Re-Alkalization Effect Experiment and a New Re-Alkalization Control Model of Carbonated Concrete

Te re-alkalization technology of carbonated concrete can efectively restore the alkalinity of concrete, but the efect of re-alkalization is afected by many factors, such as the voltage, electrolyte concentration, and current conduction time. In this paper, the re-alkalization efect of carbonated concrete under diferent parameter changes is studied, and the feasibility of re-alkalization technology and the selection range of parameters in its application are determined. Results show that the voltage should be controlled below 12V or above 16V and can be adjusted properly; the electrolyte concentration should be controlled at 0.5 molL − 1 ; and the current conduction time is determined according to the control model. Finally, a new re-alkalization control model is obtained by regression of test data to control the re-alkalization process.


Introduction
Te corrosion of reinforcement is the main reason for the lack of durability of concrete. Terefore, methods for preventing the corrosion of reinforcements are an urgent problem to be solved. One of the main reasons for the corrosion of reinforcement is the carbonation of concrete [1], and the corrosion of reinforcement and the carbonation of concrete structures caused by carbon dioxide is very serious [2]. A new solution to carbonation-induced rebar corrosion is the use of electrochemical methods, which include a cathodic protection method [3][4][5], an electrochemical chloride extraction treatment [6], and an electrochemical re-alkalization technique [7]. In terms of operation and feasibility, the electrochemical re-alkalization technique is a new and more convenient method for repairing and solving the problem of carbonated concrete [8,9].
Some studies of electrochemical re-alkalization techniques have focused on transport mechanisms [7,10,11] and the efects of concrete materials [12][13][14][15]. However, parameters such as the voltage, electrolyte concentration and type, current conduction time, and steel corrosion level will all infuence the re-alkalization efect. To control the realkalization process and achieve the best efect, some studies on the process parameters of the re-alkalization efect have been reported.
Matsumoto [16] used Li 3 BO 3 and Na 2 CO 3 solutions as electrolytes in the re-alkalization test. Te results showed that the Na 2 CO 3 solution as the electrolyte easily underwent alkali-aggregate reactions, which led to the expansion of concrete in the process of re-alkalization; however, the addition of the Li 3 BO 3 solution as the electrolyte could decrease the frequency of alkali-aggregate reactions. González [13] found that compared with other solutions, K 2 CO 3 could play an efective role in preventing silicon alkali reactions and maintaining long-term stability after re-alkalization. Bastidas [17] studied whether re-alkalization could prevent the corrosion of steel bars in carbonated concrete and what conditions needed to be achieved to prevent the corrosion of steel bars. Te results showed that re-alkalization could prevent the corrosion of the steel bars in concrete, while the carbonation degree was not up to the surface of the steel bars. Terefore, the author thought that re-alkalization technology could be used as a preventive means and not as a repair method. Zhang [18] studied the electrode reaction and the change in the composition of the rust layer on the surface of steel after re-alkalization by electrochemical testing and scanning electron microscopy. Te results showed that additional inhibitors should be added if the surface of the internal rebar of carbonated concrete needs to be passivated by re-alkalization. Redaeli [3] studied the infuence of the corrosion level of the steel bar on the re-alkalization efect. Te results showed that re-alkalization was benefcial to the restoration and passivation of the corroded steel bar. González [19] systematically studied the infuence of rebar corrosion on the re-alkalization efect. Te results showed that re-alkalization could only restore the passivation of recently rusted reinforcement and could not repair severely rusted reinforcement. Miranda [20] studied the factors affecting rebalancing, and the results showed that rebar with severe corrosion could not be recovered after rebalancing. Franzoni [12] carried out a re-alkalization test on old concrete and mortar. Te results showed that the compressive strength of the old concrete and mortar could be improved after repair by re-alkalization technology, and the remaining service reliability of the improved concrete structure could meet structural requirements. Yeih [21] studied the changes in the physical properties of the mechanical parameters of concrete after re-alkalization, such as the elastic modulus, compressive strength, and pore structure. Te results showed that the mechanical parameters of the specimens decreased with increasing electric quantity, and the pore structure of the concrete was concentrated after re-alkalization.
In summary, re-alkalization can make carbonated concrete return to alkaline conditions. However, in the process of re-alkalization, the electrolyte concentration, electrifcation time, and physical state of the steel bar surface will infuence the efect of re-alkalization. To obtain the best efect of re-alkalization, research on the factors afecting the re-alkalization efect should be developed from a qualitative to quantitative standpoint. In addition, re-alkalization can not only restore the alkalinity of carbonated concrete but also produce other efects, such as changes in the mechanical properties of concrete. Terefore, research on the re-alkalization efect should be expanded to a wider range to prevent the possible adverse consequences of re-alkalization.
However, the specifc infuence of parameters on the realkalization efect has not been quantitatively studied. In this paper, the infuence of voltage, electrolyte concentration and current conduction time on the evaluation indexes of the realkalization efect is studied through the re-alkalization test of carbonated concrete specimens, and a new control model for the re-alkalization of carbonated concrete is established.

