Test tube experiments undertaken in this study utilized materials, including deionized water, 99% purity of urea, 96% purity of calcium chloride (CaCl₂), and 98% purity of magnesium chloride (MgCl₂) and ammonium chloride (NH₄Cl). Furthermore, the activity of urease enzyme used to promote urea hydrolysis is 1.1 U/mg. CaCl₂ was first poured into the prepared urea solution, followed by urea hydrolysis using the urease enzyme, finally producing the carbonate precipitation. This is referred to also as the ordinary EICP process, as shown in Figure 1. Furthermore, a modified EICP process was run in parallel, considering the effect of MgCl₂ and NH₄Cl and the concentration of CaCl₂, as shown in Figure 1. The mass of CaCO₃ precipitated in the ordinary and modified EICP processes was measured, and the precipitation ratio (PR) of CaCO₃ was compared against each other to quantify the precipitation efficiency. PR can be calculated using: PR = Actual   precipitation   mass Theoretical   precipitation   mass × 100 (1) where the theoretical precipitation mass is equal to C × V × M, in which C and V correspond to the concentration of urea–calcium carbonate solution in moles per liter and the total volume of EICP solution in liters respectively. M is the molar mass of calcium carbonate (100.087 g/mol) (Neupane et al., 2013; Putra et al., 2017b).

The calcite precipitation, resulting from the test tube experiments, was then applied to prevent ancient clay bricks from weathering erosion. Laboratory tests, including water absorption and air permeability tests, were conducted to verify the water-resisting properties of the EICP-treated clay bricks. Image binarization primarily aimed to evaluate the extent of the water-resistant film formed across the treated brick surface. The water absorption and air permeability of the EICP-treated bricks were defined using Eqs 2 and 3, respectively. Q S = K t (2) T = M 1 − M 2 M 1 × 100 (3) where Q is equal to the water weight measured against different time intervals of immersion, S is the cross-section area of brick immersed in water, K is the slope in the Q/S-t^0.5 diagram, also referred to as the sorptivity coefficient (g/cm²/s^0.5), t is the duration for the water absorption tests, M ₁ is the mass of brick after the water saturation, M ₂ corresponds to the mass of brick after retrieval from the constant temperature chamber against different time intervals, and T corresponds to the percentage of weight loss (%).

The test tube experiments primarily compare the carbonate precipitation of the ordinary EICP process with that of the modified EICP process when subjected to the effect of MgCl₂, NH₄Cl, and CaCl₂ additions, respectively, and reveal the enhancement and degradation mechanisms of carbonate precipitation. In light of this, the test tube experiments consisted of one control group (A1 to A6 corresponding to the ordinary EICP process) and three test groups. One test group (B1 to B9 corresponding to the modified EICP process) was to investigate the effect of Mg addition on the performance of carbonate precipitation. The second test group (C1 to C6 corresponding to the modified EICP process) was principally to investigate the effect of NH₄ ⁺ addition on the performance of carbonate precipitation. The third test group (D1 to D5 corresponding to the modified EICP process) mainly took into account the effect of CaCl₂ addition. Measurements of pH and electric conductivity (EC) were accompanied with the test tube experiments, highlighting the relation of pH and EC with the performance of carbonate precipitation. The design of the test tube experiments is summarized in Table 1.

The flowchart of the test tube experiments is depicted in Figure 1. The procedure of test tube experiments is as follows: 1) The weight of the empty tube is first measured, and 25 ml of distilled water is added. 2) Urea and CaCl₂ are added to the distilled water, 3) and a given proportion of MgCl₂ or NH₄Cl is added to the urea–CaCl₂ solution. 4) Urease enzyme is the last to be added to the solution and stirred thoroughly to ensure the uniformity of the mixed solution. 5) The pH and EC of the mixed solution are measured during the urea hydrolysis process at a temperature of 30°C. 6) After that, the calcite precipitation is retained using a filter paper. 7) The calcite precipitation is also acquisitioned in the test tube, and 8) they are dried at 60°C for 24 h and weighed for its actual mass and precipitation ratio evaluation. The most popular methods for the measurement of urease activity include the Nessler’s reagent colorimetric method. The accuracy of the measured urease activity can be enhanced by a modified method proposed by van Paassen (2011) where electric conductivity (EC) measurements bring benefits to the enhancement of the accuracy. This is to say that EC can also be considered as a key indicator that reflects the urease activity. For this reason, the present work used the measurements of EC and pH to evaluate the urease activity and subsequently the degree of urea hydrolysis instead of direct measurement.

