Conformational studies by dynamic NMR. 91. Conformational stereodynamics of tetraethylmethane and analogous C(CH2X)4 compounds

Article abstract of DOI:10.1021/jo025984k

Variable-temperature NMR studies of tetraethylmethane (1a), tetrapropylmethane (1b), tetrachloromethylmethane (1c), tetrabromomethylmethane (1d), tetracyclopropylmethylmethane (1e), and tetrabenzylmethane (1f) show a range of dynamic behavior. Separate signals for two types of conformation are observed for 1a, 1c, and 1d at low temperatures, with more than 95% of the molecules in a time-averaged D_2d conformation, and the S₄ conformation as the minor populated alternative. Compound 1e populates only S₄-type conformations but equilibrates slowly between degenerate versions of these at low temperatures. Compounds 1b and 1f show a temperature-dependent spectrum but the low-temperature limit spectrum could not be observed. Ab initio calculations agree well with experiment on the conformational equilibria and suggest in particular that compounds 1b and 1f behave similarly to compounds 1a and 1e, respectively. A crystal structure of compound 1f is reported.

Full text of DOI:10.1021/jo025984k

Con for m a tion a l Stu d ies by Dyn a m ic NMR. 91.1 Con for m a tion a l
Ster eod yn a m ics of Tetr a eth ylm eth a n e a n d An a logou s C(CH2X)4
Com p ou n d s
Stefano Grilli, Lodovico Lunazzi,* Andrea Mazzanti, and Marco Pinamonti
Department of Organic Chemistry “A. Mangini”, University of Bologna, Risorgimento, 4,
Bologna 40136, Italy
J . Edgar Anderson*
Chemistry Department, University College, Gower St., London, WC1E 6BT, UK
C. V. Ramana, Priti S. Koranne, and Mukund K. Gurjar
National Chemical Laboratory, Pune-411008, India
lunazzi@ms.fci.unibo.it
Received May 29, 2002
Variable-temperature NMR studies of tetraethylmethane (1a ), tetrapropylmethane (1b), tetra-
chloromethylmethane (1c), tetrabromomethylmethane (1d ), tetracyclopropylmethylmethane (1e),
and tetrabenzylmethane (1f) show a range of dynamic behavior. Separate signals for two types of
conformation are observed for 1a , 1c, and 1d at low temperatures, with more than 95% of the
molecules in a time-averaged D2d conformation, and the S4 conformation as the minor populated
alternative. Compound 1e populates only S4-type conformations but equilibrates slowly between
degenerate versions of these at low temperatures. Compounds 1b and 1f show a temperature-
dependent spectrum but the low-temperature limit spectrum could not be observed. Ab initio
calculations agree well with experiment on the conformational equilibria and suggest in particular
that compounds 1b and 1f behave similarly to compounds 1a and 1e, respectively. A crystal
structure of compound 1f is reported.
In tr od u ction
least one parallel 1,3-interaction between X-groups
(Scheme 1).
The conformational analysis of a quaternary carbon
center with four radiating primary alkyl chains is
potentially complex since each C-CH2 bond, at least in
acyclic molecules which are the subject of this paper, has
three staggered conformations, so there are 34 ) 81
possible conformations. Fortunately with four identical
alkyl groups, i.e., C(CH2 X)4 1, symmetry reduces the 81
possible structures to 6 types with various levels of
degeneracy. In important seminal work in this series for
1a (X ) Me) and 1b (X ) Et), Alder and his collaborators
have shown2-4 that two of these six types of conformation
are of very similar stability and are much more stable
than the other four. The relative instability of these latter
four conformational types results from their having at
An electron diffraction study showed that for 1a , both
of the stable conformations are populated in the gas
phase.4 Calculations2,3 confirm this observation and sug-
gest a similar situation for 1b. Crystal structure deter-
mination of a large range of analogous compounds cen-
tered on a quaternary nitrogen atom, i.e., N+(CH2R)4 Y-,
show many examples of these two stable conformational
types.3
Scheme 1 shows the six conformational types in ideal-
ized form, and how they are related by successive 120°
rotations of an appropriate CH2X group. The two most
stable conformations are labeled T1 and T3 since they
are separated by ∼120° rotations of tw o CH2X groups,
and T2 is intermediate between these two. The set of
conformations will be considered further in the Discus-
sion.
(1) Part 90: Grilli, S.; Lunazzi, L.; Mazzanti, A.; Pinamonti, M. J .
Org. Chem. 2002, 67, 5733. See also Part 89: Gasparrini, F.; Grilli,
S.; Leardini, R.; Lunazzi, L.; Mazzanti, A.; Nanni, D.; Pierini, M.;
Pinamonti, M. J . Org. Chem. 2002, 67, 3089.
We now want to report the dynamics of several
molecules 1a -f of this general type, using calculations
(2) Alder, R. W.; Maunder, C. M.; Orpen, A. G. Tetrahedron Lett.
1990, 31, 671.
(3) Alder, R. W.; Allen, P. R.; Anderson, K. R.; Butts, C. P.; Khosravi,
E.; Martin, A.; Maunder, C. M.; Orpen, A. G.; St. Pourc¸ain, C. B. J .
Chem. Soc., Perkin Trans. 2 1998, 2083.
(4) Alder; R. W.; Allen, P. R.; Hnyk, D.; Rankin, D. W. H.; Robertson,
H. E.; Smart, B. A.; Gillespie, R. J .; Byetheway, I. J . Org. Chem. 1999,
64, 4226.
and dynamic NMR spectroscopic observations to show
what conformations are populated, how conformations
10.1021/jo025984k CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/15/2002
J . Org. Chem. 2002, 67, 6387-6394
6387

