Construction and Building Materials 200 (2019) 36–45

Contents lists available at ScienceDirect

Construction and Building Materials

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

Activated chemicals of cementitious capillary crystalline waterproofing

materials and their self-healing behaviour
Guangyan Li a, Xiaofeng Huang a, Jiesheng Lin b, Xiang Jiang a,⇑, Xinya Zhang a,⇑
a
School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510641, China
b
Guangdong Carpoly Greenway Technology Materials Co. Ltd., Jiangmen 529085, Guangdong, China

h i g h l i g h t s

Sodium carbonate, sodium silicate, sodium aluminate, tetrasodium EDTA and glycine proposed as activated chemicals of CCCW.

The optimal proportion of activated chemicals of CCCW was achieved through permeability and pre-pressure self-healing test.

White crystals with an average diameter of 0.2 mm were generated in the self-healing process of CCCW.

The self-healing mechanism of CCCW was proposed based on the self-healing experimental results.

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

Article history: Activated chemicals play a key role in the waterproof performance of cementitious capillary crystalline
Received 21 January 2018 waterproofing materials (CCCW). We proposed sodium carbonate, sodium silicate, sodium aluminate,
Received in revised form 14 December 2018 tetrasodium EDTA and glycine as activated chemicals of CCCW. The proportion of activated chemicals
Accepted 15 December 2018
of CCCW was optimized by two key parameters evaluation of the second permeability ratio and
Available online 21 December 2018
pre-pressure self-healing rate for their self-healing ability of CCCW. Results show that sodium silicate
has the most significant impact on the self-healing ability of CCCW. In the self-healing progress of
Keywords:
CCCW, the consumption of Ca(OH)2 (CH) and the production of Calcium Silicate Hydrate (CSH) gel,
Cementitious capillary crystalline
waterproofing
calcium carbonate and ettringite were observed through XRD and SEM analysis. The white crystals with
Activated chemicals an average diameter of 0.2 mm are produced on both sides of cracks, and consequently the cracks were
Self-healing healed by the accumulation of crystals. The self-healing mechanism of CCCW was proposed based on the
Permeability self-healing experimental results.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction nomic losses [3,5]. According to De Rooij and Van Tittelboom [6],
the UK spends 45% of the annual construction costs to maintain
Concrete is widely used as construction materials due to its the existing concrete structure. It is quite imperative to improve
high durability and low cost. But the unexpected cracks caused the durability of concrete for both economic and social benefits.
by construction loading or shrinkage remain a challenge during Conventional ways to repair concrete are using polymer mate-
varying temperature, commonly leading to a decrease on concrete rials as waterproof roll [7] or injecting materials that coagulate
strength and exposing the interior of concrete exposed to the envi- rapidly by grouting, which requires considerable efforts and
ronment [1]. And concrete is porous, consisting of tunnel-like cap- money. And they do not always work well because locating the
illaries, which give access to water and other unwanted chemicals. leakage is not easy [8]. Nowadays, self-healing materials are
As a consequence, concrete is susceptible to chemical attack, prob- attracting plenty of attentions and have been applied in many
ably ending up with the corrosion of the inside steel bars [2,3]. In fields [9–12]. With self-healing ability, the number and size of
addition, larger cracks are likely to appear over time and would racks could be reduced, which makes concrete more durable
form a path for water outside, which not only impair the durability [13,14].
of concrete, but also destroy the water tightness of the cement There are many approaches to achieve self-healing functionality
structure [4]. The damage of cracks in buildings caused severe eco- such as using shape memory materials [15], proper bacteria [3] and
microcapsules [13]. Unfortunately, shape memory materials cost
⇑ Corresponding authors. too much. Utilizing bacteria is not always effective in the case of
E-mail addresses: jiangx@scut.edu.cn (X. Jiang), cexyzh@scut.edu.cn (X. Zhang). multiple damage events. Environmental problems, especially

https://doi.org/10.1016/j.conbuildmat.2018.12.093
0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.
 G. Li et al. / Construction and Building Materials 200 (2019) 36–45 37

