Eclipsed Conformation of the Exocyclic N-CH2 Bond in N-Neopentylpiperidines and the Stereodynamic Consequences As Studied by Dynamic NMR Spectroscopy and Molecular Mechanics Calculations

Article abstract of DOI:10.1021/jo9720344

The dynamic stereochemistry of a range of N-neopentylpiperidines 1 with an eclipsed N-CH₂-t-Bu bond is compared with that of the corresponding range of N-ethylpiperidines 2 with a gauche N-CH₂Me bond. By using dynamic NMR spectroscopy, the relative importance of ring inversion, exocyclic bond rotation, and nitrogen inversion are elucidated and barriers are reported and discussed. Minor populations of the neopentyl-axial-eclipsed conformation are detected directly and identified with the help of molecular mechanics calculations. The corresponding N-alkylpyrrolidines are also reported.

Full text of DOI:10.1021/jo9720344

3
310
J . Org. Chem. 1998, 63, 3310-3317
2
Eclip sed Con for m a tion of th e Exocyclic N-CH Bon d in
N-Neop en tylp ip er id in es a n d th e Ster eod yn a m ic Con sequ en ces As
Stu d ied by Dyn a m ic NMR Sp ectr oscop y a n d Molecu la r Mech a n ics
Ca lcu la tion s
J . Edgar Anderson,* Anthony I. Ijeh, and Christine Storch
Chemistry Department, University College London, Gower Street, London WC1E 6BT, U.K.
Daniele Casarini and Lodovico Lunazzi*
Dipartimento di Chimica Organica “A. Mangini”, Universit a` di Bologna, Viale Risorgimento 4,
Bologna 40136, Italy
Received November 5, 1997
The dynamic stereochemistry of a range of N-neopentylpiperidines 1 with an eclipsed N-CH
Bu bond is compared with that of the corresponding range of N-ethylpiperidines 2 with a gauche
N-CH Me bond. By using dynamic NMR spectroscopy, the relative importance of ring inversion,
2
-t-
2
exocyclic bond rotation, and nitrogen inversion are elucidated and barriers are reported and
discussed. Minor populations of the neopentyl-axial-eclipsed conformation are detected directly
and identified with the help of molecular mechanics calculations. The corresponding N-alkylpyr-
rolidines are also reported.
In tr od u ction
There has been much recent interest in molecules that
cube that encompasses these possibilities and shows the
intermediate structures II-IV and VI-VIII, which may
be involved and may or may not be significantly popu-
lated. Dynamic NMR is suitable for studying the overall
interconversion of I and its equivalent form V, for
geminal protons on the ring are diastereotopic when
interconversion of these structures is slow on the NMR
time scale and this is in fact observed at suitably low
temperatures.
Thus, if ring inversion first becomes slow on the NMR
time scale, I and less stable structures II, IV, and VII
on the left-hand side are distinct from V (and III, VI,
and VIII) on the right-hand side of the cube. If it is
nitrogen inversion that first becomes slow, structure I
and others on the top face become distinct from structure
V and others on the bottom face of the cube. If it is
contain saturated bonds that prefer to adopt an eclipsed
conformation.1 Particularly striking is the demonstra-
tion that the exocyclic N-C bond in N-neopentylpiperi-
-3
3
2i,j
dine and analogues and the exocyclic C-O bond in tert-
butoxycyclohexane2h are eclipsed.
Such eclipsing has possible consequences for the in-
terconversion of molecular conformations,2 so we have
synthesized and investigated a set of compounds chosen
to show a range of stereodynamic behavior, N-neopen-
tylpiperidine and some of its derivatives 1a -e, the
equivalent N-ethylpiperidines 2a -e as noneclipsed com-
parisons, and the two N-alkylpyrrolidines 1f and 2f.
f,3b
O
O
O
CH3
CH3
CH3
(1) (a) Rabideau, P. W.; Lipkowitz, K. B.; Nachbar, R. B., J r. J . Am.
Chem. Soc. 1984, 106, 3119. (b) Dhar, R. K.; Sygula, A.; Fronczek,
F.R.; Rabideau, P. W. Tetrahedron 1992, 48, 9417. (c) J uaristi, E.;
Martinez, R.; Mendez, R.; Toscano, R. A.; Soriana-Garcia, M.; Eliel,
E. L.; Petsom, A.; Glass, R. S. J . Org. Chem. 1987, 52, 3806. (d)
Gordillo, D.; J uaristi, E.; Martinez, R.; Toscano, R. A.; White, P. S.;
Eliel, E. L. J . Am. Chem. Soc. 1992, 114, 2157. (e) Cremonini, M. A.;
Lunazzi, L.; Placucci, G.; Okazaki, R.; Yamamoto, G. J . Am. Chem.
Soc. 1990, 112, 2915. (f) Goren, Z.; Biali, S. E. J . Am. Chem. Soc. 1990,
112, 893. (g) Columbus, I.; Hoffman, R. E.; Biali, S. E. J . Am. Chem.
Soc. 1996, 111, 6891 and earlier work cited therein.
N
N
N
N
N
N
CH2R
CH2R
CH2R
CH2R
CH2R
CH2R
a
b
c
d
e
f
1
, R = tert-butyl 2, R = methyl
CH2tBu
N
H2C
N
CH2
(
2) (a) Anderson, J . E. J . Chem. Soc., Perkin Trans. 2 1991, 299.
N
(
(
b) Anderson, J . E.; Watson, D. G. J . Am. Chem. Soc. 1992, 114, 1517.
c) Anderson, J . E. J . Chem. Soc., Perkin Trans. 2 1992, 1343. (d)
t-BuH2C
C
H2
CH2tBu
Anderson, J . E. J . Chem. Soc., Perkin Trans. 2 1993, 441. (e) Anderson,
J . E.; Corrie, J . E. T.; Lunazzi, L.; Tocher, D. A. J . Am. Chem. Soc.
