Zt. Gu et al., HYDROGEN-BONDING EFFECTS ON AMINE ROTATION RATES IN CRYSTALLINE AMINO-ACIDS, Solid state nuclear magnetic resonance, 7(3), 1996, pp. 161-172
Rates of rotation for amines in a variety of crystalline environments
are reported, and the trends are explained in terms of the strengths o
f local hydrogen bonding interactions. Proton spin-lattice relaxation
times (T-1) and deuterium broad-line NMR spectra have been measured fo
r D-, DL- and L- aspartic acid, two polymorphs of glycine, alanine, an
d leucine in the temperature range from -40 to 110 degrees C. The ener
gy barriers for amine rotation are 27 +/- 2 kJ mol(-1) for D- or L-asp
artic acid and 22 +/- 2 kJ mol(-1) for DL-aspartic acid; these energie
s are slightly lower than the previously reported value for the L form
based on direct proton T-1 measurements at 60 MHz. The values for the
alpha and gamma forms of glycine were 24 +/- 2 and 30 +/- 2 kJ mol(-1
) respectively, that for L-alanine was 40 +/- 2 and that for L-leucine
was 49 +/- 3 kJ mol(-1). These are all in rough agreement with previo
usly reported values (although the differences for the polymorphs of g
lycine and for L- vs. DL-aspartic acid were not reported). Crystal str
uctures of these amino acids indicate differences in hydrogen bonding
environments around the R-NH3+ groups that are probably responsible fo
r the different activation barriers. A molecular mechanics calculation
of the rotation energy barriers for L- and DL-aspartic acid based on
the crystal structures gave satisfactory agreement with experimental r
esults if a uniform (and arbitrarily chosen) dielectric constant of 2.
5 was assumed. Differences between L- and DL-aspartic acids and betwee
n two polymorphs of glycine were well represented qualitatively. Inclu
ding additional neighboring molecules not involved in the hydrogen bon
ding or including periodic boundary conditions to describe the crystal
packing did not significantly affect these results. If vacuum dielect
ric constants are used, the barriers are uniformly overestimated, and
if the experimental macroscopic dielectric constant values are used, t
he barriers are generally underestimated. Dielectric constants differ
substantially from one amino acid to another and significantly affect
the estimated barriers; in fact, the bulk dielectric constants appear
to be the major difference between the highest and the lowest values.
The difficulty of accurately including dielectric relaxation into mole
cular mechanics calculations resulted in the disagreement between expe
rimental measurements and theoretical calculations.