drinking water pollution, cannot be ignored when using bacteria or 2. Experimental program and methodology
microcapsules.
Neville [16] mentioned the self-healing behavior of concrete. 2.1. Experimental program
The general Portland concrete consists of 2CaOSiO2 (C2S),
3CaOSiO2 (C3S), 3 CaOAl2O3 (C3A), 4 CaOAl2O3Fe2O3 (C4AF), CCCW is a kind of rigid waterproof material used for cement
SiO2, Al2O3, Fe2O3, MgO, SO3, Na2O, K2O, TiO2, P2O5, CaSO4, concrete. After interaction with water, the active chemicals con-
CaSO42H2O, CaSO40.5H2O and CaO. When water is added into tained in CCCW penetrate the concrete by water as the carrier,
the general Portland cement, C2S and C3S hydrate then produce and acicular crystals insoluble in water with cement hydration
CSH gel and CH. C3A reacts with the set retarder gypsum and water products are generated, consequently filling capillary pores and
as follows: micro-cracks, thereby improving the compactness and waterproof-
ing of concrete. In order In order to achieve the best waterproof
3CaO  Al2 O3 þ 3ðCaSO4  2H2 OÞ þ 26H2 O effect by the activated chemicals, the ratio of the active chemicals
! 3CaO  Al2 O3  3CaSO4  32H2 O ð1Þ in CCCW should be optimized.
At present, there are few evaluation criteria for self-healing per-
Shortly after mixing, the formed ettringite layer covers the sur- formance of CCCW. During the preparation process of tested
face of the cement grains [17,18]. Thus there is always unreacted blocks, the manual insertion and ramming process and the com-
cement left. On the other hand, the crystallization of calcium car- plexity of cement itself, the performance of different batches of
bonates contributes for the self-healing process [19–22]. Calcium cement blocks might be varied, thus the tested blocks of different
silicates could react with CO2 to generate calcium carbonates and groups could not be completely consistent. A simple comparison of
silica gel, which blocks cracks and results in low porosity and high first and second permeability may cause serious deviation. There-
density of concrete [23,24]. CH, the hydration product of cement, fore, appropriate adjustment of evaluation parameters is essential.
provides free calcium ions that react with CO2 to form calcium car- The first evaluation index was set as the ratio of second permeabil-
bonates on both sides of cracks and fill the gaps eventually [15]. It ity, which was the ratio between the measured value of second
is believed that the formation of calcium carbonates is the main permeability of CCCW and permeability of this group of standard
reason for later self-healing of concrete [25]. In general, self- blocks for 14d curing periods (which is also called as the first
healing process could be triggered by the expansion of hydrated permeability).
cementitious matrix, the formation of calcium carbonates crystals, Corresponding to second permeability ratio, the pre-pressure
insoluble impurities present in water and the further hydration of self-healing rate is noted as another evaluation index reflecting
unreacted cement [26]. the self-healing ability of CCCW. The difference between perme-
Considering on the above, CCCW, based on the development of ability and pre-pressure test is that the pre-pressure test would
crystallization technology and thoroughly understanding of con- bring relatively large cracks with irregular connectivity. Thereby
crete’s chemical and physical makeup, satisfies economic and envi- the pre-pressure self-healing rate and second permeability ratio
ronmental requirements at the same time. Researchers realize the together serve as the evaluation index of self-healing performance
possibility to fill these cracks and capillaries to prevent the pene- of CCCW.
tration of water and other liquids with a chemical treatment Due to the particularity of concrete material itself, impervious
[26]. By diffusion, the activated chemicals in CCCW enter with tests would consume a lot of time and materials, and pre-
water as a migration medium and move down the capillaries or pressure tests would also consume a lot of time and materials.
cracks of concrete. This process facilitates the chemical reaction Orthogonal experiment is a multi-factor and multi-level experi-
between CCCW, moisture and the byproducts of cement hydration, mental method. Through orthogonal experiments, some represen-
forming a new insoluble crystal structure. This integral structure tative points could be selected from the overall experiments, thus
filling the capillary tracts endows concrete the ability of water- greatly reducing the workload of the experiments [30].In this
resisting. CCCW shortens inspection and maintenance time by paper, orthogonal experiment is used to optimize the activated
self-healing before the leakage occurs. Meanwhile, there is no need chemicals.
to worry about the environmental problems triggered by the Choose active chemicals basic formula of sodium carbonate,
byproducts of bacteria or microcapsules. sodium silicate, sodium aluminate, tetrasodium EDTA and glycine
Although amounts of test results have demonstrated the as five factors obtained by previous exploratory experiment.
advantages of CCCW, it is an agreed fact that the major compo- Among these activated chemicals, sodium carbonate, sodium sili-
nent or proportion has not been fully reported in literatures cate are not easy to cause side effects, and thus their amounts
[13,27–29]. CCCW contains cement, activated chemicals, aggre- could be slightly larger. Excessive sodium aluminate would seri-
gate, water reducer and other additives. The self-healing ability ously affect the performance of concrete. Large amount of tetra-
of CCCW could be influenced by the aggregate-cement ratio and sodium EDTA might do harmful for the self-healing effect, and
water-cement ratio (W/C ratio). However, it is the activated glycine cost is relative higher. Based the above, each of the 5 active
chemicals that determine the self-healing ability of CCCW, which chemical factors was set four levels in the following amount range:
means with activated chemicals, CCCW could be applied in all sodium carbonate of 0%–5.0%, sodium silicate of 1.0%–6.0%, sodium
cementitious circumstances to achieve an outstanding waterproof aluminate of 1.0%–4.0%, tetrasodium EDTA of 0%–0%, and glycine of
capability despite of its components, aggregate-cement ratio or 0–2.0%.The contents of the active chemicals are calculated as the
W/C ratio. mass percentage of the cement, and replaced the cement by the
In this research, sodium carbonate, sodium silicate, sodium same amount. The orthogonal table of L16(45) was designed, and
aluminate, tetrasodium EDTA and glycine were proposed as the factor level table was shown in Table 1.
activated chemicals of CCCW. The second permeability and
pre-pressure self-healing tests were conducted to evaluate the
self-healing behavior and further optimize the proportion of acti- 2.2. Permeability test
vated chemicals of CCCW. Based on analysis of the self-healing
process by X-ray diffraction (XRD) and scanning electron Each group contains six pieces of standard mortar blocks
microscopy (SEM), the self-healing mechanism of CCCW was (70  80  30 mm) prepared simultaneously with the cement con-
proposed. tent of 17.56% and the water to cement ratio of 0.6 (that is a
38 G. Li et al. / Construction and Building Materials 200 (2019) 36–45

Table 1
Factors and levels for orthogonal test.