3
1
993, 115, 3493. (f) Anderson, J . E.; Casarini, D.; Corrie, J . E. T.;
Lunazzi, L. J . Chem. Soc., Perkin Trans. 2 1993, 1299. (g) Anderson
J . E. Conformational Analysis of Acyclic and Alicyclic Saturated
Hydrocarbons. In The Chemistry of Alkanes and Cycloalkanes; Patai,
S., Rappoport, Z., Eds.; Wiley: Chichester, 1992; Chapter 3.II.D. (h)
Anderson, J . E.; Ijeh, A. I. J . Chem. Soc., Perkin Trans. 2 1994, 1965.
Scheme 1 shows the ground-state chair conformation
I of N-neopentylpiperidine 1a , which is known to have
2
the substituent equatorial and the exocyclic N-CH -t-
Bu bond eclipsed.3 There is an equivalent form V
a
(i) Anderson, J . E.; Casarini, D.; Ijeh, A. I.; Lunazzi, L.; Tocher, D. A.
reached from I by three processes, namely ring inversion,
nitrogen inversion, and N-CH bond rotation, which,
2
putatively, can take place in any sequence, some more
likely than others. Scheme 1 is thus a conformational
J . Am. Chem. Soc. 1995, 117, 3054. (j) Anderson, J . E.; Casarini, D.;
Ijeh, A. I.; Lunazzi, L. J . Am. Chem. Soc. 1997, 119, 8050.
(
3) (a) Forsyth, D. A.; J ohnson, S. M. J . Am. Chem. Soc. 1993, 115,
364. (b) Forsyth, D. A.; J ohnson, S. M. J . Am. Chem. Soc. 1994, 116,
11481.
3
S0022-3263(97)02034-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/30/1998

Exocyclic N-CH
2
Bond in N-Neopentylpiperidines
J . Org. Chem., Vol. 63, No. 10, 1998 3311
Ta ble 1. MM3-Ca lcu la ted Rela tivea En th a lp ies ERel a n d
b
Sch em e 1
c
Exocyclic Bon d Con for m a tion a l Min im a for
N-Neop en tyl- a n d N-Eth ylp ip er id in e a n d P yr r olid in e
Der iva tives 1 a n d 2 (Selected MMX Resu lts Ar e Sh ow n in
P a r en th eses)
equatorial minima
eclipsed gauche
anti
axial min
measured
ERel
φ
ERel
φ
ERel
φ
ERel
φ
ERel
1
1
1
1
1
a
b
c
d
e
0.00
2
3.60 171 2.12 (2.24)
3.53 172 2.13 (2.57)
3.67 171 4.05 (4.44)
3.68 172 1.98 (2.51)
9
1.04
0.78
0.00 13
0.00 16
0.00 17
0.00 12
16
14
9
1
9
0.03
1.35
3.56 172 1.15d (1.24)
e
2
.10 (2.31)
1
2
2
2
2
2
f
0.00
0
3.13 166
a
b
c
d
e
0.00 51 0.76 179
0.00 53 0.77 179
0.00 51 0.79 179
0.00 52 0.77 180
0.00 51 0.77 179
2.25
2.29
5.04
1.95
1.40d
53
53
-42
51
50
50
f
f
f
f
f
f
2
.30e
2
f
0.00 54 1.48 179
N-CH
2
bond rotation that first becomes slow, structure
a
The MM3 final steric energy (kcal/mol) of the most stable
I on the rear face becomes distinct from structure V on
the front face. Geminal protons becoming diastereotopic
thus does not indicate which of the three processes has
become slow on the NMR time scale. The answer may
emerge from other evidence and turns out to be different
for different N-alkylpiperidines studied. If forms such
as II, III, and IV are of sufficiently low relative energy
to be significantly populated, a second set of dynamic
NMR changes due to slowing of a second of these
processes on the NMR time scale may be observed at even
lower temperatures, resulting in two sets of NMR signals
of different intensity.
We will show examples with separate sets of spectra
for a minor conformation, but in another case, when the
minor population is less than about 2%, the slowing of
the interconversion will appear only as a broadening of
signals, followed by their narrowing once again as the
temperature is progressively lowered, the expected minor
signal being too weak to be discerned from noise.4
Finally, when there is no explicit or implicit NMR
evidence for them, other minor conformations will be
described only from molecular mechanics calculations of
their structure and energy.
conformation is as follows 1a , 21.92; 1b, 22.88; 1c, 25.41; 1d , 22.08;
1e, 36.65; 1f, 31.55; 2a , 14.82; 2c, 18.44; 2d , 14.99; 2e, 29.51; 2f,
2
3.85. b kcal/mol. c φ is the lp-N-C-t-Bu torsion angle in com-
d
pounds 1, the lp-N-C-Me torsion angle in compounds 2. Methyl
e
axial, φ for the equatorial CH2R group. N-CH2R group axial.
f
Separate sets of signals for different kinds of conformations were
not observed for compounds 2.
conformational minima, eclipsed, anti, or gauche for the
compounds 1a -f and 2a -f, in both CH
CH R-axial conformations. Some axial/equatorial energy
differences were also determined by MMX calculations,
2
R-equatorial and
2
7
and values are shown in parentheses. For neopentylpi-
peridines 1a -e, the equatorial-neopentyl conformations
I and V with the tert-butyl group eclipsing the lone pair
are the most stable, with the axial eclipsed conformations
III and VII the next most likely. No minima are found
with a gauche arrangement, and the anti-equatorial
conformations II and VI are about 3.6 kcal/mol higher
in energy. Conformations such as IV and VIII and
nonchair conformations were calculated to be of so high
relative energy that they are not considered further nor
reported in Table 1.
The most stable conformations for neopentylpyrrolidine
1f are calculated to be gauche, where the lone-pair
2
The eclipsing of the N-CH bond has been demon-
3
2i
strated for 1a and for N,N′,N′′-trineopentyltriazane, 3,
but there is no simple NMR method for deciding whether
further analogues 1b-f are also eclipsed, although of
N-CH -t-Bu torsion angles are 22°. There is a barrier
2
of 0.31 kcal/mol to rotation to the eclipsed conformation,
which is only 0.24 kcal/mol less stable. The different
geometry around the nitrogen center in the five-mem-
bered ring reduces the compression that leads to eclipsing
in neopentylpiperidines and acyclic neopentylamines, and
there is a broad potential well with two gauche and one
eclipsed minima.
