*3.2. Crack-Healing Quantification under Different Supplementary Cementitious Materials* 3.2.1. The Crack-Healing Ratio

The crack-healing ratios of mortar specimens containing different SCMs with various content are shown in Figure 7. As can be seen, the addition of SCMs changed the autogenous self-healing behavior of surface cracks in mortar specimens. The influence of SCMs on the crack-healing ratio is related to their type and content. For BFS specimens, the crack-healing ratio decreased with the increase in BFS content. Moreover, 20BFS exhibited a higher crack-healing ratio compared to the reference; however, 40BFS showed a lower crack-healing ratio compared to the reference. For FA specimens, the crack-healing ratio also decreased with the increase in FA content. However, both 20FA and 40FA showed a lower crack-healing ratio compared to the reference after a healing period of 28 days. In

general, it is not beneficial to replace cement with FA or BFS for the self-healing of early age surface cracks (pre-cracking age of 3 days) in mortar specimens (except 20BFS specimen). Comparing FA with BFS, at a given content of 20%, BFS specimens exhibited a higher crack-healing ratio compared to FA specimens. However, FA specimens showed a higher crack-healing ratio compared to BFS specimens when the content was increased to 40%.

**Figure 7.** Crack-healing ratio of specimens containing different SCMs with various content (*d* denotes days).

The autogenous self-healing of surface cracks is mainly attributable to two mechanisms: (1) ongoing hydration of unreacted cement or SCMs particles; (2) formation of calcium carbonate precipitation. The crack-healing ratio of specimens with SCMs depends on the coupled effect of the abovementioned two aspects. In this study, early age cracks (pre-cracking age of 3 days) were generated. After cracking, specimens with SCMs had more unreacted binder particles due to slower hydration of FA or BFS. The ongoing hydration of unreacted binder particles is beneficial to promote self-healing of the crack. However, the ongoing hydration of FA or BFS needs to be activated by Ca(OH)2, which is mainly attributed to cement hydration. In specimens with SCMs, the portlandite content is lower than that of the referenced specimens due to less cement composition and the partial consumption of portlandite by the hydration reaction of FA or BFS [18,27]. The reduction in Ca(OH)2 greatly influenced the formation of calcium carbonate precipitations and ongoing hydration of FA or BFS at the surface cracks. When the formation of calcium carbonate precipitation is the main mechanism, FA or BFS may not be good for the autogenous self-healing of surface cracks. It was noted that 20BFS exhibited a higher crack-healing ratio compared to the reference. This is because that there are still large amounts of unhydrated cement particles in early cracking specimens when the BFS content is low; further hydration of unreacted cement particles provides additional Ca(OH)2 to promote the hydration of BFS and the formation of calcium carbonate precipitation at the crack's surface. The 40FA specimen exhibited a higher crack-healing ratio compared to the 40BFS specimens, which may be attributed to the lower hydration activity of FA, resulting in more Ca(OH)2 for calcium carbonate formation.

### 3.2.2. Water Permeability Test

Figure 8 shows the change in relative permeability of specimens containing different SCMs with various content over different healing times. The changing trend of regaining water tightness is different from that of the crack-healing ratio. This is because regaining the water tightness of the pre-cracked mortar specimens is not only related to the sealing of surface crack but also to the healing of internal cracks. For BFS mortars, both 20BFS and 40BFS showed better regaining abilities of water tightness compared to the reference specimens. Although a lower crack-healing ratio after healing for 28 days was observed for 40BFS mortar compared to the reference, the higher regaining capacity for water tightness may be attributed to more healing products being formed in internal cracks due to the ongoing hydration of more unreacted BFS particles. As for FA mortars, the regaining of water tightness declined with the increasing FA content. This is because that the hydration activity of FA is low and not enough calcium hydroxide can be utilized to activate the hydration of unreacted FA particles in internal cracks for mortars with a high content of FA. Moreover, it was noted that 20FA and 40FA exhibited a lower decline in the relative permeability coefficient after healing for 3 days compared to the reference, but almost the same decline in the relative permeability coefficient after healing for 28 days was observed compared to the reference. This is because unreacted FA particles in the crack cannot be activated in the early healing process but can be activated as the healing time increases. It also can be found that BFS mortars exhibited a better regaining of water tightness compared to FA mortars; this is due to their higher hydration activity and CaO content for slag.