Levels Factors
Sodium carbonate (%) Sodium silicate (%) Sodium aluminate (%) Tetrasodium EDTA(%) Glycine (%)
1 0 1.0 1.0 0 0
2 1.0 2.0 2.0 1.0 0.5
3 3.0 4.0 3.0 2.0 1.0
4 5.0 6.0 4.0 3.0 2.0

standard cement mortar proportion of P.O 42.5R ordinary Portland 2.3. Pre-pressure test
cement of 330 g, ISO standard sand of 1350 g, water of 198 g and
cellulose of 0.5 g). To carry out the experiments, we coated CCCW Corresponding to the permeability test, there is a pre-pressure
on four random mortar blocks each group and another two blocks test. The pre-pressure test is an experiment that reflects the self-
as blank samples. The CCCW with a total amount of 1.5 kg/m2 was healing ability of CCCW regarding the healing efficiency of com-
coated again when the first coating film was dried to touch. After pressive strength. But too much pressure could cause large cracks
the second coating, these specimens were moved to the standard or even smashed in the concrete. Generally, concrete is a kind of
conservation box (Yh-40b standard constant temperature and brittle material, when the ambient pressure is strong enough, large
humidity curing box, Hebei Shuangxin testing instrument manu- cracks are likely to generate in normal concrete or even broken of
facturing co., LTD, China) curing in water under the condition of concrete. In this section, fiber reinforced concrete was selected as
20 °C and humidity of 90% for 14 days. With drying the surface of the standard specimen for the pre-pressure experiments. Fiber lim-
the cured specimens, the side of the specimens is coated with a its the local fracture of concrete, thus overcoming the brittleness
melted sealing material (sealing wax) adhered to the permeable and low strain capacity of concrete to a certain extent [31], result-
molds. It is note that the two ends of the specimen could not apply ing in more micro-cracks [32]. These characteristics allow fiber
any sealing material. These specimens were then installed on the reinforced concrete to obtain more reliable data in pre-pressure
permeable mortar instrument (Digital display mortar penetrome- tests.
ter SJS-15, Shanghai meiyu instrument equipment co., LTD, China.) Some studies have pointed out [33] that when the ratio of the
as shown in Fig. 1. The water pressure of the permeable mortar received pressure load r to the maximum pressure load r0 is less
instrument was set at 0.2 MPa and hold for 2 h. Then the water than 0.3, there are virtually no new cracks inside the concrete.
pressure was increased by 0.1 MPa per hour until the mortar When r/r0 is between 0.4–0.5, new cracks are beginning to
blocks leaked. The final value of water pressure minus 0.1 MPa appear and expand. When the ratio is between 0.7–0.8, the cracks
was marked down as the first permeability pressure of the exper- continue to spread and intersect with the increasing pressure load.
imental group or the blank group. It is noted that the permeability Large cracks would form when r/r0 = 0.9 and the specimen would
test was invalid if the permeability of blank samples is not of be damaged when r/r0 = 1. Based on this, in this pre-pressure
0.3 MPa or 0.4 MPa. Then this group of permeable specimens was experiments, 0.6r0 pressure load on fiber reinforced concrete
moved to the standard conservation room soaking in water for could give rise to micro cracks and avoid larger cracks generated
another 14 days with its permeable molds. The same permeability which have serious impact on mechanical properties of mortar.
test was proceeding again to get the second permeability of the In this part, three pieces of 40  40  160 mm fiber-reinforced
experimental group and the blank group. Thus the permeability concrete specimens were prepared according to the ISO 679-
ratio can be calculated as following: 2009 standard ‘‘Cement Test methods Determination of strength”
at the same time. The mixing proportions of the fiber-reinforced
S1 concrete used are given in Table 2. For the fiber-reinforced con-
SP1 ¼  100%
S0 crete, P.O 42.5R ordinary Portland cement, sand with the average
particle size of 180 lm, 12 mm polypropylene fibers provided by
S2 Changsha Bo Sai Te Construction Materials Co., Ltd, water reducing
SP2 ¼  100%
S0 agent (MELMENTÒ F10) from BASF SE were used. The activated
chemicals were used to substitute the same quantity of cement.
wherein, SP is the permeability ratio (%), S2, S1 are the second and
The 16 different proportion of activated chemicals in CCCW
first permeability for mortar blocks with CCCW, and S0 is the per-
numbered as E1-E16 shown in Table 3 were used to prepare the
meability for blank samples.

Fig. 1. Permeability test instrument and specimen.

 G. Li et al. / Construction and Building Materials 200 (2019) 36–45 39

Table 2
The Mixing proportions of the fiber-reinforced concrete (kg/m3).

Materials Cement Aggregate Water Water reducing agent Fiber

Fiber-reinforced concrete 685 730 411 2 14

Table 3
L16(45) orthogonal array testing for permeability of CCCW.

NO. Sodium Sodium Sodium Tetra- sodium Glycine First permeability Second permeability Second permeability
carbonate silicate aluminate EDTA ratio (%)
Blank Blocks with Blank Blocks with
CCCW CCCW
E1 1 1 1 1 1 0.5 0.9 0.5 0.8 160.0
E2 1 2 2 2 2 0.3 0.6 0.3 0.8 266.7
E3 1 3 3 3 3 0.3 0.5 0.4 0.7 175.0
E4 1 4 4 4 4 0.4 0.6 0.5 0.7 140.0
E5 2 1 2 3 4 0.4 0.6 0.4 0.7 175.0
E6 2 2 1 4 3 0.3 0.9 0.3 0.9 300.0
E7 2 3 4 1 2 0.3 0.6 0.3 0.6 200.0
E8 2 4 3 2 1 0.3 0.3 0.4 0.7 175.0
E9 3 1 3 4 2 0.3 0.8 0.4 0.9 225.0
E10 3 2 4 3 1 0.3 0.6 0.3 0.7 233.3
E11 3 3 1 2 4 0.4 0.8 0.4 0.9 225.0
E12 3 4 2 1 3 0.4 0.7 0.5 0.8 160.0
E13 4 1 4 2 3 0.4 0.6 0.4 0.7 175.0
E14 4 2 3 1 4 0.3 0.6 0.3 0.7 233.3
E15 4 3 2 4 1 0.3 0.6 0.4 0.7 175.0
E16 4 4 1 3 2 0.3 0.8 0.4 0.9 225.0