5
,6
course the bond environment is very similar. MM3 and
MMX molecular mechanics calculations have reliably
predicted the eclipsing found for 1a and 3 and so are
used to predict the structure and energy of the various
plausible stable conformations in the series of compounds
7
2
i,3
1
and 2.
For the N-ethylpiperidines 2, the conventional confor-
mations with a staggered equatorial N-ethyl group are
most stable, and the staggered axial conformation is
usually at least 2 kcal/mol less stable: in fact, it is
relatively more improbable than the axial neopentyl
group conformation in the series 1. Nitrogen inversion
Resu lts
Molecu la r Mech a n ics Ca lcu la tion s. Table 1 lists
relevant details of MM3 calculations of N-CH bond
2
(4) Anet, F. A. L.; Basus, V. J . J . Magn. Reson. 1978, 32, 339. See
also: Sandstr o¨ m, J . Dynamic NMR Spectroscopy; Academic Press:
London, 1982; p 84.
in this series 2 is accompanied by about 60° of N-CH
2
bond rotation to reach a new staggered conformation, so
(
5) Allinger, N. L.; Yuh, Y. H.; Lii, J .-H. J . Am. Chem. Soc. 1989,
11, 8551.
6) Allinger, N. L.; Rahman, M.; Lii, J .-H. J . Am. Chem. Soc. 1990,
12, 8293.
7) MMX force field as in PCMODEL, Serena Software, Blooming-
for 2 the cube of Scheme 1 should be modified to reflect
1
1
(
this point. Barriers to rotation about equatorial N-CH
bonds were calculated and are similar within each series
and 2, since changes in substitution are remote from
this bond. For the N-CH Me bond in equatorial confor-
2
R
(
1
ton, IN. See: Gajewski, J . J .; Gilbert, K. K.; McKelvey, J . In Advances
in Molecular Modelling; J AI Press: Greenwich, CT, 1992; Vol. 2.
2

3
312 J . Org. Chem., Vol. 63, No. 10, 1998
Anderson et al.
q
Ta ble 2. Dyn a m ic NMR Ba r r ier s (∆G , k ca l/m ol) to
For neopentylamines of which the compounds of series
are examples, there is no clear indication what barriers
Con for m a tion a l P r ocesses, Rin g In ver sion , R.I., Nitr ogen
In ver sion , N.I., or Bon d Rota tion , Rot.
1
to pure nitrogen inversion should be, since, from a
ground-state A this single process takes a neopentyl
group from an eclipsed conformation to an anti one, which
is unpopulated, so the process cannot be observed in a
dynamic NMR experiment. Rotation of the neopentyl
group through 180°, however, may then lead on to a
stable populated conformation B. The overall intercon-
version A / B may be studied by dynamic NMR, yielding
a barrier, but experience suggests that it is due to the
second, rotation process. Thus, it is known that ground-
state strain lowers barriers to nitrogen inversion, so the
strain that leads to eclipsed conformations for neopen-
tylamines may also lead to pure nitrogen inversion
barriers significantly lower in the series 1 than that of
compd barrier (T (°C)) process compd barrier (T (°C)) process
1
1
a
10.6 (-40)
.56 (-)
10.96 (-40)
R.I.
N.I.
R.I.
N.I.
R.I.
Rot.
R.I.
R.I.
N.I.
Rot.
2a
2b
11.50 (-15)
11.42 (-20)
R.I.
R.I.
7
b
8
.25 (-80)
1
1
c
d
11.8 (-35)
2c
2d
11.4 (-45)
7.60 (-100)
R.I.
N.I.
9.3 (-75)
6
.5 (-120)
9.6 (-40)
.0 (-80)
8.1 (-90)
1
1
e
f
2e
2f
no equilibrium
8
8.0 (-95)
N.I.
mations of 2a -f, bond rotation takes place in steps of
20° with a small barrier of ∼2.5 kcal/mol for a methyl
group rotating from +gauche past the lone pair to
gauche and of about 5 kcal/mol for rotation from gauche
1
8
.0 kcal/mol found for 2f. We have suggested that in
-
N,N′,N′′-trineopentyl-1,3,5-triazane a barrier of 6.4 kcal/
mol measured is to be associated with nitrogen inversion,
although a hindered neopentyl group rotation process
that should be detectable was not located.
The barrier of 8.1 kcal/mol measured for interconver-
sion of two equivalent conformations of neopentylpyyr-
rolidine 1f thus seems rather high to be attributed to
nitrogen inversion, but in good agreement with that of
2
to anti past the N-CH bond of the ring. This agrees
well with experimental measurements8,9 of such N-ethyl
rotational barriers.
2
In contrast for the N-CH -t-Bu bond in compounds
1
1
a -e, barriers of about 10 kcal/mol are calculated for
80° rotation from the eclipsed conformation I to the anti-
2
f,3b
conformation II. It has been pointed out previously
that eclipsed bonds often have high rotational barriers.
The interaction of the tert-butyl group with the equatorial
hydrogen atoms at C2 and C6 (e.g., H in I) causes both
A
8
.5 kcal/mol calculated for neopentyl group rotation. We
thus assign this barrier to such a rotation, noting that
calculations for the equivalent barrier in neopentylpip-
eridines that we will now discuss are about 1.5 kcal/mol
higher.
eclipsing and a relatively high barrier to rotation since,
in contrast to simple staggered bonds, rotation involves
the increase in this repulsion over 120° of bond rotation
until maximium interaction is reached near the point
where the tert-butyl group eclipses the C-C bond of the
1
The H spectrum of each of 1a -c shows broadening of
signals from the protons on the ring below about -20 °C
and eventually splitting into equal multiplets below about
-40 °C (see, for instance, the spectra of 1c in Figure 1).
ring. As an experimental point of reference, it is notable
that in eclipsed N,N-diethylneopentylamine2j,3b the ro-
1
3
There are no corresponding changes in the C spectra
except for doubling of the geminal dimethyl singlet for
tational barrier is 9.4-9.5 kcal/mol. For neopentylpyr-
rolidine 1f with a broad potential energy minimum
described above, the calculated barrier for 180° is dis-
1
c and of the methylenedioxy singlet in 1b. Very similar
behavior is observed for the corresponding N-ethyl com-
pounds 2a -c. In each case, interconversion of structures
such as I and V has become slow, so atoms or groups
attached to ring carbon atoms are either axial or equato-
rial on the NMR time scale at -40 °C. Barriers are in
the range 10.6-11.8 kcal/mol (see Table 2), close in size
to that of 11.8 kcal/mol observed for ring inversion of
N-methylpiperidine,10 so it is reasonable to conclude that
the slowing of ring inversion is responsible for these
changes.
tinctly lower at 8.5 kcal/mol.
