“…Unexpectedly, the pK a of Ti 5c OH − is 5.61, smaller than that of Ti 5c OH 2 , which contradicts the general understanding that the second pK a is larger than first pK a for the same surface species. 30 By converting the pK a values into the free energy of reaction 12, we obtain the ΔG PT = −0.16 eV, which confirms our speculation and thereby rationalizes the stability of Ti 5c O 2− . Moreover, we verify this conclusion using PBE-D3, SCAN, and HSE06 functional with simplified model, as shown in Table 3 (see the Supporting Information for computational details).…”