specimens. The fresh fiber-reinforced concrete specimens were carried out with the orthogonal array L16(45) as shown in Table 3.
demolded after 24 h and moved to the standard conservation room 16 groups of mortar blocks with CCCW, along with the correspond-
curing for 28 days. After curing, the specimens were first broken ing blank mortar blocks were prepared. The second permeability
into two parts using the rupture strength tester. A random part ratio is shown in the last column of Table 3, the first and second
of each fiber-reinforced concrete specimen was chosen to conduct permeability is shown simultaneously for better understanding.
the pre-pressure test. The resulted value was noted as the original It can be seen from Table 3, the first permeability and secondary
compressive strength P0. Then the 0.6 P0 was pre-pressured on permeability of CCCW are more than that of blank cement. How-
another part of the specimen and held for 1 min. The pre- ever, a high first permeability groups do not correspond to the
pressured part was moved back to the standard conservation again ground with the high secondary permeability. This is because the
and soaking in water for 28 days. After the curing process, the pre- first permeability of CCCW is more related to the surface compact-
pressured part was used for the pre-pressure test again to achieve ness resulted from the film on the basic block, but the secondary
the compressive strength P2. The pre-pressure self-healing rate is permeability is mainly impacted by the self-healing effect of the
defined as the ratio of the compressive strength P2 after 28 days block of waterproof coating on the basic block. Therefore the first
curing to the original compressive strength correspondingly P0 permeability and second permeability have no strict corresponding
and calculated as following: relation. This is one reason why we choose the second permeability
P2 ratio as a parameter for evaluating self-healing ability of CCCW.
PR ¼  100% Based on Table 3, the influence of each activated chemical on
P0
second permeability ratio and their preferred values could be
obtained through orthogonal analysis shown in Table 4. In the
2.4. Characterization orthogonal analysis, K1-K4 represent the corresponding second
permeability ratio at each level of activated chemicals, R denotes
XRD analysis of crystals scraped off from the cracks was carried the difference between the largest value and the smallest of K1-
out on Bruker D8 ADVANCE diffractometer. The samples were K4. The activated chemical possessing the larger R has more signif-
scanned from 5-to-80° (2h) using a D/max-III type automatic X- icant influence on second permeability ratio. Therefore we rank the
ray diffractometer with Cu-Ka radiation. The diffractometer has a influence order by R value that sodium silicate (83.3) > glycine
wavelength of 1.54 Å and an integrated step scan of 0.017° (2h). (43.5) > sodium aluminate (40.5) > sodium carbonate (27.0)
Crystals scraped off from the cracks were investigated by scanning > tetrasodium EDTA (22.3).
electronic microscopy (SEM, ZEISS Merlin, Germany). Specimens According to Table 4, the influence of the level of activated
were fixed by a conducting resin and then coated with gold and chemical on second permeability ratio is illustrated in Fig. 2. It
examined by SEM at 5 kV. Optical microscope images of the crys- shows that through adding sodium carbonate and glycine, CCCW
tals were captured by 80X optical microscope (MG10085-2, China). possesses higher second permeability ratio. When the proportion
of sodium carbonate reaches 1.0%, or the proportion of glycine
3. Results and discussion comes to 0.5%, it has the highest second permeability ratio of
212.5% or 229.3%. When the proportion of sodium carbonate or
3.1. Orthogonal array testing for permeability of CCCW the proportion of glycine further increases, second permeability
ratio shows a sustained decline, but it’s higher than that one with-
Herein, four levels of orthogonal test for these 5 activated out sodium carbonate or glycine even though. It verifies that
chemicals are determined in Table 1, and the orthogonal test was sodium carbonate and glycine be conducive to the self-healing
40 G. Li et al. / Construction and Building Materials 200 (2019) 36–45

Table 4
Analysis of orthogonal test results.

Factor Sodium carbonate Sodium silicate Sodium aluminate Tetrasodium EDTA Glycine
Second permeability ratio (%) K1 185.5 183.8 227.5 188.3 185.8
K2 212.5 258.3 194.3 210.5 229.3
K3 210.8 193.8 202.0 202.0 202.5
K4 202.0 175.0 187.0 210.0 193.3
R 27.0 83.3 40.5 22.3 43.5
260
240
Second permeability ratio (%)
220
200
180

sodium carbonate

tetrasodium EDTA
0
1
3
5

1
2
3
4

0
1
2
3

0
0.5
glycine

1
2
sodium aluminate
1
2
4
6
sodium silicate

Proportion of each activated chemicals (%)

Fig. 2. Range analysis of second permeability ratio as index.