Dyn a m ic NMR Sp ectr a . The 1H and 13C NMR
spectra of the compounds studied are temperature de-
pendent, showing splitting of signals as the temperature
is lowered. A full-line shape matching of experimental
q
and calculated spectra yields barriers ∆G for the process
responsible; see Table 2. Some of the spectral changes
are clearer than others and will be described in some
detail below. Tables 3 and 4 in the Experimental Section
1
13
describe H and C spectra for each compound at various
significant temperatures. Results for the conformation-
ally biased 3-methyl-substituted compounds 1e and 2e
will be discussed after those for the other compounds.
It is useful to begin by reporting the results for
N-neopentylpyrrolidine 1f and N-ethylpyrrolidine 2f.
The piperidones 1d and 2d display analogous spectral
changes but at considerably lower temperatures. Ring
inversion is expected to have a much lower barrier than
in the piperidine derivatives since the piperidone ring is
flattened around the carbonyl group. A comparison of
1
1
1
3
ring-inversion barriers for cyclohexanone and cyclohex-
ane12 (4.0 and 10.3 kcal/mol, respectively) dramatically
illustrates this effect. In fact, broadening appears in the
proton NMR as the temperature is lowered below about
There are no changes in the C NMR spectrum, and in
1
the H NMR spectrum only geminal protons of the five-
membered ring become nonequivalent at low tempera-
ture, yielding barriers of about 8.0 kcal/mol in each case.
The only plausible process in molecule 2f is nitrogen
inversion, and it is reasonable to expect that barriers of
a similar size will be obtained for nitrogen inversion in
other members of the series 2.
-50 °C for 1d and below about -70 °C for 2d with
eventual splitting of the signals for geminal protons as
before. Once more, the intercoversion of structures such
as I and V of Scheme 1 has become slow on the NMR
time scale with barriers of 9.3 and 7.6 kcal/mol, respec-
(
8) Bushweller, C. H.; Fleischman, S. H.; Grady, G. L.; McGoff, P.;
Rithner, C. D.; Whalon, M. R.; Brennan, J . G.; Marcantonio, R. P.;
Domingue, R. P. J . Am. Chem. Soc. 1982, 104, 6224.
(10) Lambert, J . B.; Keske, R. G.; Carhart, R. E.; J ovanovich, A. P.
J . Am. Chem. Soc. 1967, 89, 3761.
(
9) (a) Anderson, J . E.; Casarini, D.; Lunazzi, L. J . Chem. Soc.,
(11) Anet, F. A. L.; Chmurny, G. N.; Krane, J . J . Am. Chem. Soc.
1973, 95, 4423.
(12) Anet, F. A. L.; Bourn, A. J . R. J . Am. Chem. Soc. 1967, 89, 760.
Perkin Trans. 2 1991, 1431. (b) Casarini, D.; Davalli, S.; Lunazzi, L.;
Macciantelli, D. J . Org. Chem. 1989, 54, 4616.

Exocyclic N-CH
2
Bond in N-Neopentylpiperidines
J . Org. Chem., Vol. 63, No. 10, 1998 3313
F igu r e 2. Carbon-13 NMR spectra of N-neopentylpiperidine
1
a at ambient (top), and at low temperature, when two sets
of signals are seen for equatorial (97%) and axial (3%)
neopentyl chair conformations. CD Cl solvent appears at δ
4.0.
2
2
5
maximum signal broadening by the method of Anet and
Basus are 7.6 and 8.2 kcal/mol, respectively.
4
F igu r e 1. Proton NMR spectra of 3,3-dimethyl-N-neopen-
tylpiperidine 1c at ambient temperature (top) and at -120°
when geminal groups on the ring and the geminal neopentyl
protons are nonequivalent on the NMR time scale.
The MM3-calculated energies (see Table 1) suggest
that the second set of signals comes from the conforma-
tion III (≡ VII) with the neopentyl group axial-eclipsed
(see Scheme 1). The failure to observe a second confor-
tively, significantly different from each other when set
alongside the similar barriers of 1f and 2f and from the
ring inversions found for 1a -c and 2a -c.
mational process for 1c fits well with “neopentyl axial”
as the second conformation of 1a and 1e, for the axial
methyl group at ring position 3 in 1c disfavors such a
conformation. The agreement between the calculated
axial/equatorial energy difference (about 2 kcal/mol) and
that measured experimentally (about 1 kcal/mol) is only
moderately good. Interconversions of I and VII and of
V and III thus have also become slow on the NMR time
scale in these compounds at -120 °C. This happens
The barrier of 7.6 kcal/mol measured for N-ethylpip-
eridone 2d is of the size found for what is certainly
nitrogen inversion of 2f, and the same process in N-meth-
yl-4-piperidone13 has a barrier of 8.6 kcal/mol, so the
barrier measured for 2d is assigned to nitrogen inversion.
The barrier of 9.3 kcal/mol for N-neopentylpiperidone 1d
is significantly larger than that of 2d and much greater
than expected for either ring inversion or nitrogen
inversion, as discussed above. It agrees satisfactorily
with the calculated and experimental barriers mentioned
above for rotation about the N-neopentyl bond so is
assigned to this rotation.
2
when either nitrogen inversion or N-CH bond rotation
is slow, but the assignment of barriers to one or other
process, rotation or nitrogen inversion, needs careful
consideration.
