Quantenchemische Rechnungen an Stickstoffbasen Teil 3*. Ab initio- und semiempirische Rechnungen zur Protonenaffinität einiger sekundärer, tertiärer sowie verzweigter primärer aliphatischer Amine

Abstract

Our investigations concerning the protonaffmity of aliphatic amines have been confined so far to primary unbranched species. In order to gain further information about the applicability and the limitations of the statements derived from the results of the previous calculations, we extended our investigations to some tertiary, secondary and branched primary amines. Correlations between the experimental protonaffinities and the Mulliken charges of the nitrogen atom and the NHx-group (x = 0, 1, 2) of the amines, as well as those of the NHx + 1 ,-group of the corresponding ammonium cations have been found. However, these correlations only hold within the different groups of amines (primary, secondary and tertiary) and not between them. A somewhat better correlation throughout all groups is found if one compares the charges of the protons in the NHx + 1-groups of the ammonium cations with the experimental protonaffinities of the parent amines but deviations from the expected order are found in this case, too. This shows the limited value of charges as calculated by the methods used here, in predicting the relative order of protonaffinities of amines if they belong to different groups.

Calculation of the Proton Affinities of Primary, Secondary, and Tertiary Amines Using Semiempirical and ab initio Methods

“…In this case the mean difference between the calculated (PA cal ) and measured ( Nonempirical calculations at the Hartree-Fock level employing standard structural parameters [20] and minimal basis sets like STO-3G result in values which also correlate nicely (r = 0.9955) with the experimental proton affinities (Fig. 2) although they are much higher than the corresponding experimental values (∆PA = 48.9kcal/mol, see also [21,22]). Moreover, the correlations within the groups of the primary (line a) and secondary amines (line b) are better than the overall correlation.…”

“…The only method besides MSINDO which correctly predicts the order of the proton affinities of the mono-, di-, and trimethyl-and ethyl amines (RNH 2 < R 2 NH < R 3 N) is AM1 [24], while the overall correlation is still rather poor. PM3 also gives the correct order for the ethyl amines, however the calculated proton affinities of the corresponding methyl compounds erroneously decrease in the order MeNH 2 > Me 2 NH > Me 3 N. Both MINDO/3 [21,25] and MNDO [26] yield proton affinities which increase from the mono-to the dimethyl-and the diethyl compounds. However, with both methods the proton affinities are lower for the tri-than for the dialkyl bases.…”

“…For numbering of compounds see Figure 1. and EtNH 2 [19]. Their values (215.1, 221.7, 227.1, and 218.2 kcal/mol) are essentially identical with our results.Nonempirical calculations at the Hartree-Fock level employing standard structural parameters[20] and minimal basis sets like STO-3G result in values which also correlate nicely (r = 0.9955) with the experimental proton affinities (Fig.2) although they are much higher than the corresponding experimental values (∆PA = 48.9kcal/mol, see also[21,22]). Moreover, the correlations within the groups of the primary (line a) and secondary amines (line b) are better than the overall correlation.Complete geometry optimizations at the DFT level employing the B-P86 functional[23] and a valence triplebasis set (TZVP) including polarization functions result in an overall correlation (r = 0.9966, k = 0.9931) which is only slightly worse than the one obtained at the ZPE+MP2/6-311++G**//HF/6-311++G** level of conventional ab initio theory (Table 1, Fig.3), while the mean difference between…”