capability of concrete. However, as the activated chemicals par- the part of sodium aluminate, the second permeability ratio shows
tially substitute cement, the proportion of cement would decrease a decrease trend generally but still rises a little at 3.0% proportion
when the proportion of activated chemicals increases. The reduced of sodium aluminate. Sodium aluminate speeds up the hardening
cement is detrimental to second permeability ratio. The second process of concrete, in other words, under the definite water con-
permeability ratio decreases when the increased activated chemi- dition, the more sodium aluminate, the smaller compactness the
cals do more harm than good. The variation tendency of second concrete has. It’s easier for concrete to produce macro pores or
permeability ratio of sodium silicate is almost the same as that even obvious cracks when the concrete has a worse compactness.
of sodium carbonate and glycine. The difference is the highest sec- That decreases the anti-permeability of CCCW coating, which
ond permeability ratio of 258.3% is much higher than that of other eventually leads to the big jump of second permeability ratio. At
activated chemicals. When the proportion of sodium silicate is the 3.0% proportion of sodium aluminate, it seems that the beneficial
maximum, the second permeability ratio is lower than that of the effect of the CSH gel triggered by sodium aluminate keeps the
blank group. The second permeability ratio increases when tetra- anti-permeability of CCCW coating from dropping to some extent.
sodium EDTA added. It decreases when the proportion of tetra- As stated above, the second permeability ratio is the index for
sodium EDTA comes to 2.0%. When the proportion comes to 3.0%, choosing the optimal proportion of the activated chemical. Mean-
the second permeability ratio rises to the same value as the pro- while, the economic cost should be taken into account as well. As
portion of tetrasodium EDTA is 1.0%. When tetrasodium EDTA is the second permeability ratio is the same when the proportion of
added as an additive, the ettringite would grow as long thick noo- tetrasodium EDTA is 1.0% or 3.0%, according to the economic prin-
dles [34]. It’s easier for CCCW to heal the cracks using more acerose ciple, it’s better to choose 1.0% of tetrasodium EDTA than 3.0%.
ettrigite. The reason for the decrease of second permeability ratio Take into account the comprehensive consideration of cost and
at 2.0% proportion of tetrasodium EDTA is that when there are performance of CCCW, the optimal proportion of activated chemi-
too much tetrasodium EDTA, more calcium ions would be chelated, cals in CCCW is as follows: sodium carbonate 1.0%, sodium silicate
which imparted the formation of ettrigite or CSH gel. Jolanta Pry- 2.0%, sodium aluminate 1.0%, tetrasodium EDTA 1.0%, glycine 0.5%.
wer [35] showed a tendency that the higher concentration of tetra- This optimal proportion of activated chemicals is not listed in the
sodium EDTA there is, the higher aspect ratios for calcium orthogonal array. It is normal that the optimal array parameter
carbonate crystals have. This explained the re-increase of second obtained is not included in orthogonal experiments. In this case,
permeability ratio at 3.0% proportion of tetrasodium EDTA. For we should still use the optimal combination parameters to regard
 G. Li et al. / Construction and Building Materials 200 (2019) 36–45 41

a result, and cost should be considered at the same time. Perme- the condition of constant water-cement ratio, excessive sodium
ability test using the optimal proportion shows a second perme- aluminate would lead to the decrease of concrete compactness as
ability ratio of 333.3%, which is higher than all the second well as the formation of micropores or cracks more easily in the
permeability ratios in Table 3. concrete. As the proportion of sodium aluminate in E3 is of 3%,
It should be emphasized that the proposed method applied is to while it’s of 2% in E15, the one with a larger proportion of sodium
optimize the second permeability ratio within the given levels aluminate has more micropores or cracks, and consequently a
shown in Table 2. If a further optimization is required, the experi- lower pre-pressure self-healing rate.
mental procedure could be repeated within the new levels in a
minor area near the optimal second permeability ratio. 3.3. X-ray diffraction analysis

For better understanding of the self-healing process of CCCW,

3.2. pre-pressure experiments
the X-ray diffraction analysis was conducted on crystals scraped
off from the cracks of blank specimens and the specimens with
Specimens for this test were prepared through the procedure
CCCW after soaking for 28 days. The samples were immersed in
mentioned in Sec. 2.3. Each group of the pre-pressure experiment
anhydrous ethanol to terminate the hydration and then dried in
was consisted of three pieces of (40  40  160) mm specimens
an oven at 50 °C for 12 h. Different representative peaks were
prepared in the meantime. The pre-pressure self-healing test was
shown in Fig. 4. Herein, the activated chemicals of CCCW are
performed on each specimen. Each group of the pre-pressure
sodium carbonate 1.0%, sodium silicate 2.0%, sodium aluminate
experiments would carry out 3 times and the average pre-
1.0%, tetrasodium EDTA 1.0% and glycine 0.5%. The standard
pressure self-healing rate was obtained and shown in Fig. 3.
cement mortar with the cement content of 17.56% and the water
As can be observed in Fig. 3, even the lowest pre-pressure self-
to cement ratio of 0.6 were prepared with or without activated
healing rate of E1-E16 (103.1%) is higher than that of the blank
chemicals.
group (87.3%). The pre-pressure self-healing rate test considers
It can be seen from Fig. 4 that the main component of the crys-
CCCW as the whole part rather than considering CCCW just as a
tals scraped off from the specimens with CCCW is the same as
coating covering blocks for permeability test. Higher pre-pressure
those of the blank specimen. Peaks at 29.64, 32.32, 49.97 are
self-healing rate means the better self-healing ability. Comparing
related to CSH and peaks at 18.20, 34.27, 47.27, 50.94 are related
to second permeability ratio in Table 3, higher pre-pressure self-
to CH are shown in XRD patterns. From the XRD patterns of the
healing rate corresponds to higher second permeability ratio. How-
crystals from the specimens with CCCW, the intensities of the
ever, the pre-pressure self-healing rate and second permeability
peaks characterizing CSH increase and meanwhile the intensities
ratio are not corresponded strictly. For example, E1 has a higher
of the peaks charactering CH decrease comparing to the blank
pre-pressure self-healing rate while E3 has a higher second perme-
specimen. This attributed to the consumption of CH caused by
ability ratio, that’s contribute to the total proportion of the acti-
the addition of CCCW. CCCW reacted with CH and thus lead to
vated chemicals. In this experiment the activated chemicals
more CSH generation, which participated in pores and cracks dur-
substitute cement partially, the higher proportion of the activated
ing the curing process. This is also associated with the higher pre-
chemicals in CCCW means less cement in CCCW. The Less cement
pressure self-healing rate of the specimen with CCCW than that of
is harmful to the compressive strength of CCCW and meanwhile
the blank group.
limits the self-healing ability relatively. Meanwhile, there are also
some groups like E3 and E15 with the identical second permeabil-
3.4. Scanning electron microscopy
ity ratio in the permeability test but have different pre-pressure
self-healing rates in pre-pressure self-healing test. In terms of total
Specimens for SEM were prepared by the same procedure of the
proportion of the activated chemicals, E3 has a total proportion of
specimens for XRD analysis. Fig. 5 shows SEM images of crystals
the activated chemicals of 13% while E15 has of 14%. We conclude
from cracks of cementitious mortar blocks without and with CCCW
that the difference between the total proportions of the activated
after soaking for 28 days. Herein, the activated chemicals of CCCW
chemicals is not the major cause leads to the different pre-
are sodium carbonate 1.0%, sodium silicate 2.0%, sodium aluminate
pressure self-healing rates. The main reason is the different pro-
1.0%, tetrasodium EDTA 1.0% and glycine 0.5%. The standard
portion of sodium aluminate. As the above analysis, sodium alumi-
nate can promote the hydration of unhydrated cement, but under
Blank
180 CCCW
CSH
160
Calcium carbonate
Pre-pressure self-healing rate, %