The barriers of 7.6-8.3 kcal/mol for 1a ,b,e determined
4
using the approximation of Anet and Basus are signifi-
A second dynamic NMR process is seen in the spectra
of 1a ,b,d ,e, but not of 1c and 1f nor for any of the
compounds 2a -f. This second process involves the
appearance of two sets of signals as summarized in
Tables 2 and 3 in the Experimental Section. Changes
are much clearer in the 13C spectra, which in each case
broaden below about -80 °C and then sharpen, appear-
ing at about -120 °C as a major and a minor set of
signals. Figure 2 shows the spectrum for 1a with 3% of
the minor structure, which is thus 1.04 kcal/mol less
stable at -120 °C. For 1b, there is 8% of the minor
structure, which is thus 0.72 kcal/mol less stable at -120
cantly lower than that of 9.3 kcal/mol certainly assigned
to neopentyl group rotation in 1d , yet the immediate
environment of the neopentyl group is identical in all four
cases. On the other hand, there is no reason nitrogen
inversion should have barriers in the range 7.6-8.3 kcal/
mol. The best interpretation is that these barriers
represent hindered rotation but that their magnitude is
somewhat uncertain. We note immediately below for
compound 1d that the population of equatorial and axial
conformations is very temperature dependent. This, and
4
the assumptions of the Anet and Basus approximation,
may be responsible for the 7.6-9.3 kcal/mol range for
neopentyl group rotation barriers that we report.
The second set of changes in the spectrum of the
piperidone 1d occur at about 20 °C lower in temperature,
where the 13C NMR show two sets of spectra at -145 °C
°
C. Barriers to this second process determined from the
(13) Lehn, J . M.; Wagner, J . J . Chem. Soc., Chem. Commun. 1970
,
414.

3
314 J . Org. Chem., Vol. 63, No. 10, 1998
Anderson et al.
Sch em e 2
barrier10 is itself 1.5 kcal/mol higher than that of cyclo-
1
2
hexane.
A further interesting point can be derived from the 13C
chemical shifts in Table 4 (see also Figure 3). Signals
2 2
are referenced to internal CD Cl , and all show small
upfield shifts as the temperature is lowered. The qua-
ternary tert-butyl carbon signal, which does not split,
illustrates this point being at δ 34.0 at room temperature
and δ 33.0 at -145 °C. Correcting for this small intrinsic
temperature dedpendence of shifts, the averaged signals
F igu r e 3. Upfield region of carbon-13 NMR spectra for
N-neopentyl-4-piperidone 1d at ambient temperature (top) and
at -145°, showing two sets of signals of almost equal intensity
for equatorial- and axial-neopentyl chair conformations. Simi-
lar doubling is seen for the carbonyl carbon. Slow rotation of
the tert-butyl group affects the signal at δ 27.0. Chemical shifts
suggest that one of these two conformations substantially
predominates at room temperature.
1
3
13
at room temperature [ C(3,5), 42.2, C(2,6), 56.8, and
1
3
2
CH -t-Bu, 69.4 ] correspond closely to those for the
major conformational isomer signal at low temperature.
The compound apppears to have a normal marked
preference for the equatorial conformation at room tem-
perature, contrasting with the observed 53:47 conforma-
tion ratio at -145 °C and thus reinforcing the oddity of
the latter result.
of relative intensity 53:47 (see Figure 3). This suggests
two kinds of conformation for 1d differing in energy by
only 0.03 kcal/mol at that temperature and interconvert-
ing slowly, with a barrier calculated to be 6.5 kcal/mol.
The analogous N-ethyl compounds 2a -d show only a
single dynamic process in the NMR in which geminal
groups on the ring become nonequivalent on the NMR
time scale (see Table 2). Arguments as used above for
the series 1 suggest that for 2a -c ring inversion is the
process involved, whereas for compound 2d , discussed
above, the process is nitrogen inversion. The magnitudes
of the barriers fit well with these assignments. The
absence of a very low temperature set of spectral changes
for 2a -d probably means that the ethyl-axial conforma-
tion of 2 is unpopulated since by calculation it is
relatively less stable than the neopentyl-axial conforma-
tion of 1. This agrees with the high axial/equatorial
energy difference of 2.7 kcal/mol reported for N-meth-
ylpiperidine.15
This experimental energy difference of 0.03 kcal/mol
for 1d is so small that it would be unwise to rely on
calculations for deciding which is the major conformation,
although they admit no doubt that the two conformations
in question are equatorial eclipsed and axial eclipsed.
Since the signal of a carbon atom antiperiplanar to a lone
pair appears upfield from that of a gauche carbon atom,14
the relative upfield shift of the minor signal for carbons
3
and 5 of the ring fits with the minor conformation being
the one with an axial eclipsed neopentyl group. Calcula-
tions suggest that this conformation is less stable than
the eclipsed equatorial by 1.98 (MM3) or 2.51 (MMX)
kcal/mol, much larger than experimental values, repre-
senting even poorer agreement with experiment for 1d
than was reported above for 1a and 1b.
For the 3-methylpiperidines 1e and 2e, the conforma-
tional cube is distorted. For 1e, the equatorial,equatorial
conformation IX is calculated to be the most stable, with
the 3-methyl-axial conformation X and the neopentyl-
axial conformation XI being 1.15 and 2.10 kcal/mol less
stable, respectively; see Scheme 2, noting that a set of
Here, inspection of Scheme 1 suggests that these
changes cannot be due to slowing down of nitrogen
inversion because the populated conformations, eclipsed-
equatorial and eclipsed-axial, can still interconvert by
ring inversion. The anti-conformations II, IV, VI, and
VIII are calculated to be so high in energy as to be
unpopulated. This implies that the second process slow-
ing down in 1d , to produce two sets of signals, is ring
inversion (e.g., I-III), with a barrier of 6.5 kcal/mol at
enantiomeric drawings is required for the enantiomer of
13
1
e. In the proton-decoupled C NMR spectrum of 1e as
the temperature is lowered, there is a broadening of some
-
120 °C. This barrier is significantly higher than in
(
15) (a) Delpuech, J .-J . In Cyclic Organonitrogen Stereodynamics;
Lambert, J . B., Takeuchi, Y., Eds.; VCH: New York, 1992; Chapter 6.
b) Crowley, P. J .; Robinson, M. J , T.; Ward, M. G. J . Chem. Soc. Chem.
cyclohexanone, but the N-methylpiperidine ring inversion
(
Commun. 1974, 825. (c) Crowley, P. J .; Robinson, M. J . T.; Ward, M.