Ancient brick specimens were dried at 110°C for 24 h and then placed in an incubator for cooling (30°C). Subsequently, their top surfaces were treated with an EICP solution of 25 ml for four times. The other surfaces, including bottom and peripheral surfaces, were thoroughly sealed with melting paraffin, thereby allowing a water seepage through the treated surface only. The brick specimens were then immersed in the deionized water, and their weights were measured at a given time interval. The sorptivity coefficient K was adopted as an indicator for assessing the water-resisting properties after the treatment (see Eq. 2). The higher the K, the poorer the water-resisting properties.

Similarly, the top surface of the ancient brick specimens was treated with the EICP solution for four times. Prior to immersion in deionized water, they were weighed. They were then immersed in the deionized water to reach their saturation. The other surfaces of the brick specimens were sealed with melting paraffin right after their saturation. The T value was adopted to assess the weight loss against different time intervals when they were heated at 80°C (see Eq. 3). The lower air permeability can be attained using a smaller T value.

The temporal relations of EC and pH in the ordinary EICP process when subjected to various concentrations of urea–CaCl₂ are shown in Figure 2. EC shows a small change in the very beginning of the ordinary EICP process. This is not attributed to the ionization of urea but to the dissolution of urea. EC goes up very quickly after pouring CaCl₂ into the urea solution because the ionization of CaCl₂ produces a great number of ions that are capable of carrying electric charges. There are two chemical reactions initiated after pouring urease enzyme; these are urea hydrolysis, initiated by urease enzyme and produces NH₄ ⁺ and CO₃ ²⁻ ions (Eqs 4, 5), and carbonate precipitation, resulting from a chemical reaction between Ca²⁺ ions and CO₃ ²⁻ ions, and consumes CO₃ ²⁻ ions (Eq. 6). C H 4 N 2 O + 4 H 2 O → H 2 C O 3 + 2 N H 4 + + 2 O H − (4) HC O 3 − + O H − → C O 3 2 − + H 2 O (5) C a 2 + + C O 3 2 − → CaC O 3 (6)

The effect of the urea hydrolysis on EC outweighs the effect of carbonate precipitation, which is deemed as the main cause leading to a steadily increasing tendency of EC. EC consistently increases until the end of the ordinary EICP process. On the other hand, pH remains nearly unchanged after the introduction of CaCl₂, but it shows a significant change when pouring urease enzyme into the urea–CaCl₂ solution. This is due to the production of OH⁻ ions, induced by urea hydrolysis using urease enzyme. Then the pH goes into a decline until the end of the ordinary EICP process. The consumption of OH⁻ ions in the course of carbonate precipitation causes pH to decrease gently after reaching a peak value of 8.9.

The relations of the actual mass of carbonate precipitation and the precipitation ratio (PR) versus the urea–CaCl₂ concentration are shown in Figure 3. It is clear that the carbonate precipitation versus urea–CaCl₂ concentration relation behaves in a “bell” shape, meaning that the highest carbonate precipitation cannot be achieved when subjected to higher or lower urea-CaCl₂ concentrations. The lower urea-CaCl₂ concentrations cannot provide enough basal constituents that are in great necessity of the EICP process toward causing a degradation of the carbonate precipitation, whereas for the higher urea–CaCl₂ concentrations, the lower urease enzyme concentrations, in turn, impede the urea hydrolysis, leading to a degradation of the carbonate precipitation as well. In contrast, the PR versus urea–CaCl₂ concentration relation presents a descending tendency all the way to the end of the EICP process, indicating that the lower the urea–CaCl₂ concentration, the higher the PR. These results also indicate that it is either the highest carbonate precipitation or the highest PR, and we cannot have it both ways. Alternatively, a point where two curves intersect each other may provide a pathway to achieve both higher carbonate precipitation and PR. This point, for the ordinary EICP process, presents at the urea–CaCl₂ concentration being equal to 0.3 mol/L.

The temporal relations of EC and pH when subjected to the effect of MgCl₂ addition are shown in Figure 4. Given the higher carbonate precipitation and PR at 0.3 mol/L for the ordinary EICP process, the concentration of CaCl₂ decreases from 0.3 to 0.1 mol/L, while the concentration of MgCl₂ increases from 0 to 0.2 mol/L. This is to say that such an experimental design principally does not allow a total concentration in excess of 0.3 mol/L. When MgCl₂ starts involving in the EICP process, the higher the concentration of MgCl₂, the more significant the change in EC. However, compared with lower concentrations of MgCl₂, higher concentrations of MgCl₂ appear to cause some difficulty in elevating EC in the later stage of the modified EICP process. Similarly, pH shows a significant change when introducing urease enzyme to the urea-CaCl₂ solution. The highest value being greater than 9 is higher than that of the ordinary EICP process. Subsequently, pH goes into a decline until the end of the modified EICP process. The relations of the carbonate precipitation and PR versus the concentration of MgCl₂ are shown in Figure 5.