Principle of Re-Alkalization
In the process of re-alkalization, the current forces the anodic reaction as given in the following equation on the surface of the reinforcement to stop completely.
As there are only active metal cations with a small algebraic value of electrode potential in the solution, hydrogen ions will preferentially obtain electrons. When oxygen is involved in the reaction, a cathodic reaction (2) will take place on the surface of the reinforcement. If the current between the anode and cathode is too large, a cathodic overprotection phenomenon will occur, and an electrolytic water reaction will occur (3). Part of the OH − generated by (2) and (3) migrates to the anode, and part of the OH − is retained around the reinforcement, which increases the alkalinity of the concrete.
Te anions in the anode solution contain carbonate and hydroxide ions. Te carbonate ions are not easy to oxidize and will not discharge at the anode, and the hydroxide ions in the solution will preferentially discharge at the anode and produce oxygen, as shown in the following equation. In addition, the electrolytic water reaction is also conducted at the anode (5). Additionally, the highly alkaline electrolyte solution penetrates into the carbonated concrete through the porous concrete matrix under the applied voltage.
As seen from the above, the re-alkalization efect of concrete is mainly based on two aspects: the electrochemical reaction of the cathode and the electro-osmosis of the alkaline solution.

Casting and Curing of Concrete Specimens.
Te cement used in the study was a 42.5R high early strength Portland cement that complies with the Chinese standard GB 175-2007 [22]. Te coarse aggregate used was continuously graded crushed gravel with a size from 5 to 20 mm. Te fne aggregate used was locally available natural river sand. Te properties and chemical components of cement, fne aggregate, and coarse aggregate are shown in Tables 1 to 3.
In this study, the water-binder ratios were set at 0.7, and the subsequent concrete mixtures and properties are presented in Table 4.
Nineteen diferent batches of specimens, as described in Table 5, were used for the three experiments in this study: specimens for the infuence of voltage experiment (V), specimens for the infuence of electrolyte concentration experiment (C), and specimens for the infuence of current conduction time experiment (T). U4, C3, and T3 in Table 5 are from the same batch. Each batch had three specimens, and there were a total of ffty-seven specimens in this experiment. Te mean values of three test specimens were used as the results of each batch.
All test specimens were prisms with dimensions of 100 mm × 100 mm × 400 mm. One rebar with a diameter of 2 Advances in Materials Science and Engineering 10 mm and a length of 360 mm was embedded in the specimens. Te details of the test specimens are shown in Figure 1. In addition, two control specimens were cubes with dimensions of 100 mm × 100 mm × 100 mm to test and compare the carbonation degree in each experiment. Te dimensions of the specimen used for the compressive strength test of each batch were 150 mm × 150 mm × 150 mm.

Carbonation Test of Specimens.
After the specimens were poured, the mould was removed 24 hours later, and the test and control specimens were placed in a curing room for 7 days of standard curing. Ten, the specimens were immediately placed in an oven at 60°C. All the specimens were removed after 48 hours, and the test specimens were sealed with silicone sealant on fve of the surfaces, with one surface set apart so that the re-alkalization process only occurred on     10 0. 5 7 Advances in Materials Science and Engineering one surface. After the sealing was completed, all the specimens were placed in a tank for carbonation. Te carbon dioxide concentration in the tank was controlled at (20 ± 3)%, the temperature was controlled at (20 ± 2)°C, and the relative humidity was controlled at (70 ± 5)%. All specimens were removed after 2 months of carbonation, and then the control specimens were split and titrated into split sections with a phenolphthalein reagent. All the control group specimens were completely carbonated, which also indicated that the test specimens were completely carbonated.

