The synthesis and binding properties for carbohydrates of several artificial, acyclic receptors containing two or three heterocyclic recognition units covalently attached to a phenyl spacer is described. These host molecules having uncharged hydrogen-bonding sites were used in a systematic study towards the evaluation of recognition motifs for carbohydrates. A novel effective, acyclic hydrogen-bonding receptor possessing naphthyridine-amide moieties as heterocyclic recognition units has been developed.

The mechanisms of cycloaddition of thioformaldehyde S-methylide and thioacetone S-methylide, as models for an alkyl-substituted ylide, to thioformaldehyde and thioacetone, as well as to ethene as a model for a C=C double bond have been studied by ab initio calculations. Restricted and unrestricted B3LYP/6-31G* calculations were performed for the geometries of ground states, transition structures, and intermediates. Although basis sets with more polarization functions were tested, the 6-31G* basis set was applied throughout. Single-point CASPT2 calculations are reported for analysis of the unsubstituted system. The stabilities of structures with high biradical character seem to be overestimated by DFT methods in comparison to CASPT2. The general trends of the results are independent of the level of theory. Thioformaldehyde adds to thioformaldehyde S-methylide without activation energy, and the activation energies for two-step biradical pathways to 1,3-dithiolane are low. C,S biradicals are more stable than C,C biradicals. The two-step cycloaddition is not competitive with the concerted cycloaddition. Methyl substitution in the 1,3-dipole and the dipolarophile does not change the mechanistic relationships. TSs for the concerted formation of the regioisomeric cycloadducts of thioacetone Smethylide and thioacetone were located. Concerted addition remains the preferred reaction. The reactivity of the C=S double bond is high relative to that of the C=C double bond.

RHF/3-21G ab initio, density functional theory (B3LYP/3-21G), and single-point calculations using an effective core potential (B3LYP/6-31G*-ECP(S)//3-21G) are performed for different mechanisms of polymerization of acrylic acid and methyl acrylate by dicyclopentadienyl−zirconocene enolate complexes. Polymerization is considered involving a cationic zirconocene complex and a neutral zirconocene complex, respectively. In addition, a mechanism proposed by Collins is treated which involves a neutral and a cationic zirconocene complex in the CC bond-forming step at the same time. Catalytic cycles can be devised in all cases which suggest that all three mechanisms may be realizable under suitable conditions. However, the “bimetallic” mechanism proposed by Collins shows the lowest energy of activation for the propagation step. It is shown that bridging of the cyclopentadienyl rings should be important not only with respect to reactivity but also with respect to the possible catalytic cycle. Methyl methacrylate is not included in the treatment; i.e., the problem of tacticity in these polymerizations is not addressed.

Non-heme iron(III) complexes of 14-membered tetraaza macrocycles have previously been found to catalytically decompose hydrogen peroxide to water and molecular oxygen, like the native enzyme catalase. Here the mechanism of this reaction is theoretically investigated by DFT calculations at the (U)B3LYP/6-31G* level, with focus on the reactivity of the possible spin states of the FeIII complexes. The computations suggest that H2O2 decomposition follows a homolytic route with intermediate formation of an iron(IV) oxo radical cation species (L.+FeIV==O) that resembles Compound I of natural iron porphyrin systems. Along the whole catalytic cycle, no significant energetic differences were found for the reaction proceeding on the doublet (S=1/2) or on the quartet (S=3/2) hypersurface, with the single exception of the rate-determining O--O bond cleavage of the first associated hydrogen peroxide molecule, for which reaction via the doublet state is preferred. The sextet (S=5/2) state of the FeIII complexes appears to be unreactive in catalase-like reactions.

The cycloaddition of thiobenzophenone S-methylide to thiobenzophenone, an experimentally well-known reaction, was studied, using (U)HF/3-21G* for finding stationary points and (U)B3LYP/6-31G*//(U)HF/3-21G* single-point calculations for energies. Some optimizations were performed by (U)B3LYP/ 6-31G* to check the reliability of the calculations. The comparison of the concerted pathways and stepwise reactions via C,C-biradicals and C,S-zwitterions showed that the formation of a tetraphenyl-substituted C,C-biradical and its ring closure to 4,4,5,5-tetraphenyl-1,3-dithiolane constitutes the energetically most probable pathway of product formation, despite the fact that the regioisomeric 2,2,4,4-tetraphenyl-substituted product is more favorable by 17 kcal mol(-1). Model calculations on bond dissociation energies showed that (U)B3LYP with various basis sets overestimates radical stabilization, whereas CBS-QB3 closely reproduced experimental values. Results with the BLYP functional are similar to those with B3LYP. The consequences of the overestimation of radical stability for the cycloaddition mechanism involving biradicals are discussed. Thiobenzophenone S-methylide, if not captured by a dipolarophile, dimerizes to 2,2,3,3-tetraphenyl-1,4-dithiane. Calculation disclosed likewise a tetraphenyl-substituted C,C-biradical as intermediate.