140 CH

120
CPS

100

80

60

40

20
10 20 30 40 50 60 70 80
0
Blank E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14 E15 E16 2 °

Fig. 3. The pre-pressure self-healing rate of E1-E16 and the blank group. Fig. 4. XRD analysis of crystals from blank specimen and the specimens with CCCW.
42 G. Li et al. / Construction and Building Materials 200 (2019) 36–45

Fig. 5. SEM of crystals from cracks of specimens: a) blank specimen; b)  e) specimens with CCCW.

cement mortar with the cement content of 17.56% and the water to 3.5. Visual assessment of self-healing behavior
cement ratio of 0.6 were prepared with or without activated
chemicals. Specimens with CCCW were pre-cracked at the age of 28 days
As shown in Fig. 5, lots of CSH are generated in specimens with and then immersed into water soaking for another 28 days. It can
CCCW comparing to the blank specimen. There are also less CH and be seen clearly that the cracks were healed and covered by white
more ettringite in specimens with CCCW, which is helpful for the precipitates in Fig. 6. When cracks were healed by white crystals,
compactness and self-healing ability of concrete. water could not penetrate through origin cracks as expected.

Fig. 6. Cracks of specimens with CCCW after self-healing.

 G. Li et al. / Construction and Building Materials 200 (2019) 36–45 43

Herein, the activated chemicals of CCCW are sodium carbonate state of crystals generated on the specimens with CCCW, we dry
1.0%, sodium silicate 2.0%, sodium aluminate 1.0%, tetrasodium the wetted specimens with white crystals under room tempera-
EDTA 1.0% and glycine 0.5%. The standard cement mortar with ture. The crystals are still attached to the fracture surface as well
the cement content of 17.56% and the water to cement ratio of under the dried condition in Fig. 7c. It verifies that these numerous
0.6 were prepared with or without activated chemicals. crystals fill full of the cracks and thus the cracks are blocking to
In order to obverse the crystals directly, the pre-cracked speci- accomplish the self-healing process.
mens with CCCW through 28 days soaking into water were tested. Furthermore, Fig. 7d shows the fiber-reinforced specimens with
For comparison, the pre-cracked specimens without CCCW with CCCW after 28 days curing. There are large amount of white crys-
the same conditions were also tested. There are not any white crys- tals attached to fibers clearly on the fracture surface. 80X optical
tals presented on the surface of specimen from Fig. 7a. But from microscope was used to observe the crystals attached to fibers in
Fig. 7b, there are numerous of large visible white crystals gener- Fig. 8. The crystals with an average diameter of 0.2 mm and
ated. These crystals are attached on the fracture surface and play not any crystals with very large diameter presented. These crystals
an important role in healing process. For further investigate the stack together for blocking cracks, and thus display the self-healing

Fig. 7. Images of fracture surface of specimens: a) blank; b) wetted; c) dried; and d) with fiber-reinforced.

Fig. 8. 80X optical microscope images of the crystals attached to fibers.

44 G. Li et al. / Construction and Building Materials 200 (2019) 36–45

Fig. 9. 80X optical microscope images of cracks: a) before healing; b) after healing.