Testing Process.
After the carbonation test, the specimens were subjected to re-alkalization. Te specimen was laid fat on a foam cushion block in a container with the unsealed surface of the specimen perpendicular to the bottom of the container. Te K 2 CO 3 solution was injected into the container until the solution submerged a portion of the unsealed surface of the specimen while not coming into contact with the rebar. A wire welded to the rebar was connected to the cathode of a power supply, and the wire attached to the steel wire mesh was connected to the anode of the power supply.
To ensure the reliability of the data obtained during the test, three test specimens in each batch were tested at the same time, and the mean value of the obtained data was taken as the test data.
Te whole test process was continuous and uninterrupted, and the current was observed and recorded at appropriate intervals. After current conduction was completed, the test specimens were split vertically on the surface that was not sealed and titrated with a phenolphthalein solution. Te re-alkalization depth was measured by observing the depth of the pink area, thus fnally obtaining Figure 2 realkalization area.
Some phenomena occurred in the testing and are described below: (1) Te anode of the steel wire mesh was corroded during the test, and the corrosion was more serious at the location exposed to the solution. Te reason was that the location exposed to the solution lost electrons easily and oxidized in the presence of air. (2) Tere were many bubbles in the solution, especially near the wire mesh. According to the principle of realkalization, the electrochemical reaction in the process of re-alkalization was the production of electrolytic water, which would inevitably produce gas, especially when the steel wire mesh was used as the anode. During the experiment, the container was hot, and the temperature of the electrolyte solution clearly increased. (3) After the test, many white crystals appeared on the realkalized surface of the specimen, as shown in Figure 2(a). Tis phenomenon was explained as follows: a certain amount of K 2 CO 3 solution would be retained on the surface of concrete, and K 2 CO 3 would undergo a hydrolysis reaction during the process of re-alkalization and electrifcation, thereby leading to the formation of crystals. (4) For the specimen with a large applied voltage, cracks appeared along the direction of reinforcement on the realkalized surface, as shown in Figure 2(b). Te reason was that the larger the applied voltage, the larger the corresponding current. Te larger current would result in a greater thermal efect on the steel bar, causing it to expand and causing the concrete to crack along the direction of the steel bar. In this test, only the U7 specimens exhibited small cracks, and no obvious cracks were found on the other specimens.

Infuences of Voltage on the Re-Alkalization Efect.
Te U1-U7 batches were tested under diferent voltages, while keeping the same electrolyte concentration and current conduction time. Te relationship between the current and time is shown in Figure 3(a). trend from the beginning and then tend to be stable or slightly decreasing. Te reason is that only a small amount of conductive material is present in the concrete before realkalization; thus, the current is small at the beginning. As the re-alkalization process progresses, the electrolyte solution slowly penetrates from the surface to the interior of the concrete, resulting in an increase in concrete conductivity. Te current gradually increases until the electrolyte in the concrete gradually approaches saturation, which inhibits the resistance from continuing to decrease; thus, the current fnally stabilizes at a certain value. Te current change trend meets the expected requirements, although the current cannot develop according to the ideal state curve due to the existence of other external factors, such as the temperature and environment. Figure 3(a) also shows that the current intensity through the specimens reaches a peak value in the process of re-alkalization. Te peak current indicates that the conductive solution penetrating into the concrete of the specimens reaches saturation, and the efect of re-alkalization is best at this point. Te times corresponding to the peak current are diferent for the specimens, indicating that the re-alkalization speeds are diferent.
To determine the infuence of increasing voltage on the re-alkalization speed, the time to reach the peak current is shown in Figure 3(b). Te fgure shows that the time corresponding to the peak current decreases and the re-alkalization speed increases with an increase in voltage.
In this experiment, the re-alkalization degree was evaluated by using a phenolphthalein indicator and measuring the re-alkalization depth of the specimens. From the centre around the rebar, the larger the re-alkalization area is, the larger degree of re-alkalization. From the concrete surface to the centre of the rebar, the re-alkalization depth Advances in Materials Science and Engineering was measured at angles of 0, 45, 90, 135 and 180°, which correspond to points 1, 2, 3, 4 and 5 in Figure 4. Te mean value of the fve depths is used as the re-alkalization depth of a specimen (s), and the mean value of the results of three specimens is used as the re-alkalization depth of a batch (s). Figure 5 shows the re-alkalization depth of a batch (s) at diferent voltages in the test. Basically, the s increases when the voltage is below 12 V or above 16 V, but the re-alkalization depth decreases when the voltage is between 12 and 16 V. Te reason is that the electric energy is mainly used for the electrolytic water reaction instead of the other electrochemical reactions when the voltage is more than 12 V according to electrochemical literature. Terefore, only a small portion of the electric energy is used to promote the penetration of the alkaline electrolyte in the concrete when the voltage reaches 12 V; thus, the re-alkalization depth decreases. In addition, the electroosmosis efect is very strong due to the excessive voltage when the voltage increases to more than 16 V because the electrolyzed water reaction is very vigorous, thus promoting the re-alkalization depth to continue increasing. Terefore, the efect of realkalization will be enhanced with increasing voltage when the voltage is below 12 V or above 16 V, but the efect of realkalization will be lessened when the voltage is between 12 and 16 V. As shown in Figure 5, the voltage should be in the range of 8-12 V after considering the above efects and economic cost.
Te points in Figure 4 are connected with straight lines after obtaining the re-alkalization depth to form a closed fgure. Te area of the closed fgure is the re-alkalization area of a specimen (A). Te obtained mean re-alkalization areas of batches (A) are shown in Table 6.
Te re-alkalization area of U1 is basically zero in Table 6, which indicates that there is basically no re-alkalization efect at this voltage, and the change in the re-alkalization area of U2-U7 is similar to the change in the re-alkalization depth. Terefore, the change in re-alkalization depth can essentially refect the change in re-alkalization area, and only an analysis of the re-alkalization depth is needed when analysing the re-alkalization efect.

