T. D. W. Claridge et al. / Tetrahedron: Asymmetry 24 (2013) 947–952
949
(in 0.1 equiv portions) to the solution of (S)-6Li-15N-827 produced
no significant difference in the 5.5:1.0 ratio of 9 to 10, again consis-
tent with this being the equilibrium ratio but hence precluding any
potential insight into the relative consumption of 9 versus 10 dur-
ing the course of the conjugate addition reaction. Interestingly,
however, two new doublet signals appeared in the 6Li NMR spec-
trum (at dLi 0.35 and 1.36 ppm) as tert-butyl cinnamate was added;
in the presence of 0.5 equiv of tert-butyl cinnamate these were the
only detectable signals, suggesting that a 1:1 lithium b-amino (Z)-
enolate/lithium amide complex (in which each 6Li nucleus is con-
nected to only one 15N nucleus) had formed in preference to either
of dimers 9 or 10 (Fig. 2).
dimers 11 and 12 (in a 2:1 ratio, respectively). A 6Li–6Li EXSY
experiment indicated exchange between both of the ‘outer’ triplets
and the ‘inner’ triplet, supporting the presence of 11 (with the Li
nuclei in non-equivalent environments) which undergoes ex-
change via a non-dissociative mechanism with 12 (with the Li nu-
clei in equivalent environments). Direct exchange between the two
triplet peaks associated with 11 was not observed and neither was
there any evidence for exchange of 9 or 10 with 11 or 12, again
consistent with full dissociation and reformation of the various di-
mers being slow relative to the NMR timescale. The 15N NMR spec-
trum17 of (RS)-6Li-15N-8 at 173 K revealed the presence of four,
1:2:3:2:1-quintet signals in the ratios 6.3:5.2:2.6:1.0. As before,
two of these signals correspond to dimers trans-(RS,RS)-9 and cis-
(RS,RS)-10, in the ratio 6.3:1.0, whilst the 2:1 ratio of the other
two signals compares favourably with the ratio of cis-(RS,SR)-11
to trans-(RS,SR)-12 obtained from the 6Li NMR spectrum. From
these data, it can be concluded that (RS)-6Li-15N-8 exists as a mix-
ture of all possible diastereoisomeric dimers 9, 10, 11 and 12 in the
ratio 6.3:1.0:5.2:2.6, respectively. The ratio of 9 to 10 is 6.3:1.0 [in
contrast to 5.5:1.0 observed for (S)-6Li-15N-8] and the ratio of 11 to
12 is 2.0:1.0, whilst the ratio of (9 + 10) to (11 + 12) is 1.0:1.132
(Fig. 3).
Variable temperature 6Li and 15N NMR spectra (recorded be-
tween 173 K and 273 K, in 10 K intervals) revealed similar behav-
iour: peak broadening was initially noted, with eventual
coalescence at approximately 223 K, consistent with much more
rapid exchange at higher temperatures. A single peak was observed
at 273 K and is attributed to the time averaged populations of 6Li
and 15N nuclei in various oligomers at this temperature. It is note-
worthy that the 5.5:1.0 ratio of 9 to 10 decreased before coales-
cence, indicating an increase in the population of 10 at higher
temperatures: for example, at 193 K the ratio of 9 to 10 was
3.6:1.0. From these VT 6Li NMR spectra it is possible to evaluate
the approximate value of the free energy for the process that ef-
fects the coalescence of the two triplet peaks attributed to dimer
9.28 The Eyring equation may be written as:
Variable temperature 6Li and 15N NMR spectroscopic analysis of
(RS)-6Li-15N-8 showed similar behaviour to that observed for
(S)-6Li-15N-8, with initial line broadening and eventual coalescence
being observed as the temperature of the sample was increased,
consistent with an increased rate of exchange between dimers
9–12. Empirical analysis also indicated a change in the dimer ratio
at higher temperatures, consistent with a shift in the position of
equilibrium. The simultaneous increase in the rate of exchange
and shift in the position of equilibrium precluded calculation of
the rates of exchange (and hence free energy of activation) from
line-broadening analysis in this case.