Grilli et al.
SCHEME 1. On e Ver sion Ea ch of th e Six Con for m a tion a l Typ es (T1-T6) of C(CH2X)4 Der iva tives (1a -1f)a
a
Types actually populated are T1 and T3. The versions shown for types T1 and T3-T6 can each be converted to T2 by rotation of the
one CH2X group indicated. Structures are drawn with the labels for hydrogen atoms omitted and with idealized staggered conformations
although minimum energy forms may have torsion angles skewed. Each type is uniquely defined by two criteria: (i) the number of
anti,anti arrangements (a,a) of X-C-Cq-C-X chains that can be identified, thus T1 has two of these, XA-C-Cq-C-XB and XC-C-
Cq-C-XD, and (ii) the number of 1,3-parallel arrangements of substituents (X,X), thus T6 has one of these, XA with XC. For each type,
the number of degenerate/enantiomeric versions that exist is indicated.
interconvert, and what the barrier to these interconver-
sion processes are.
by a resharpening as in Figure 1. Such an effect is not
noticeable for the Me signal. The Cq signal was too weak
for quantitative determinations, but the CH2 signal
shows maximum incremental broadening of 10 Hz (i.e.
13-3 Hz) at -138 °C. From the relationship5,6 k ) 2π-
(10), the rate constant for the conformational process is
63 s-1, hence ∆Gq ) 6.6 kcal mol-1 (Table 1). At -155 °C
a CH2 signal due to the minor conformer was detected,
166 Hz (1.65 ppm) to lower field of the major one, with
∼3% relative intensity. By allowing for the degeneracy
of types T3 and T1 (see Scheme 1) and assuming ∆S° )
0 this proportion should become 5% at -138 °C. Using
this value, the computer line shape simulation provides,
at -138 °C, ∆Gq ) 6.6 kcal mol-1, in agreement with the
approximate formula above.
Resu lts
All compounds showed relatively simple NMR spectra
at room temperature with dynamic broadening of signals
at considerably lower temperatures. For compounds 1a -
d , signals sharpened at even lower temperatures and,
except for 1b, two sets of signals of different intensity
were seen. For 1e resharpening at lowest temperatures
produced 1:1 doubling of the three 1H signals arising from
methylene protons, but for 1f, which by calculation is
conformationally similar, resharpening of signals was not
observed at the lowest temperatures attainable. Because
of solubility problems at very low temperatures, useful
13C NMR data were obtained only for compounds 1a and
1b, so 13C NMR of 1c-f will not be mentioned. Each
compound will now be discussed in turn.
1
The 400 MHz H NMR spectrum also shows incremen-
tal broadening of 3.3 Hz at -142 °C for the CH2 quartet
whereas the CH3 triplet is essentially unaffected. This
corresponds, as above,5,6 to k ) 21 s-1 and thence to
Tetr a eth ylm eth a n e (C(CH2Me)4, 1a ). The 13C NMR
spectrum of Et4C features three signals at ambient
temperature. On progressively lowering the temperature,
broadening of the Cq and CH2 signals occurs, followed
(5) (a) Anet, F. A. L.; Basus, V. J . J . Magn. Reson. 1978, 32, 339.
(b) Sandstro¨m, J . Dynamic NMR Spectroscopy; Academic Press:
London, New York, 1982; p 84.
(6) Grilli, S.; Lunazzi, L.; Mazzanti, A. J . Org. Chem. 2001, 66, 4444.
6388 J . Org. Chem., Vol. 67, No. 18, 2002