The carbonate precipitation for lower concentrations of MgCl₂ (0–0.02 mol/L) increases with the increase in MgCl₂ concentration, while for higher concentrations of MgCl₂ (0.02–0.20 mol/L), it decreases with the increase in MgCl₂ concentration. The highest calcite precipitation of 0.67 g for the modified EICP process under the effect of MgCl₂ addition presents when the concentration of MgCl₂ reaches 0.02 mol/L, which is higher than the highest calcite precipitation of 0.6 g for the ordinary EICP process (see Figures 3, 5). It is well known that the EICP process starts with urea hydrolysis toward discharging CO₃ ²⁻. The carbonate precipitation is attained when these CO₃ ²⁻ catch cations. The cations, including Ca²⁺ ions and Mg²⁺ ions, in solution are, however, bonded by water dipoles, and they have to be dehydrated prior to attracting CO₃ ²⁻. Compared with Ca²⁺ ions, Mg²⁺ ions are more difficult to dehydrate because of its higher hydration enthalpy, thereby precipitating CaCO₃ (calcite) crystals instead of MgCO₃ (aragonite) crystals. Despite the difficulty in Mg²⁺ ion dehydration, the hydrated Mg²⁺ ions can poison the crystal surface. Mg²⁺ ions can only be bonded by two oxygen atoms resulting from two CO₃ ^2- groups, which connection is relatively weaker than Ca²⁺ ions that are bonded by at least three oxygen atoms. For this reason, Mg²⁺ ions will be expelled by the following Ca²⁺ ions and cannot poison the growth of aragonite. Given a relatively higher solubility of magnesium calcite compared with pure calcite, the growth of calcite is inhibited. These results lead us to conclude that Mg²⁺ ions inhibit the growth of calcite and promotes aragonite precipitation.

The highest PR appears when the concentration of MgCl₂ reaches 0.15 mol/L. To tackle the conflict, a point where two curves meet each other at a concentration of MgCl₂ being about 0.05 mol/L prevents the degradation of carbonate precipitation and also secures the performance of PR. Furthermore, the lower carbonate precipitation, accompanied with higher concentrations of MgCl₂, is observed. As discussed, although Mg²⁺ ions inhibit the growth of calcite, they promote another carbonate precipitation, namely, aragonite. Since aragonite is featured with smaller molecular weight compared with calcite, this causes the decreasing carbonate precipitation with the increasing concentration of MgCl₂. On the other hand, it can be observed from Figures 3, 5 that Mg²⁺ ions cause the pH to reach its highest value 25 min after the beginning of the process; however, it reaches its highest value 15 min after the commencement of the process when Mg²⁺ ions do not take part in the process. This is to say that Mg²⁺ ions added to the urea-CaCl₂ solution are difficult to dehydrate and delay the rate of urea hydrolysis, discharging OH⁻ ions at a slower rate. Notwithstanding that, Mg²⁺ ions promote the growth of aragonite in addition to calcite toward gearing up the performance of PR to higher than 114%.

The temporal relations of EC and pH when subjected to the effect of NH₄Cl addition are shown in Figure 6. EC shows a rapid increase when NH₄Cl takes part in the process. The higher the concentration of NH₄Cl, the more significant the increase in EC. Higher concentrations of NH₄Cl also result in higher levels of EC. Notwithstanding that, given that NH₄Cl discharges NH₄ ⁺ ions in the EICP process, the higher concentrations of NH₄Cl, in turn, depress the urea hydrolysis (denoted by the blue area). Eq. 7 commences to move toward the opposite direction. EC, therefore, presents a small increase, followed by a negligible increase toward the end of the process. C H 4 N 2 O + 4 H 2 O ← H 2 C O 3 + 2 N H 4 + + 2 O H − (7)

As to the temporal relation of pH, NH₄Cl added to the urea–CaCl₂ solution do not cause pH to change notably (denoted by the gray area). The reverse EICP process causes a reduction in the discharge of OH⁻. In light of this, the highest value reduces to approximately 8.5 which is lower than 8.9 of the ordinary EICP process. The higher NH₄Cl concentrations are deemed as the main cause leading to the decrease in pH. As a result, the higher the concentration of NH₄Cl, the smaller the number of OH⁻ discharged, and the smaller the number of CO₃ ²⁻, thereby degrading the carbonate precipitation. This also causes PR to monotonically decrease with the increase in the NH₄Cl concentration, as shown in Figure 7.