Infuences of the Electrolyte Concentration on the Re-Alkalization Efect.
Te C1-C7 batches were at diferent electrolyte concentrations while keeping the same voltage and current conduction time. Te relationship between the current and time is shown in Figure 6(a).
As seen from Figure 6(a), the current change trend of C1-C7 is basically the same as that of U1-U7. Te current is small at the beginning and then gradually increases to a certain limit value, at which point the current fuctuates above and below this limit.
Te time to reach the peak current for C1-C7 is shown in Figure 6(a). Te fgure shows that the time corresponding to the peak current decreases and the re-alkalization speed increases with an increase in the electrolyte concentration when the concentration is below 0.7 molL −1 . However, electrolyte concentrations above 0.7 molL −1 do not continue to increase the re-alkalization speed.
Te relationship between the re-alkalization depth and the electrolyte concentration is shown in Figure 8. Te fgure shows that the higher the electrolyte concentration is, the greater the re-alkalization depth is; thus, the better the efect of re-alkalization is. Te fgure also shows that the increase in re-alkalization depth decreases and that the re-alkalization efect cannot continue to increase when the electrolyte concentration is above 0.7 molL −1 . Tis result is consistent with the results of the previous current change analysis.
Terefore, the electrolyte concentration should be properly controlled below 0.7 molL −1 in the process of realkalization because the re-alkalization efect cannot continue to improve above 0.7 molL −1 ; thus, the use of additional electrolyte would be uneconomical. As shown in Figure 7, the electrolyte concentration should be between 0.5 and 0.7 molL −1 when considering the above efect and economic cost.
Similar to the voltage, the obtained re-alkalization areas of the C1-C7 batches are shown in Table 7.

Infuences of the Current Conduction Time on the Re-Alkalization
Efect. Te T1-T7 batches are at diferent current conduction times while keeping the same voltage and electrolyte concentration. Te relationship between the current and time is shown in Figure 8(a).     As shown in Figure 8(a), the current changes in T1, T2 and T3 basically show an upward trend without a downward trend. Tese results indicate that the alkaline electrolyte solution in the specimen has not reached saturation and that the re-alkalization efect is always increasing when the electrifcation time is less than three days. However, the current intensity of T4-T7 begins to gradually or slightly decrease upon reaching a certain re-alkalization degree after more than three days of electrifcation. Terefore, the best re-alkalization efect of the specimen can be achieved only after being electrifed for a certain period of time, and the realkalization efect cannot increase indefnitely with an extended electrifcation time.
Te time to reach the peak current for T1-T7 is shown in Figure 8(b). Te fgure shows that the time for each specimen to reach its peak current is mostly within 40-60 hours, except T1. Tese results show that the solution in the concrete specimen basically reaches saturation when the current conduction time is between 40 and 60 hours.
According to the results, the changes in the re-alkalization depth of the T1-T7 batches at diferent current conduction times are shown in Figure 9. As seen from the fgure, the re-alkalization depth basically shows an upward trend after a few days of electrifcation but does not continue to increase after four days.
Terefore, as seen from Figure 9, the re-alkalization efect cannot continue to improve after four days; thus, continuing electrifcation for more than four days is uneconomical. Terefore, the current conduction time should be properly controlled at four days after considering the above efects and economic cost.
Similar to when the voltage was constant, the obtained re-alkalization areas of the T1-T7 batches are shown in Table 8.