G. Tetrahedron 1977, 33, 915. (d) Appleton, D. C.; McKenna, J .;
McKenna, J . M. J . Am. Chem. Soc. 1976, 98, 292.
(
14) Eliel, E. L.; Pietrusiewicz, K. M. Top. 13C NMR Spectrosc. 1979,
, 171.
3

Exocyclic N-CH
2
Bond in N-Neopentylpiperidines
J . Org. Chem., Vol. 63, No. 10, 1998 3315
signals with a maximum breadth at about -45 °C and
then a sharpening. On further cooling, before a minor
set of signals can be detected, there is a new broadening
of a different set of signals with a maximum at about
the most stable conformation is the equatorial eclipsed
one shown in the conformational cube of Scheme 1 as
structures I and V and that the next most stable, except
for 1c, is the axial eclipsed one III and VII.
-
85 °C and then sharpening of signals until, at -105
The results for 1a -c with barriers over 10 kcal/mol
undoubtedly represent ring inversion, and the changes
in the NMR spectrum of the 3-methyl compound 1e are
assigned to ring inversion as well, since for reasons
already explained the somewhat low barrier of 9.6 kcal/
mol is only an estimate.
The only clearly identified barrier to rotation about the
piperidine nitrogen-neopentyl bond is that of 9.3 kcal/
mol for the N-neopentylpiperidone 1d . This is justified
as it is markedly higher than nitrogen inversion barriers
in similar compounds and the expected ring inversion
barrier. When nitrogen inversion subsequently becomes
slow, there should be no effect on the spectrum since the
two populated structures can interconvert by ring inver-
sion. Thus, no barrier of about 8 kcal/mol should be or
is in fact observed. The 6.5 kcal/mol barrier is then
assigned to ring inversion becoming slow.
°
C, two sets of signals (relative intensity 98:2) are seen
for some carbon positions (see Table 4).
Two processes become slow as the temperature is
lowered but only one minor set of signals of 2% relative
intensity is observed. Results from other neopentyl
piperidines suggest that the first process to slow is ring
inversion, so the minor conformation whose signals
should then emerge is X, the only low-energy conforma-
tion with the methyl group axial. The interconversion
of IX and XI is the second process becoming slow, and
as before, this is seen when either neopentyl group
rotation or nitrogen inversion becomes slow. The barriers
to these two processes are determined by the Anet and
4
Basus method to be about 9.6 and 8.0 kcal/mol, respec-
tively, at the temperature of maximum broadening.
1
H NMR shows a less well-defined set of changes, but
in spectra taken at -100 °C minor signals are seen
adjacent to the major signals of some groups, and in
particular, there is a minor isomer doublet for the methyl
Calculations suggest, however, that the barrier to
rotation is about 10 kcal/mol in all neopentylpiperidines.
In the NMR of 1a and 1b, as the temperature is lowered,
after ring inversion is slow, a second process might be
expected at intermediate temperatures where intercon-
version of structures I and III is slow due to hindered
rotation. We believe that in 1a and 1b (and also perhaps
in 1e), with very small amounts of the minor conforma-
tion, the overlap of the dynamic changes in the spectrum
with temperature and the uncertainties of the Anet and
Basus approximation make detection of the minor isomer
difficult until both processes are slow. The clear dem-
onstration of bond rotation in 1d emerges since it appears
in the NMR as geminal protons becoming nonequivalent.
It is interesting to consider the suite of changes
experienced by a molecule of type 1. Both rotation and
nitrogen inversion lead to anti conformations II and IV
at least 3.6 kcal/mol higher in energy than eclipsed by
calculation, so further very rapid changes are expected.
Molecules are likely either to return to the original
conformation by reversal of the first process or to move
on to give the axial eclipsed conformation by nitrogen
inversion or rotation, respectively. Overall, this two-step
passage through an unstable intermediate not signifi-
cantly populated is a single process as far as changes in
the NMR are concerned.
group at δ 0.83: when compared with the downfield 13
C
sideband of the tert-butyl signal of the major isomer this
allows a direct estimate of population of the minor isomer
(about 2.0%). The fact that the minor methyl doublet
signal is downfield from the major does not correspond
to a methyl axial location and so favors XI as the second
populated conformation. The coupling constants in these
two equatorial methyl doublets are markedly different,
namely 7.2 Hz in the major and 5.6 Hz in the minor
conformation.
The assignment of structure XI to the 2% conformation
agrees with the observations for other neopentylpip-
eridines and is further confirmed by the relative upfield
shift of the minor C3 signal, suggesting that C3 is
antiperiplanar to a nitrogen lone pair.14 Although cal-
culations suggest that structure X, not XI, is the more
stable, they have been shown to underestimate the
population of axial-N-neopentyl conformations in other
compounds of the series 1. The absence of a second minor
spectrum that we would assign to structure X probably
reflects the experimental difficulty in detecting popula-
tions less than about 1% and, possibly, a temperature-
dependent population. Derivative 2e does not display
any line broadening attributable to conformational pro-
cesses, thus suggesting that essentially only a single
conformation is populated.
Chair-chair interconversion of the six-membered rings
of type 1 also involves unstable intermediate twist
conformations, so whatever is the first process taking
place in Scheme 1, there is a significant likelihood that
the reverse process take place. This means that the rate
of crossing the first barrier may be significantly higher
than indicated by dynamic NMR measurements, so the
barriers reported in Table 2 should be reduced by up to
Discu ssion
The series of barriers measured for the compounds 1a -
f, and 2a -f are shown together in Table 2 along with
assignment to the processes involved. The pyrrolidine
0
.3 kcal/mol. In other words, a transmission coefficient
2
f provides a good model for nitrogen inversion.
less than 1 in value should be used in converting a rate
of site exchange derived from the NMR into a potential
barrier.
In the N-ethylpiperidines 2, calculations agree that
only equatorial staggered conformations deserve consid-
eration. Only one dynamic NMR process is ever observed
experimentally for each compound and is assigned on the
basis of barrier size to ring inversion (2a -c) or nitrogen
inversion (2d ,f), so little further discussion of the assign-
ments of these is needed.
