Bacterial carbonate precipitation as an
alternative surface treatment for concrete, Constr. Build. Mater. 22 (5) (2008)
875–885.
[6] Z. Yang, J. Hollar, X. He, et al., A self-healing cementitious composite using oil
core/silica gel shell microcapsules, Cem. Concr. Compos. 33 (4) (2011) 506–
Fig. 8. The average chloride content in different depths. 512.
 H. Ling, C. Qian / Construction and Building Materials 144 (2017) 406–411
[7] Y. Yang, M.D. Lepech, E.H. Yang, et al., Autogenous healing of engineered
cementitious composites under wet–dry cycles, Cem. Concr. Res. 39 (5) (2009)
382–390.
[8] Y. Yang, E.H. Yang, V.C. Li, Autogenous healing of engineered cementitious
composites at early age, Cem. Concr. Res. 41 (2) (2011) 176–183.
[9] V.C. Li, E. Herbert, Robust self-healing concrete for sustainable infrastructure, J.
Adv. Concr. Technol. 10 (6) (2012) 207–218.
[10] S.S. Bang, J.J. Lippert, U. Yerra, et al., Microbial calcite, a bio-based smart
nanomaterial in concrete remediation, Int. J. Smart Nano Mater. 1 (1) (2010)
28–39.
[11] C.X. Qian, Q.F. Pan, R.X. Wang, Cementation of sand grains based on carbonate
precipitation induced by microorganism, Sci. China Technol. Sci. 53 (8) (2010)
2198–2206.
[12] W. De Muynck, N. De Belie, W. Verstraete, Microbial carbonate precipitation in
construction materials: a review, Ecol. Eng. 36 (2) (2010) 118–136.
[13] V. Wiktor, H.M. Jonkers, Self-healing of cracks in bacterial concrete, in: K. van
Breugel, G. Ye, Y. Yuan (Eds.), 2nd International Symposium on Service Life
Design for Infrastructures, RILEM Publications SARL, 2010, pp. 825–831.
[14] M. Luo, C. Qian, R. Li, Factors affecting crack repairing capacity of bacteria-
based self-healing concrete, Constr. Build. Mater. 87 (2015) 1–7.
[15] C. Qian, R. Li, Q. Pan, Microbial self-healing effects of concrete cracks, J.
Southeast Univ. Nat. Sci. Ed. 43 (2) (2013) 360–364.
411
[16] J.Y. Wang, N. De Belie, W. Verstraete, Diatomaceous earth as a protective
vehicle for bacteria applied for self-healing concrete, J. Ind. Microbiol.
Biotechnol. 39 (4) (2012) 567–577.
[17] V. Achal, A. Mukherjee, S. Goyal, et al., Corrosion prevention of reinforced
concrete with microbial calcite precipitation, ACI Mater. J. 109 (2) (2012) 157–
164.
[18] Z. He, C. Shi, X. Hu, et al., Development on migration characteristic and
interactions of chloride ion in cement-based materials under applied voltages,
J. Chin. Ceram. Soc. 8 (2015) 016.
[19] W. Morris, A. Vico, M. Vázquez, Chloride induced corrosion of reinforcing steel
evaluated by concrete resistivity measurements, Electrochim. Acta 49 (25)
(2004) 4447–4453.
[20] S. Qiu, F. Cao, Z. Sun, et al., Applications of potentiodynamic polarization
technique in corrosion behavior measurements of prosthodontic alloys, Acta
Acad. Med. Militaris Tertiae 23 (2) (2001) 228 (in Chinese).
[21] J. Liu, L. Lv, Z. He, Evaluation of corrosion degree of reinforcing bar based on
corrosion current, in: Tenth National Conference on Cement and Concrete
Chemistry and Application Technology, 2007, p. 1 (in chinese).
[22] M. Ismail, A. Toumi, R. François, et al., Effect of crack opening on the local
diffusion of chloride in cracked mortar samples, Cem. Concr. Res. 38 (8) (2008)
1106–1111.