Nt. Anh et al., A THEORETICAL-STUDY OF ALKYL SUBSTITUENT EFFECTS IN INTRAMOLECULAR SN2 REACTIONS .2. SUBSTITUTIONS IN NONIDENTICAL EXCHANGE-REACTIONS, New journal of chemistry, 18(4), 1994, pp. 483-487
This paper examines the effects of alkyl substituents on S(N)2 transit
ion structures, using NH3, NH2Me, NHMe2 and NMe3 as nucleophiles and M
eCl, EtCl, i-PrCl, i-BuCl and t-BuCl as substrates. The reactions of M
e- and Et- with MeCl and EtCl are taken as model systems that give ear
ly transition states. In late transition states, according to ab initi
o (3-21 G or 6-31 G) calculations, substituting the nucleophile lengt
hens the (nucleophile-reaction center) bond and shortens the (reaction
center-leaving group) bond by approximately 0.115 angstrom. Monosubst
itution of the reaction center lengthens the Nu-C and C-Cl bonds by ap
proximately 0.10 angstrom and approximately 0.05 angstrom, respectivel
y. The leaving group is also deviated by approximately 15-degrees. Dis
ubstitution of the reaction center leads to an angular deviation of ap
proximately 10-degrees for both the nucleophile and the leaving group.
The bond lengthening is approximately 0.20 angstrom for Nu-C and appr
oximately 0.10 angstrom for C-Cl. In early transition states, substitu
ting the nucleophile leaves the TS practically unchanged. The reason f
or this is that steric hindrance is markedly reduced by the ''outwards
'' pyramidalization of the reagent and by the large intermolecular dis
tance. Monosubstitution of the reaction center leads to a deviation of
approximately (0.10-degrees for the nucleophile and no Nu-C bond leng
thening. The leaving group is deviated by approximately 5-degrees and
the C-Cl bond lengthens by approximately 0.05 angstrom. The estimated
maximal values are approximately 20-degrees for the angular deviation
and approximately 0.20 angstrom for the bond lengthening. These are sl
ightly larger than the values observed for identity exchanges and are
due to the dissymmetrical nature of the transition structures. Substit
uting both the nucleophile and the reaction center gives approximately
additive effects. Even with neutral nucleophiles likes amines, AM1 ca
lculations still overestimate the basicity and underestimate the nucle
ophilicity. Compared with the 6-31G results, the AM1 central barriers
, which were slightly on the lower side for charged nucleophiles, beco
me systematically too high when the reagent is neutral. This may be tr
aced back to the fact that AM1 transition states are not only tighter
but also much later than ab initio (6-31 G or 3-21 G) transition stat
es, as indicated by the more advanced Walden inversion and LG departur
e. The combination of tightness and lateness enhances steric hindrance
and raises the central barrier. The 3-21G barriers are again too low,
although not so badly for neutral as for charged nucleophiles. In fac
t, with the standard 3-21G basis set, carbanions react with RCl withou
t a barrier and, in order to obtain transition states, the electronega
tivity of the charged carbon atom must be raised (by setting its nucle
ar charge Z equal to 6.5 or 6.8).