Combination of two of the processes into a single
process, different from two processes taking place one
after the other, may occur in some symmetrical staggered
16
amines. Since it is likely to combine the greater part
Following the calculations and previous results,2i,3
there is no doubt that in all the N-neopentylpiperidines
(16) Bushweller, C. H.; Anderson, W. G. Tetrahedron Lett. 1972, 129
and references therein.

3
316 J . Org. Chem., Vol. 63, No. 10, 1998
Anderson et al.
1
Ta ble 3.
H Ch em ica l Sh ifts (400 MHz) of Com p ou n d s 1A-1f a n d 2A-2f a t Va r iou s Tem p er a tu r es (p p m fr om Me4Si in
CCl2D2/CHF 2Cl/CHClF 2)
observed signals
compd
T (°C)
+25
NCH2-t-Bu
C(CH3)3
CH2(2,6)
CH2(3,5)
CH2(4)
1
1
1
a
b
c
1.89
1.88
0.76
0.74
2.34
2.67
1.42
1.41
1.28
1.55
-
130
2
.01
0.98
3.82 [OCH2]
3.80
+25
2.08
2.06
0.89
0.87
2.61
2.69
1.68
1.68
1.51
1.45
-120
2
.39
2.14
.27
1.92
+25
1.85
1.83
0.75
0.72
1.04
0.84 [C(CH3)2]
2
-120
1.27
1.65
1.19
0.89
0.72
0.99
2
2
2
.25
.47
.65
1
1
1
d
e
f
+25
2.11
2.09
0.82
0.80
2.28
2.40
2.75
2.91
2.61
-
120
2
.22
-
-30
1.98
1.97
0.89
0.80
2.73
2.72
(1.50-1.83)
(1.41-1.75)
0.82 [ HCCH3]
-120
0.75 (98%)
.06 (2%)
1
+25
2.30
2.25
0.89
0.89
2.64
3.01
1.78
1.72
-120
2
.98
observed signals
compd
T (°C)
+25
NCH2CH3
CH3
CH2(2,6)
CH2(3,5)
CH2(4)
2
2
2
a
b
c
2.22
2.21
0.93
0.91
2.24
2.84
1.46
1.40
1.38
1.75
1.75
1.60
1.49
1.33
1.06
1.63
-
120
1
.54
+25
2.45
2.30
1.10
1.09
2.52
2.81
3.94 [OCH2]
3.85
-120
1
.94
2.21
.91
2.75
+25
2.20
2.22
0.93
0.91
1.13
0.84[C(CH3)2]
1
-120
1.47
1.39
0.91
0.82
0.71
2
1
1
.41
.55
.24
2
d
+25
2.42
2.45
1.03
1.08
2.32
2.40
2.62
3.06
2.16
-
120
2
.02
2
2
e
f
-60
+25
2.29
2.48
2.28
1.10
0.98
0.98
2.86
2.40
2.88
(1.5-1.7)
0.78 [HCCH3]
1.65
1.60
-
120
1
.68
of the activation energy for each process, such a combina-
tion is improbable in the present molecules. This does
not preclude one process, say nitrogen inversion flatten-
ing, taking place to some small extent so as to reduce
somewhat the activation energy of another, say bond
rotation. This could be represented by joining vertexes
of the cube of Scheme 1 by lines curved inward rather
than straight. A combined process is more plausible
when the intermediate minimum is of high relative
energy, existing as a shallow depression on the potential
surface.
as described in detail below. Piperidones were prepared from
the correponding ethylenedioxy ketal 1b or 2b by acid-
catalyzed hydrolysis. Pyrrolidine and piperidine and its
3
-methyl, its 3,3-dimethyl, and its 4-ethylenedioxy derivatives
were commercially available (Aldrich).
P r ep a r a tion of Acyla ted Am in es. Pyrrolidine, or the
appropriately substituted piperidine (95 mmol), and pyridine
(7.7 mL, 95 mmol) were dissolved in chloroform (75 mL), and
the solution was cooled to 0 °C, acyl chloride (95 mmol) was
added dropwise. After 18 h of reflux, a white precipitate had
formed. The reaction mixture was allowed to cool to room
temperature, and 50 mL of water was added. The reaction
mixture was extracted with ether (3 × 50 mL), and the ether
extracts were dried with anhydrous magnesium sulfate. The
solvent was distilled off under reduced pressure, and the
product, typically produced in about 85% yield, was used
directly for the reduction step.
Exp er im en ta l Section
Tables 3 and 4 show details of NMR spectra at a range of
temperatures. Molecular mechanics calculations using Al-
5
,6
linger’s MM3 program yield torsion angles between the tert-
butyl (or methyl) group and the N-C2 and N-C6 bonds. The
torsion angle with the lone pair is derived from these by
assuming that the lone pair is located on the external bisector
of the C2-N-C6 angle in the projection.
Red u ction of Acyla ted Heter ocycles. The product from
the acylation reaction was added under nitrogen to a suspen-
sion of lithium aluminum hydride (4.7 g, 122 mmol) in diethyl
ether (150 mL) cooled to 0 °C. After 18 h of stirring, 20 mL
water was added dropwise until effervescence subsided. The
organic layer was extracted with ether (3 × 50 mL), and the
ether extracts were dried and concentrated as before to yield
the crude products 1 or 2, which were purified by distillation
under reduced pressure.
Gen er a l Syn th etic Meth od . All N-neopentyl- and N-ethyl
compounds, with the exception of the piperidones 1d and 2d ,
were prepared from the corresponding piperidine or pyrrolidine
by acylation followed by lithium aluminum hydride reduction

Exocyclic N-CH
2
Bond in N-Neopentylpiperidines
J . Org. Chem., Vol. 63, No. 10, 1998 3317
1
3
Ta ble 4.
