Predicted Gas‐Phase and Liquid‐Phase Acidities of Carborane Carboxylic and Dicarboxylic Acids

Abstract

By means of MP2 and DFT computations we predict gas‐phase acidities and liquid‐phase (MeCN) acidities of (di)carboxylic acids derived from icosahedral ortho, meta, and para‐carboranes. For comparative purpose, we include the benzoic and phthalic acids. Substitution of benzene by a carborane cage – cage effect – strikingly increases the gas‐phase acidity (lower GA) for the (di)carboxylic acids, being the ortho isomers always the most acidic, following the order ortho ≫ meta > para. The computed GA of the dicarboxylic acid derived from ortho‐carborane is far lower than sulphuric acid, due to an enhanced stabilization of the carboxylate through an intramolecular OHO bridge connection, also taking place in phthalic acid. The change of GA relative to ortho, meta and para positions of the carboxylic groups ‐ isomer effect ‐ is larger for carboranes. As regards to liquid‐phase (MeCN), the computations show that carborane (di)carboxylic acids also show a larger acidity (lower pKa) as compared to the phthalic acids and that the dicarboxylic ortho‐carborane is also a superacid in the liquid phase (MeCN), due to the OHO bridge connection in the carboxylate, as in the gas phase. Additional computations show how much of this isomeric effect is to be attributed to the electronic delocalization.

Icosahedral Carbaboranes with Peripheral Hydrogen–Chalcogenide Groups: Structures from Gas Electron Diffraction and Chemical Shielding in Solution

“…We computed the gas‐phase acidities (GAs) of monoacids of 1 and 2 using the MP2/AE1 method and found that 2 is slightly more acidic than 1 (1271.4 vs. 1288.2 kJ mol −1 for 2 and 1 , respectively). These values are comparable with that of 1279.5 kJ mol −1 computed for the monoacid of 1,2‐(COOH) 2 ‐ closo ‐1,2‐C 2 B 10 H 10 . However, the corresponding computed GA values for the diacids are not so pronounced, which can be ascribed to formation of the CCEH ⋅⋅⋅ E pentagonal rings that hinder further deprotonations and suggest that the difficulty in preparing 1,2‐(EH) 2 ‐ closo ‐1,2‐C 2 B 10 H 10 , particularly for E=Se, is because the Se ⋅⋅⋅ H contact is shorter than the S ⋅⋅⋅ H contact.…”