78
Y.-P. Niu et al. / Journal of Molecular Structure 1092 (2015) 72–80
can be accomplished through one-ring flip of one of two anti-arms.
The PB helical changes of SAA conformer would also promise the
two anti-arms equal opportunities to flip. Therefore, it was
regarded that the inter-conformational conversions between the
SSA and SAA conformers, a one-ring flip process, would control
the overall arm exchanges rates of them both. And as a result, these
would ultimately surpass or flatten the difference in donor
exchange rates between them, indicating that they will behave like
one conformer. The strong resonances in the spectra were once
thought to be correlated only with SSA conformer. However, this
easy assignment is obviously inconsistent with the above analysis
that the arm exchanges rates of inter-conformational process
(one-ring flip, SSA M SAA) are faster than that of the intra-con-
formational process (two-ring flip, the PB process of SSA and the
HBB process of SAA). That is to say, on the basis of combined con-
siderations of both intra- and inter-conformational changes, we
would like to believe that protons of SAA conformer should also
have contributions to these strong signals.
N
N
Helical Change
N
N
OH
N
OH
N
M-propeller
P-propeller
H
O
time-averaged conformation
Proton Sponge
H
O
O
H
The internal steric hindrance between the three anti-arms were
thought to be the only factor which stabilizes AAA conformer and
distinguishes itself from the other conformers. AAA conformer was
thought to be thermodynamically unstable because of a high inter-
nal energy. Logically, this conformer can only be generated from
SSA conformer. The conversion needs firstly the breaking of
intra-molecular hydrogen bond of SAA conformer, then following
the ring flip of the only syn-arm (Fig. 3. blue colored). And
therefore, it should be a minor conformer. The resonances of the
hydroxyl and aryl protons of the AAA conformer might be too weak
to be visible or be overlapped by the resonances of the other con-
formers (Fig. 4 and S1.). However, in the high field region of spectra
(Fig. S2a), the emerging signal, which presents firstly as a sharp
shoulder peak, then grows up gradually to a separate peak and
finally vanished with temperature increasing, can be assigned to
the methyl protons of this specific conformer. The variation trend
of the methyl resonance was thought to be consistent with the
opposing effects of the increased temperature on the thermal sta-
bility and the dynamic generation of this specific conformer. The
broadened methyl signals in high temperature spectra should indi-
cate the configuration changes of nitrogen atoms (between trigonal
pyramidal and planar), which is caused by the increase of the
molecular thermal motion ability. With the temperature decreas-
ing, the sharp trend of its methyl signal should indicate a frozen
conformation of C3 symmetry. The helical conversions of AAA con-
former were thought, at least in low temperature condition, to be
difficult because of the large steric hindrance and the low thermal
motion ability. With temperature increasing, AAA conformer
reaches its highest proportion at around À30 °C with a ratio of
3.80/100, calculated by the integrated area of its methyl signals
with respective to that of the strong methyl signals at around
2.26 ppm.
Scheme 3. Assumed conformational changes and time-averaged conformation of
SSS conformer It was generally agreed that TAMs exist as racemic mixtures of M-
and P-propeller structures [8].
arm exchanges). The PB helical changes of SAA conformer are low-
barrier intra-conformational conversions following zero-ring flip
pathway. However, the HBB helical changes of SAA conformer
could occur only if the hydrogen-bond barrier and the two-ring flip
barrier are crossed simultaneously.
The intra-conformational analysis shows that in SAA con-
formational states, the nitrogen atom of the only syn-arm has more
opportunities to form hydrogen bond with the hydroxyl proton
than that of the other two anti-arms. The intra-conformational
analysis also shows that the PI process of SSA conformer should
be more fast than the HBB process of SAA conformer. The two con-
siderations reveal a significant difference between the overall
donor (N) exchange rates between SSA and SAA conformers,
indicating that the time-averaged chemical shifts of their attached
protons should also be different. The above analysis show that only
intra-conformational changes cannot meet the requirement of
chemical equivalence and, the inter-conversions between the both
conformers should be considered as factors.
The inter-conformational conversions via two-ring flip process,
SSS M SAA and SSA M AAA, were thought to be unlikely to occur,
because of a relatively high transition state barrier. As already
mentioned, the interconversions between the enantiomeric pair
of SSA and SAA conformers all need to break the specific con-
formational states of individual conformer. If not considering these
inter-conformational conversions between SSA conformer and SAA
conformer, there should be a big difference in probabilities that
their nitrogen atoms of syn-arms and anti-arms to form hydrogen
bonds with the hydroxyl proton. The interconversions of the enan-
tiomeric pair of the SSA and SAA conformers can be accomplished
via two one-ring flip process, respectively through SSS and AAA
conformational states. While, the conversions between twelve con-
formational states of SSA conformer and that of SAA conformer
(Scheme 2 and 4), each to each, only need one one-ring flip process
(Fig. 3, equilibrium black colored). The conversion from SSA con-
former to SAA conformer can be easily accomplished through the
one-ring flip of the non-binding syn-arm. Or else, this conversion
can also be accomplished through a PI helical change of SSA con-
former, and then a subsequent ring flip of the newborn non-bind-
ing syn-arm. That indicates that PI helical changes would afford
equal opportunities for the two syn-arms of SSA conformer to flip.
Similarly, the conversion from SAA conformer to SSA conformer
Additional variable-temperature 1H NMR experiments, from
À60 to 60 °C, were carried out in methanol-d4 (Fig. S7). The
variation trends of conformers in methanol are similar to that in
chloroform (Fig. S2), except the absence of AAA conformer.
Calculation results show that the steric hindrance among the
groups, as well as thus produced the inward orientation of the
lonely pair of nitrogen atoms, counts against the tertiary alcohol
to form the NÁ Á ÁHAOCH3 inter-molecular hydrogen bonds with
methanol (solvent). Instead, methanol would like to form inter-
molecular hydrogen bonds with the hydroxyl group of the tertiary
alcohol (Fig. 5). Comparing with the calculation results of the pre-
ceding models (Fig. 3, first row), the relative smaller energies of the
present models in Fig. 5, 119.746 (SSS3), 116.128 (SSA3), 122.119
(SAA3) and 137.472 (AAA3) kcal/mol, should indicate more stable
conformations. At the same temperature, the line width of the
methyl signal of SSS conformer in methanol is broader than that