The basal constituents applied to the process generally consisted of urea and CaCl₂. To deepen our understanding of the effect of CaCl₂ addition on the carbonate precipitation, test tube experiments, associated with higher concentrations of basal constituents, were conducted. The temporal relations of EC and pH under the effect of CaCl₂ addition are depicted in Figures 8, 9. The higher the concentration of CaCl₂, the more significant the change in EC. It can also be seen that the temporal relations of EC throughout the urea hydrolysis are relatively stable, indicating that the effect of carbonate precipitation counterbalances the effect of urea hydrolysis. In contrast, the concentration of CaCl₂ appears to have a negligible effect on pH. Notwithstanding that, pH during the urea hydrolysis can also be characterized as a rapid increase and then a gentle decrease toward the end of the process. It can be regarded that the urea hydrolysis accompanies the discharge of OH⁻ toward the elevating pH very quickly. Then the carbonate precipitation commences to consume OH⁻, causing pH to go into a decline after reaching a peak. It is worth noting that the highest pH of 8.6, induced by the modified EICP process under the effect of CaCl₂ addition is attained 30 min after the commencement of the process. Given that in the discharge of OH⁻ during the urea hydrolysis Ca(OH)₂ will form indicates an occupation of OH⁻. The occupation of OH⁻ causes a lack of CO₃ ²⁻, leading to a reduction in the carbonate precipitation. Furthermore, the occupation of OH⁻ also reduces pH, and for this reason, Ca(OH)₂ that is just formed commences discharging OH⁻ to counterbalance the reduction of pH. This is considered as the main cause leading to a delay in the attendance of the highest pH (compared with the ordinary EICP process). Compared with PR of the modified EICP process under the effect of NH₄Cl addition, the effect of CaCl₂ addition degrades PR a step further by converting all Ca²⁺ ions into carbonate precipitation (see Eq. 1).

Figure 10 presents the image binarization results of the brick specimens without EICP process and with once, twice, three times, and four times of EICP process where the white color area indicates carbonate precipitation. The percentage extent of carbonate precipitation across the brick specimen surface elevates from 38.9% for once of the EICP process to 91.3% for four times of EICP process. The brick specimens with four times of EICP process were, therefore, applied to the subsequent water absorption and air permeability tests. The temporal relations of Q/S for the brick specimens subjected to the ordinary and modified EICP processes, respectively, are shown in Figure 11A. The Q/S increases sharply at the very beginning, and then the Q/S shows a little change 40 min after the commencement of the water absorption tests. The modified EICP-treated specimen subjected to the effect of higher MgCl₂ concentrations is featured with the lowest water absorption. In contrast, the specimen without the EICP process is featured with the highest water absorption. Compared with the study of Liu et al. (2020b), their water adsorption is measured as 0.15 g/cm² compared with 0.04 g/cm² from the present work, although it may involve bacterial culture and inoculation. On the other hand, the temporal relations of T for the specimens subjected to the ordinary and modified EICP processes, respectively, are shown in Figure 11B. The modified EICP-treated specimen subjected to the effect of higher MgCl₂ concentrations has the lowest mass loss corresponding to the lowest air permeability. The specimen without the EICP process has the highest mass loss, which also indicates the highest air permeability. These results not only represent the excellent water-resisting ability of the specimens but highlight the potential use of the EICP process for the protection of heritage buildings in NW China.

On the whole, the modified EICP process under the effect of MgCl₂ addition performs the best, with the highest carbonate precipitation, followed by the ordinary EICP process (Figure 12A). In contrast, the modified EICP process when subjected to the effect of NH₄Cl or CaCl₂ addition performs the worst, with the lowest carbonate precipitation. Mg²⁺ ions added to the urea–CaCl₂ solution delays the rate of urea hydrolysis due to the difficulty in their dehydration. Although Mg²⁺ ions inhibit the growth of calcite, they, however, favor aragonite precipitation toward producing more carbonate precipitation. These results reveal the enhancement mechanism of carbonate precipitation (see Figure 12B). The effect of NH₄Cl addition not only reduces the discharge of OH⁻ by degrading the urease activity but also causes the EICP process to move toward the opposite direction. The lack of OH⁻ and reverse EICP process reduce the carbonate precipitation and reveal the degradation mechanism of carbonate precipitation (see Figure 12C). Last but not the least, the effect of CaCl₂ addition leads to the occupation of OH⁻ and subsequently the lack of CO₃ ²⁻, associated with the occupation of OH⁻, which causes a reduction in the carbonate precipitation, revealing the hijacking mechanism of carbonate precipitation (Figure 12D).