ability of CCCW. It is worth that the average diameter of 0.2 mm Sodium aluminate helps generate ettringite and sodium
agrees well with the fact that CCCW works effectively only when hydroxide in cement. Sodium hydroxide speeds up the absorption
crack width is no more than 0.4 mm generally. If cracks are too of carbon dioxide to form carbonate ion thus accelerates the for-
wide, it’s hard for these crystals to stack compacted to block cracks mation of calcium carbonate. The hydrolysis reaction of sodium
and easily to be washed away by permeated water. aluminate and the reactions that generate ettringite are as follows:
For better understanding of self-healing ability of CCCW, 80X
NaAlO2 þ 2H2 O ! AlðOHÞ3 þ NaOH ð3Þ
optical microscope was used to observe the cracks of the fiber-
reinforced specimens with CCCW before and after self-healing in
Fig. 9. 2NaAlO2 þ 3CaO þ 7H2 O ! 3CaO  Al2 O3  6H2 O þ 2NaOH ð4Þ
It can be seen from Fig. 9a that there are origin cracks of 0.2 mm
wide. Fig. 9b shows that the cracks are healed after the self-healing 3CaO  Al2 O3  6H2 O þ 3ðCaSO4  2H2 OÞ þ 20H2 O
process. White crystals are generated and the crystals stacked ! 3CaO  Al2 O3  3CaSO4  32H2 O ð5Þ
together and consequently sealed the cracks. With the generation
of numerous crystals on both sides of cracks, the self-healing Tetrasodium EDTA, as one of the calcium chelators, attaches to
behavior of CCCW is achieved. the calcium carbonate crystal and decreases its interfacial energy,
which changes the crystal shape of calcium carbonate [37]. When
Tetrasodium EDTA added, calcium carbonate grows as long thick
3.6. Self-healing mechanism analysis noodles according to AM Cody [34]. Calcium carbonate crystals
with higher aspect ratios could block the cracks more efficiently.
As we know that, activated chemicals of CCCW are soluble in Glycine is a corrosion inhibitor of steel bars. When CCCW faced
water. The capillary pore structure in the concrete structure makes with water, there are plenty of calcium ions in the system. Glycine
it easy for water to permeate, but simultaneously it also gives the and rich calcium ions chelates to generate calcium glycine, thus
condition for the active chemicals of CCCW to permeate into the plays an important role in calcium ions accumulation. When the
concrete [36]. hardened concrete is cracked, there is CH on both sides of cracks.
Sodium carbonate is one kind of pH buffer. If the cementitious However, calcium ions provided by the solution of CH are not
material cracked, acid rain is easy to permeate into the interior enough for the self-healing process. When water permeated into
of concrete and even corrode the steel bars. Sodium carbonate the interior of concrete, calcium glycine dissolves in the relatively
reacts with the acid rain to maintain the alkaline environment of restricted space and releases a part of calcium ions. This enhances
cement, which is essential for the protection of steel bars and the the concentration of calcium ions and thus accelerates the process
durability of concrete. What’s more, when CH dissolved in water of the combination of calcium ions and silicate ions in concrete.
partially, the carbonate ions penetrate into the interior of concrete When the self-healing process finished, there would be a little
with the penetrated water and react with sodium carbonate. The amount of water at the original cracks. The released glycine che-
osmotic pressure caused by the calcium ion concentration differ- lates with calcium ions dissolved from CH to form calcium glycine,
ence between inside and outside of the concrete also helps the which is prepared for the next self-healing process and consume
penetration of carbonate ions. Sodium carbonate reacts with the CH beforehand. Besides, when water is permeated deep enough
hydration product CH as follows: to reach the steel bars, calcium glycine could convert to ferrous bis-
glycinate due to the calcium glycine is more unstable. Ferrous bis-
Na2 CO3 þ CaðOHÞ2 ! CaCO3 # þ2NaOH ð2Þ glycinate attached to the steel bars slows down the corrosion
process of steel bars while the released calcium ions participate
This reaction could generate dense precipitation of calcium car-
bonate and consume CH, as shown in Fig. 3. The solubility of CH in the self-healing process continuously.
makes it a threat for the compactness of concrete. In the meantime,
byproduct sodium hydroxide could maintain the alkalinity of the 4. Conclusions
concrete and accelerate the hardening of cement.
Sodium silicate dissolves in water easily. Dissolved silicate ions This research focuses on the components and proportions of
react with the calcium ions and generate CSH gel to heal the cracks. activated chemicals of CCCW instead of its aggregate ratio or
From Figs. 1 and 4, it can be seen that the reaction of calcium ions W/C ratio to obtain the optimal self-healing capability. The second
with the sodium silicate plays an important role in the self-healing permeability ratio and pre-pressure self-healing rate as two key
process. evaluation indexes were carried out. The optimal proportion of
 G. Li et al. / Construction and Building Materials 200 (2019) 36–45 45