Relationship between Area/Depth (A/s) and the
Parameters. Taking the area/depth (A/s) at the same electrolyte concentration in Tables 6-8 as the object, the relationships between A/s and the parameters are shown in Figure 10.
As shown in Figure 10(a), A/s basically remains unchanged when the voltage changes. Figures 10(b) and 10(c) show that A/s changes with concentration and time according to (6) and (7). Tese results indicate that the value of A/s is independent of the voltage and related to the electrolyte concentration and current conduction time.

New Control Model of the Carbonated Concrete Realkalization.
It can be seen from the published literature, the re-alkalization efect is mainly based on two aspects: the electrochemical reaction at the cathode and electroosmosis of the alkaline solution. Zhang [23] studied the infuence of electrochemistry and electroosmosis on the re-alkalization of carbonated concrete and established a model of realkalization, as shown in (8). In (8), kI e T/Fc OH − is the electrochemical efect, DUζ/4πηh · A 0 · T is the electroosmosis efect, and A · l · ρ/1 + ρ is the expected re-alkalization efect. DUζ 4πηh where D is the relative permittivity of water, U is the loop voltage, ζ is the potential on the shear plane, η is the bond coefcient of water, h is the cover thickness of concrete, k

Advances in Materials Science and Engineering
is the retention coefcient of OH − , I e is the equivalent current strength, T is the current conduction time, F is the Faraday electrolytic constant, c OH − is the concentration of OH − in the electrolyte solution, A 0 is the area of the electroosmosis path, A is the re-alkalization area, l is the buried length of the reinforcement, and ρ is the porosity of the concrete.
In this experiment, the restorative efect on the cover thickness of concrete is the main target; thus, only the efect of electroosmosis should be considered, and the model can be further simplifed as follows: From (7), the value of A/s is related to the current conduction time at the same electrolyte concentration; thus, (10) can be obtained.
Te relation extracted in Section 4.4 is related to the unique mixing ratios and test method of this research, and its use in general relation (8) lead to the extraction of a relation (10) with specifed conditions and constraints. According to the model described by Wu et al. [24], the potential ζ on the shear plane changes with the electrolyte concentration, and the reduction coefcient k z should be considered in the calculation of the potential ζ; thus, equation (11) can be obtained.
Te concrete porosity (ρ) in this experiment can be obtained from the relationship model deduced by Kollas [25] according to (12), and the result is ρ � 56%.
where σ c is the compressive strength of concrete, 21.3 MPa. Te relative permittivity of water (D) and the bond coefcient of water (η) in this experiment can be obtained from Tang [26]; and the results are D � 72.77 F/m and η � 1.139 cP. In this experiment, the cover thickness of concrete is h � 20 mm, the area of the electroosmosis path is A 0 � 2669 mm 2 , and the buried length of the reinforcement is l � 340 mm. Terefore, the following equation can be obtained. 678.82k z ζ · UT � 122.05 −0.1393T 2 + 4.9121T + 31.637 s. (13) Te relationship between k z ζ and UT can be obtained by the data in Tables 6-8 and (13), as shown in Table 9 and Figure 11. Te fgure shows a regression coefcient of R 2 � 0.9218, thereby expressing high reliability. Tus, (11) can be described as (14) according to Figure 11.

Conclusions
Based on the above study, the following conclusions can be drawn [24,26].
(1) Te voltage should be between reasonable level after considering the above efect and economic cost.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.  Figure 11: Relationship between k zδ and UT. 10 Advances in Materials Science and Engineering