C Ch em ica l Sh ifts (100.6 MHz) of Com p ou n d s 1a -f a n d 2a -f a t Va r iou s Tem p er a tu r es (p p m fr om Me4Si in
CCl2D2/CHF 2Cl/CHClF 2)
observed signals
compd
T (°C)
NCH2-t-Bu
C(CH3)3
C(CH3)3
C(2,6)
C(3,5)
C(4)
1
1
1
a
b
c
+25
-120
-120
+20
-120
-120
+25
72.0
70.5
63.0
69.5
72.0
65.2
71.5
34.1
32.1
33.0
32.4
33.2
33.4
34.5
28.2
27.0
a
27.1
26.8
c
58.2
57.0
53.4
53.9
56.0
53.0
58.6
27.4
25.5
19.9
35.1
35.5
29.6
24.5
32.9
24.2
32.4
42.2
42.0
38.5
26.3
32.0
26.4
31.9
29.1
23.0
23.0
25.0
22.8
24.1
107
106.5
c
9
3
7%
%
64.0 [OCH2]
66.8, 66.6
c
9
8
2%
%
28.7
38.4
28.0[C(CH3)2]
7
0.3
58.6
0.1
-80
71.0
34.2
34.0
28.1
38.2
32.4
25.0
7
1
1
d
e
+25
-145
-145
-10
69.4
67.8
62.5
70.8
28.0
27.5
27.0
26.3
56.8
55.2
53.1
65.5
212.0
213.8
213.3
33.0
b
5
4
3%
7%
33.0
33.0b
33.3
193 [HCCH3]
5
7.1
65.4
6.9
9
2
8%
%
-100
70.8
33.2
26.4
32.5
19.7
5
-100
+25
19.1
28.2
8.53
1
f
69.0
69.2
32.0
32.1
56.7
56.8
-120
observed signals
compd
T (°C)
+25
NCH2CH3
CH3
C(2,6)
C(3,5)
C(4)
2
2
2
a
b
c
53.8
52.8
52.8
51.9
53.8
12.5
12.0
12.9
12.4
12.8
55.0
53.7
52.0
50.9
27.0
25.2
35.9
34.4
38.7
25.5
24.0
108.2
107.2
31.8
-
130
+25
64.5
64.4
23.5
-
120
+25
64.2
[C(CH3)2]
66.9 55.5
2
8.2
-
120
52.7
12.0
64.5
36.8
30.1
41.9
41.5
62.9
30.5
22.1
5
5.0
2
2
2
d
e
f
+25
52.0
51.8
53.8
13.0
13.0
12.4
53.4
53.8
210.5
212.0
32.3
-
120
-30
34.0
20.2 [HCCH3]
5
4.6
26.5
+25
50.1
50.0
13.2
14.0
53.5
54.0
23.0
23.0
-
120
a
Signala for the tert-butyl methyl group of the minor isomer could not be distinguished and are presumably overlapped by the major
b
13
isomer signal. The C NMR at -140 °C. This signal is unusually broad at the lowest temperature observed in a manner quite in
c
1
keeping with rotation about the t-Bu-CH2N bond becoming slow on the NMR time scale. Signals for t-BuMe group, C4 and C(3,4) for the
minor isomer could not be distinguished and are presumably overlapped by the major isomer signal.
Hyd r olysis of P ip er id on e Keta ls. The corresponding
ketal (5 g) was added to a mixture of 5% HCl solution (60 mL)
and acetone (10 mL) and refluxed for 12 h. After cooling, 1 M
sodium hydroxide solution was added until the reaction
mixture was basic. Extraction with diethyl ether and workup
as above led to a crude product that was purified by column
chromatography on silica using 20:1 pentane/ethyl acetate as
eluant.
N-Ethylpiperidine (2a ) and N-ethylpiperidone (2d ) are
known compounds.1
7,18
4
-(Eth ylen ed ioxy)-N-eth ylp ip er id in e (2b): bp 90-92 °C/
.3 mmHg. Anal. Calcd for C : C, 63.18; H, 9.88; N,
.19. Found: C, 62.52; H, 10.24; N, 8.21.
0
8
9 2
H17NO
3
,3-Dim eth yl-N-eth ylpiper idin e (2c): bp 80-82 °C. Anal.
Calcd for C 19N: C, 76.61; H, 13.52; N, 9.87. Found: C,
6.25; H, 13.89; N, 9.82.
9
H
N-Neop en tylp ip er id in e (1a ): bp 70-72 °C/0.6 mmHg.
7
Anal. Calcd for C10
C, 77.04; H, 13.94; N, 8.93.
-(Eth ylen ed ioxy)-N-n eop en tylp ip er id in e (1b): mp 45
C. Anal. Calcd for C12 : C, 67.61; H, 10.84; N, 6.61.
Found: C, 67.56; H, 11.03; N, 6.56.
,3-Dim eth yl-N-n eop en tylp ip er id in e (1c): bp 88-90 °C/
.45 mmHg. Anal. Calcd for C12 25N: C, 78.68; H, 13.70; N,
H21N: C, 77.41; H, 13.55; N, 9.03. Found:
3
-Meth yl-N-eth ylp ip er id in e (2e): bp 139-140 °C/760
mmHg. Anal. Calcd for C 16N: C, 76.13; H, 12.78; N, 11.10.
8
H
4
Found: C, 76.28; H, 12.70; N, 11.20.
°
H23NO
2
N-Eth ylp yr r olid in e (2f): bp 100-102 °C. Anal. Calcd
3
for C H13N: C, 73.40; H, 12.78; N, 13.82. Found: C, 72.72;
6
0
7
H
H, 13.13; N, 14.14.
.61. Found: C, 78.23; H. 14.17; N, 7.50.
N-Neop en tyl-4-p ip er id on e (1d ): bp 97-99 °C/24 mmHg.
Ack n ow led gm en t. We are grateful to a reviewer for
helpful comments.
Anal. Calcd for C10
Found: C, 70.59; H, 11.52; N, 8.10.
-Meth yl-N-n eop en tylp ip er id in e (1e): bp 80-82 °C/26
mmHg. Anal. Calcd for C11 22N: C, 78.50; H, 13.18; N, 8.82.
H19NO: C, 71.01; H, 11.24; N, 8.28.
3
J O9720344
H
Found: C, 78.47; H, 13.25; N, 8.40.
N-Neop en tylp yr r olid in e (1f): bp 108-110 °C. Anal.
(
(
17) Steward, J . M. J . Am. Chem. Soc. 1954, 76, 7111.
18) Fuson, R. C.; Parham, W. E.; Reed, L. J . J . Am. Chem. Soc.
Calcd for C
9
H19N: C, 76.60; H, 13.48; N, 9.92. Found: C,
7
6.55; H, 13.27; N, 9.68.
1946, 68, 1239.