**Figure 8.** Relative permeability coefficient of specimens containing different SCMs with various content (*d* denotes days).

#### *3.3. Mineralogy of Reaction Products of Self-Healing in Early Age Cracks*

After crack-healing quantification, XRD, SEM and TG/DTG were conducted on the healing products scraped from the healed surface cracks of the referenced specimens, as well as the specimens that contained different supplementary cementitious materials (SCMs). Figure 9 shows the XRD patterns of the healing products formed in mortar specimens containing different supplementary cementitious materials (SCMs) compared to the reference. No significant differences were found for the diffraction peaks of all samples. Calcite was detected as the major crystal healing product for all samples. SEM observations indicated that micron-sized calcite crystals with a typical rhombohedral morphology were

closely packed together, as shown in Figure 10. In addition, the TG/DTG curves of the healing products in the crack mouth are shown in Figure 11. An obvious weight loss in the range of 600–800 ◦C was found, and the corresponding peaks related to the decomposition of calcium carbonate at about 750 ◦C were observed in all samples. The weight losses between 600 and 800 ◦C of reference, 20FA and 20BFS were 41.12%, 40.93% and 40.45%, respectively. According to the thermal decomposition equation of calcium carbonate, the percentages of calcium carbonate in the healing products from reference, 20FA and 20BFS were calculated as 93.46%, 93.03%, and 91.94%. The results of TG analysis were consistent with XRD and SEM, indicating that the major healing product from the healed surface cracks of the referenced specimens and the specimens containing different supplementary cementitious materials (SCMs) was calcium carbonate.

**Figure 9.** X-ray diffraction (XRD) patterns of the healing products in the mouths of the cracks.

**Figure 10.** Scanning electron microscopic (SEM) observations of the healing products in the mouths of the cracks: (**a**) Reference; (**b**) 20FA; (**c**) 20BFS.

**Figure 11.** Thermogravimetric (TG/DTG) curves of the healing products in the mouths of the cracks: (**a**) TG curves; (**b**) DTG curves.

#### **4. Conclusions**

The effects of curing conditions and supplementary cementitious materials on the autogenous self-healing of early age cracks in cement mortar were investigated. Three curing conditions (standard curing, wet–dry cycles and incubated in water) and two SCMs (FA and BFS) with various contents (cement replacement ratio at 0%, 20%, and 40%) were considered. Autogenous crack self-healing efficiency of mortar specimens was evaluated by performing a visual observation and a water permeability test. Moreover, XRD, SEM and TG/DTG were conducted to characterize the healing products. Based on the experimental results, the following conclusions can be drawn:

(1) The presence of water is essential for autogenous self-healing of early age cracks in cement mortar. Crack self-healing efficiency was highest in specimens incubated in water. However, no significant self-healing occurred in specimens exposed to standard curing. For wet–dry cycles, a longer healing time was needed to obtain good self-healing compared to those samples incubated in water.

(2) SCMs type and content significantly affected the autogenous self-healing ability of early age cracks. Crack self-healing efficiency for early age cracks decreased with the increase in FA and BFS content. BFS mortars exhibited better recovery of water penetration resistance compared to the reference and FA mortars. Almost the same regaining of water tightness and a lower crack-healing ratio after healing for 28 days in FA mortars were observed in comparison to the reference.

(3) The major healing product in surface cracks of specimens with and without SCMs was micron-sized calcite crystals with a typical rhombohedral morphology.

**Author Contributions:** Conceptualization, M.L. and H.H.; methodology, M.L., K.J. and J.B.; investigation, M.L., K.J. and J.B.; data curation, K.J. and Z.D.; writing—original draft preparation, M.L.; writing—review and editing, M.L. and H.H.; visualization, J.B. and Z.D.; supervision, M.L., D.Y., Y.G. and H.H. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the National Natural Science Foundation of China (No. 51808483), the Natural Science Foundation of Jiangsu Province (No. BK20180930), the Opening Project of State Key Laboratory of Green Building Materials, and the Opening Fund of Jiangsu Key Laboratory of Construction Materials (No. CM2018-10).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding authors.

**Conflicts of Interest:** The authors declare no conflict of interest.