D
Gz ¼ RT½lnðkBT=kchÞꢃ;
where T is the temperature of coalescence (in K), kc is the rate of
coalescence (in Hz), and the other symbols have their usual mean-
ings.29 An approximate value of kc may be determined directly from
the NMR spectrum as:
p
kc ¼
D
dvð
p=
2Þ;
3. Conclusion
where
D
dv is the difference in the chemical shifts of the two peaks
in question (in Hz).30 Taking the coalescence temperature as 223 K
and the difference in the chemical shifts of the two triplet signals
In conclusion, an efficient procedure for the synthesis of
(S)-15N-benzyl-15N-(
a-methylbenzyl)amine has been devised,
associated with 9 as 76.5 Hz, then
D
Gà ꢄ +45 kJ molꢀ1 31
.
Analysis of (RS)-6Li-15N-8 by 6Li NMR spectroscopy17 at 173 K
gave a more complex spectrum compared to that obtained from
enantiopure (S)-6Li-15N-8. Not unexpectedly, signals associated
with dimers trans-(RS,RS)-9 and cis-(RS,RS)-10, comprising two
lithium amide monomers with the same configuration, were pres-
ent. However, for (RS)-6Li-15N-8 there also exists the possibility of
formation of dimers cis-(RS,SR)-11 and trans-(RS,SR)-12, compris-
ing two lithium amide monomers with opposite configurations.
Subtraction of the 6Li NMR spectrum of (S)-6Li-15N-8 from that of
(RS)-6Li-15N-8 gave a spectrum containing three triplet peaks cen-
tred on dLi 1.37, 1.97 and 2.62 ppm in a respective 1:1:1 ratio; it
was postulated that these signals arose due to the presence of
starting from (S)-2-phenylpropanoic acid and using a Hoffman-
type rearrangement as the key step. Deprotonation of a solution
of this amine in THF-d8 with Bu6Li at low temperature gave a solu-
tion of the corresponding doubly labelled 6Li/15N lithium amide,
which is known to be a versatile enantiopure ammonia equivalent
in conjugate addition reactions to a range of
a,b-unsaturated
esters. Analysis of this species by 6Li and 15N NMR spectroscopy re-
veals the presence of lithium amide dimers as the only observable
species. Although a lithium amide dimer may not necessarily rep-
resent the reactive species for conjugate addition reaction, either a
monomeric or dimeric reactive species is accommodated by our
transition state mnemonic for this class of reaction. As rapid lith-
ium exchange within the dimers (via a ring-opening and ring-clos-
ing process) was observed, this suggests that the active species for
conjugate addition is most likely either a ring-open or ring-closed
dimeric lithium amide. The ready distinction of the various dimers
of this lithium amide, as compared to achiral (e.g., lithium dib-
CO2tBu
6Li
6Li
Ph
15
R215N
*
O
Ph
*
NR2
Ph
15N
6Li
(0.5 equiv)
enzylamide) and C2-symmetric (e.g., lithium bis-N,N-a-methylben-
zylamide) counterparts, presents unique opportunities for further
mechanistic study.
Ph
OtBu
(S)-6Li-15N-8
putative 1:1 lithium β-amino (Z)-enolate/lithium
amide complex [15NR2 = (S)-15N-benzyl-
*
solution in d8-THF
15N-(α-methylbenzyl)amino]
4. Experimental section
Ph
Ph
Ph
Ph
9 : 10 Ph
5.5 : 1.0
Ph
Ph
Li
Li
Li
Li
N
N
N
N
4.1. General experimental details
Ph
Reactions involving moisture-sensitive reagents were carried
out under a nitrogen atmosphere using standard vacuum line
techniques and glassware that was flame dried and cooled under
trans-(S,S)-9
-(
,
)-10
cis S S
Figure 2. Addition of tert-butyl cinnamate to a solution of (S)-6Li-15N-8.