A comparative study of the metal−glycine bonding for the biologically relevant Cu+ and Cu2+ pair is presented. The structure and vibrational frequencies for several coordination modes of Cu+ and Cu2+ to glycine have been determined using the hybrid three-parameter B3LYP density functional approach. Single-point calculations have also been carried out at the modified coupled pair functional (MCPF) and single- and double- (triple) excitation coupled cluster (CCSD(T)) levels of theory and using larger basis sets. Calculations have shown that the metal−glycine bonding and the energy ordering of the different conformers are very different in Cu+-glycine than in Cu2+-glycine. Whereas for Cu+-glycine, the ground state structure is found to have a bidentated η2-N,O coordination in which Cu+ interacts with the nitrogen of the amino group and the carbonyl oxygen, the ground state structure of Cu2+-glycine is the η 2 -O,O (CO2 -) one, derived from the interaction of the metal cation with the CO2 - terminus of the zwitterionic glycine. In this case, the results seem to indicate that glycine acquires an important radical character that changes the relative metal affinities of the different basic sites, which favors the interaction of the metal cation with the CO2 group compared with other coordinations.

Correlated calculations show the proton-transferred OH−H3O+ isomer to be the ground-state structure of the (H2O)2 + dimer ion, with the C 2 h hemibond structure being ca. 8 kcal/mol less stable. Modern density functionals however favor the hemibond structure, overestimating the strength of the three-electron bond by ca. 17 kcal/mol. The wrong prediction of the relative stability of the two isomers is attributed to overestimation by the exchange functionals of the self-interaction part of the exchange energy in the hemibond ion due to its delocalized electron hole. It is cautioned that this erroneous behavior of the density functionals for exchange, if unrecognized, may lead to wrong predictions for ground-state structures of systems with a three-electron bond.

Single and double proton-transfer reactions in Watson−Crick Guanine−Cytosine (GC) and Adenine−Thymine (AT) radical cations have been studied using the hybrid density functional B3LYP method. Calibration calculations for the formamidine−formamide dimer, a model system of AT, have shown that B3LYP compares well to the high level ab initio correlated method CCSD(T), both for the neutral and cationic systems. The single proton-transfer reaction is favorable in both the GC and AT radical cations; it takes place from the ionized monomer (guanine and adenine, respectively), which increases its acidity, to the neutral fragment. For the two systems, GC and AT, the nonproton transferred and single proton transferred structures are almost degenerate (ΔE = 1.2 kcal/mol), and the process presents low energy barriers (4.3 kcal/mol for GC and 1.6 kcal/mol for AT). The double proton-transfer reaction is less favorable than the single one, in contrast to what is observed for the neutral systems. The relative stability of the different structures can be understood considering two factors:  the relative stability of the asymptotes from which they derive and the number and sequence of the strong and weak hydrogen bonds formed. For the same number of strong short hydrogen bonds, the most stable structures are those in which the strong H-bonds are neighbors. Based on these considerations, a prediction for other pairings is reported.

A comparative theoretical study of a bimolecular reaction in aqueous solution and catalyzed by the enzyme catechol O-methyltransferase (COMT) has been carried out by a combination of two hybrid QM/MM techniques: statistical simulation methods and internal energy minimizations. In contrast to previous studies by other workers, we have located and characterized transition structures for the reaction in the enzyme active site, in water and in a vacuum, and our potential of mean force calculations are based upon reaction coordinates obtained from features of the potential energy surfaces in the condensed media, not from the gas phase. The AM1/CHARMM calculated free energy of activation for the reaction of S-adenosyl methionine (SAM) with catecholate catalyzed by COMT is 15 kcal mol(-1) lower the AM1/TIP3P free-energy barrier for the reaction of the trimethylsulfonium cation with the catecholate anion in water at 300 K, in agreement with previous estimates. The thermodynamically preferred form of the reactants in the uncatalyzed model reaction in water is a solvent-separated ion pair (SSIP). Conversion of the SSIP into a contact ion pair, with a structure resembling that of the Michaelis complex (MC) for the reaction in the COMT active site, is unfavorable by 7 kcal mol(-1), largely due to reorganization of the solvent. We have considered alternative ways to estimate the so-called "cratic" free energy for bringing the reactant species together in the correct orientation for reaction but conclude that direct evaluation of the free energy of association by means of molecular dynamics simulation with a simple standard-state correction is probably the best approach. The latter correction allows for the fact that the size of the unit cell employed with the periodic boundary simulations does not correspond to the standard state concentration of 1 M. Consideration of MC-like species allows a helpful decomposition of the catalytic effect into preorganization and reorganization phases. In the preorganization phase, the substrates are brought together into the MC-like species, either in water or in the enzyme active site. In the reorganization phase, the roles of the enzymic and aqueous environments may be compared directly because reorganization of the substrate is about the same in both cases. Analysis of the electric field along the reaction coordinate demonstrates that in water the TS is destabilized with respect to the MC-like species because the polarity of the solute diminishes and consequently the reaction field is also decreased. In the enzyme, the electric field is mainly a permanent field and consequently there is only a small reorganization of the environment. Therefore, destabilization of the TS is lower than in solution, and the activation barrier is smaller.