The recent experimental realization of compound Tripp-B(CO)(2) (denoted as 2a), where Tripp is 2,6-di(2,4,6-triisopropylpheny1)-phenyl), breaks through conventional knowledge that only transition metals can bind more than one CO to form multicarbonyl adducts. Compound 2a is stable in air but liberates CO under light. The B CO bonds of 2a are considered to be similar to donor acceptor bonds of transition metal complexes. To address the formation mechanism and chemical bonding of this novel type of boron compounds, we present a density functional theory study on the formation and photolysis of 2a and similar compounds. The results suggest that the formation of 2a is facile by three consecutive additions of CO to the terminal borylene metal complex, that is, the boron source of the synthesis. These CO additions can be practically accomplished via two different paths: CO direct addition and CO migration followed by addition. Such mechanisms can be excellently rationalized by the donor acceptor bonding model of the terminal borylene complex, which in turn suggests that using donor acceptor bonds for 2a is natural for understanding the mechanisms. Liberation of CO from 2a and its similar compounds has higher energy barriers at the ground states than that at the triplet states by 40 kcal/mol. These energy barriers explain the experimentally observed air stability and photolysis of these compounds. The results for the first time provide mechanistic insights for the unprecedented chemical processes; they allow evaluation of the applicability of donor acceptor bonding in main-group compounds from the new perspective of chemical reactions.