the activated chemicals of CCCW is as follows: sodium carbonate [10] A.M.G. Seneviratne, G. Sergi, C.L. Page, Performance characteristics of surface
coatings applied to concrete for control of reinforcement corrosion, Constr.
1.0%, sodium silicate 2.0%, sodium aluminate 1.0%, tetrasodium
Build. Mater. 14 (1) (2000) 55–59.
EDTA 1.0%, and glycine 0.5%. Sodium silicate has the most signifi- [11] H.Y. Moon, D.G. Shin, D.S. Choi, Evaluation of the durability of mortar and
cant impact on the self-healing ability of CCCW. The activated concrete applied with inorganic coating material and surface treatment
chemicals of CCCW show good self-healing ability both on second system, Constr. Build. Mater. 21 (2) (2007) 362–369.
[12] S.R. Sharp, G.G. Clemena, State-of-the-art survey of advanced materials and
permeability ratio and pre-pressure self-healing rate comparing to their potential application in highway infrastructure, Smart Mater. (2004).
the blank specimens. The consumption of CH and the formation of [13] L.W. Teng, R. Huang, J. Chen, et al., A study of crystalline mechanism of
CSH gel, calcium carbonate and ettringite were observed by means penetration sealer materials, Materials 7 (1) (2014) 399–412.
[14] V.C. Li, Y.M. Lim, Y.W. Chan, Feasibility study of a passive smart self-healing
of XRD and SEM. The formation of CSH gel, calcium carbonate and cementitious composite, Comp. Part B Eng. 29 (6) (1998) 819–827.
ettringite improve the compactness and self-healing ability of [15] M. Wu, B. Johannesson, M. Geiker, A review: self-healing in cementitious
cementitious materials. The CCCW with activated chemicals gener- materials and engineered cementitious composite as a self-healing material,
Constr. Build. Mater. 28 (1) (2012) 571–583.
ate white crystals with an average diameter of 0.2 mm and seal the [16] A.M. Neville, Properties of Concrete, Pitman Publishing Ltd, London, 1981.
cracks through the accumulation of white crystals. Based on the [17] H.F.W. Taylor, Cement Chemistry, Academic Press, London, 1990.
XRD and SEM analysis, the combined action of the dissolution of [18] M. Collepardi, A state-of-the-art review on delayed ettringite attack on
concrete, Cem. Concr. Comp. 25 (4–5) (2003) 401–407.
CH, generation of CSH gel and ettringite, maintenance of the alka- [19] C. Edvardsen, Water permeability and autogenous healing of cracks in
line environment of cement, adsorption of carbon dioxide, acceler- concrete, ACI Mater. J. 96 (4) (1999) 448–454.
ation of the generation of calcium carbonate, enrichment of [20] S. Qian, J. Zhou, M.R. de Rooij, et al., Self-healing behavior of strain hardening
cementitious composites incorporating local waste materials, Cem. Concr.
calcium ions by the complexing agent, complexing-precipitation
Comp. 31 (9) (2009) 613–621.
that accelerating the generation of calcium silicate precipitation [21] D. Homma, H. Mihashi, T. Nishiwaki, Self-healing capability of fibre reinforced
as well as acceleration of hydration of unhydrated cement particles cementitious composites, J. Adv. Concr. Tech. 7 (2) (2009) 217–228.
promoting the self-healing behavior of CCCW was proposed. [22] 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.
Conflict of interest [23] C. Shi, Y. Wu, Studies on some factors affecting CO2 curing of lightweight
concrete products, Resour. Conserv. Recy. 52 (8–9) (2008) 1087–1092.
[24] C. Shi, Y. Li, J. Zhang, et al., Performance enhancement of recycled concrete
None. aggregate – a review, J. Clean. Prod. 112 (1) (2016) 466–472.
[25] A. Neville, Autogenous healing: a concrete miracle, Concr. Inter. 24 (11) (2002)
76–81.
Acknowledgements
[26] A. Talaiekhozan, A. Keyvanfar, A. Shafaghat, et al., A review of self-healing
concrete research development, J. Environ. Treat. Tech. 2 (1) (2014) 1–11.
The authors would like to acknowledge the financial support of [27] Y. Mitsuki, M. Fujimoto, Y. Nakamura, An enhancement in the nature of
concrete with a multiplicative cement crystal type of concrete material Sendai,
Science and Technology Planning Project of Guangdong Province,
Japan, Proceedings of the 46th Annual Meeting of the Civil Engineering Society,
China (2015A010105008) and the National Natural Science 1992.
Foundation of China (51576070), China. [28] L.W. Teng, W.T. Lin, J. Chen, et al., The component analysis of penetration
sealer materials, Adv. Mater. Res. 11 (1) (2014) 74–77.
[29] X. Pei, M. Noël, M. Green, et al., Cementitious coatings for improved corrosion
References resistance of steel reinforcement, Surf. Coat. Tech. 315 (4) (2017) 188–195.
[30] P.S. Song, S. Hwang, Mechanical properties of high-strength steel fiber-
[1] F.H. Wittmann, Crack formation and fracture energy of normal and high reinforced concrete, Constr. Build. Mater. 18 (9) (2004) 669–673.
strength concrete, Sadhana 27 (4) (2002) 413–423. [31] R. Liu, Y. Zhang, C. Wen, et al., Study on the design and analysis methods of
[2] H.W. Reinhardt, M. Jooss, Permeability and self-healing of cracked concrete as orthogonal experiment, Exper. Tech. Manage. 9 (1) (2010) 52–55.
a function of temperature and crack width, Cem. Concr. Res. 33 (7) (2003) 981– [32] H. Liu, Q. Zhang, C. Gu, et al., Influence of microcrack self-healing behavior on
985. the permeability of engineered cementitious composites, Cem. Concr. Comp.
[3] W. Khaliq, M.B. Ehsan, Crack healing in concrete using various bio influenced 82 (1) (2017) 14–22.
self-healing techniques, Constr. Build. Mater. 102 (1) (2016) 349–357. [33] M. Zhang, A Study on Microcapsule Based Self-Healing Method and
[4] H.X.D. Lee, H.S. Wong, N.R. Buenfeld, Self-sealing of cracks in concrete using Mechanism for Cementitious Composites (Ph. D. thesis), Central South
superabsorbent polymers, Cem. Concr. Res. 79 (1) (2016) 194–208. University, Changsha, 2013.
[5] M.M. Al-Zahrani, M. Maslehuddin, S.U. Al-Dulaijan, et al., Mechanical [34] A.M. Cody, H. Lee, R.D. Cody, et al., The effects of chemical environment on the
properties and durability characteristics of polymer- and cement-based nucleation, growth, and stability of ettringite [Ca3Al(OH)6]2(SO4)326H2O, Cem.
repair materials, Cem. Concr. Comp. 25 (4–5) (2003) 527–537. Concr. Res. 34 (5) (2004) 869–881.
[6] M. Rooij, K.V. Tittelboom, N. Belie, et al., Self-Healing Phenomena in Cement- [35] J. Prywer, M. Olszynski, Influence of disodium EDTA on the nucleation and
Based Materials: State-of-the-Art Report of RILEM Technical Committee 221- growth of struvite and carbonate apatite, J. Cryst. Growth 375 (1) (2013) 108–
SHC: Self-Healing Phenomena in Cement-Based Materials, Springer, Springer 114.
Dordrecht Heidelberg New York London, 2013. [36] J. Qiu, H.S. Tan, E.H. Yang, Coupled effects of crack width, slag content, and
[7] L. Shi, J. Liu, J. Liu, Effect of polymer coating on the properties of surface layer conditioning alkalinity on autogenous healing of engineered cementitious
concrete, Procedia Eng. 27 (2012) 291–300. composites, Cem. Concr. Comp. 73 (2) (2016) 203–212.
[8] N.Z. Muhammad, A. Shafaghat, A. Keyvanfar, et al., Tests and methods of [37] K.J. Westin, A.C. Rasmuson, Nucleation of calcium carbonate in presence of
evaluating the self-healing efficiency of concrete: a review, Constr. Build. citric acid, DTPA, EDTA and pyromellitic acid, J. Colloid Interf. Sci. 282 (2)
Mater. 112 (112c) (2016) 1123–1132. (2005) 370–379.
[9] A.A. Almusallam, F.M. Khan, S.U. Dulaijan, et al., Effectiveness of surface
coatings in improving concrete durability, Cem. Concr. Comp. 25 (4–5) (2003)
473–481.