Conformational Studies by Dynamic NMR. 91
reotopic CH2 protons provide the AB part of an ABX3
spectrum, too complex to be resolved at low temperature.
The ∆G° derived from the 3% population for T3 at
about -155 °C is 0.97 kcal mol-1, which is not far from
that obtained at 293 K by electron diffraction4 (0.79 kcal
mol-1). The former value suggests that if ∆S° ) 0, the
population ratio in solution at ambient temperature
should be 73:27, which compares satisfactorily with the
66:34 ratio derived by electron diffraction measurements
in the gas phase.4 The NMR barrier of 6.6 kcal mol-1 for
the T1 to T3 interconversion agrees well with the MM2
computed value (6.6 kcal mol-1) of Alder and co-workers.3
Tetr a -n -p r op ylm eth a n e (C(CH2Et)4, 1b). In a simi-
lar way, all signals (except that of Me) in the C-13
spectrum of 1b show dynamic broadening as the tem-
perature is lowered, followed by a resharpening. It was
only in the case of the downfield CH2 signal (i.e. that next
to the quaternary carbon) that the effect was large
enough to allow reasonably accurate measurements. A
maximum incremental broadening of 24 Hz (i.e. 31-7 Hz)
was observed at -138 °C, corresponding to a ∆Gq value
of 6.35 kcal mol-1, for the major to minor interconversion,
a value only slightly lower than the 6.6 kcal mol-1 barrier
for 1a .
At -155 °C all the lines are equally sharp, but the
solubility had decreased dramatically, so that a minor
conformer with a proportion lower than 10-15% could
not be observed above the noise. Assuming a conceivable
shift difference for conformers in the range 1.5-2.50 ppm,
(i.e. a range that includes the 1.65 ppm separation found
above for 1a ), the experimental maximum line broaden-
ing suggests a proportion of the minor conformer of
12 ( 3% at -138 °C. Corrected to -155 °C with allowance
for the greater degeneracy of T3, there should be about
8% of this latter conformation, a proportion too small to
be detected in the experimental conditions we have
described, but significantly greater than the 3% observed
at the same temperature for tetraethylmethane (1a ). This
agrees with our ab initio computations (see Table 2)
predicting energy differences between these conformers
of 0.70 and 0.75 kcal mol-1 for 1b and 1a , respectively
Tetr a ch lor om eth ylm eth a n e (C(CH2Cl)4, 1c). The
400-MHz 1H NMR spectrum of C(CH2Cl)4 comprises two
signals at -158 °C as shown in Figure 2, which allows a
confident assignment of the conformational equilibrium.
The major signal (96%) is a single line (3.58 ppm), as
expected for the T1 conformer, while the minor signal
(4%) is an AB-multiplet (δ 3.92 and 3.77, J ) -11 Hz)
characteristic of the T3 conformer. Allowing for degen-
eracy, the free energy difference between conformers is
0.88 kcal mol-1 at -158 °C. Spectral simulations carried
out at -146° and -138 °C indicate an interconversion
barrier of the major into the minor conformer of 6.9 (
F IGURE 1. Methyl and methylene 13C NMR (100.6 MHz)
lines of tetraethylmethane (1a ) in CHF2Cl/CHFCl2 as function
of temperature. The inset at -155 °C displays the 15-fold
amplified CH2 line of the minor conformer.
TABLE 1. In ter con ver sion Ba r r ier s a s Deter m in ed by
NMR for C(CH2X)4, 1a -1f
1a
1b
Et
1c
1d
1e
cyPr Ph
4.3
1f
X
Me
Cl
Br
∆Gq a 6.6 (6.6)b 6.35 6.9 (10.0)c 8.0 (10.4)c 5.8
a
Experimental free energies of activation (kcal mol-1) of the
interconversion of T1 (D2d symmetry) to T3 (S4 symmetry) for
compounds 1a -1d and of the degenerate T3 interconversion for
b
1e and 1f. Computed barrier by the MM2 force field as in ref 3.
c Computed barrier by the MM3 force field assuming successive
CH2X rotations (present work).11
0.2 kcal mol-1
.
Tetr a br om om eth ylm eth a n e (C(CH2Br )4, 1d ). The
400-MHz 1H NMR spectrum at -150 °C comprises a
single line for the T1 conformer (≈98.5%) and an (≈1.5%)
AB-multiplet (δν ) 16 Hz, J ∼ -11 Hz) for the T3
conformer. Of the four lines expected for the latter
spectrum, three are clearly visible, whereas the fourth
is overlapped by the downfield 13C satellite (J C,H ) 154
Hz) of the major T1 conformer singlet line. There is a
free energy difference of 1.02 kcal mol-1 between any one
T1 and T3 conformer. The barrier, as obtained by line
∆Gq ) 6.65 kcal mol-1 at -142 °C, in agreement with
the 13C experiments. At -157 °C the minor species signal
(3% relative intensity) is seen 60 Hz (0.15 ppm) downfield
from that for the major species. While the major CH2
signal is seen as a quartet (J ) 7.2 Hz and line width
6.5 Hz) at -157 °C, the minor signal is a broad line 30
Hz wide. This much broader minor signal is compatible
with the T3 structure of S4 symmetry, whose diaste-
J . Org. Chem, Vol. 67, No. 18, 2002 6389

Grilli et al.
TABLE 2. Rela tive Com p u ted En er gies (k ca l m ol-1) for th e Tw o Most Sta ble Con for m er s of C(CH 2X)4, 1a -f
1a (X ) Me) 1b (X ) Et) 1c (X ) Cl) 1d (X ) Br) 1e (X ) cyPr)
1f (X ) Ph)a
ab initio MM3 ab initio MM3
ab initio
MM3
ab initio
MM3
ab initio
MM3
ab initio
MM3
T1
T3
0.0
0.75
0.0
0.20
0.0
0.70
0.0
0.29
0.0
0.3
0.31
0.0
0.0
0.63
0.18
0.0
1.34
0.0
0.77
0.0
3.01
0.0
3.30
0.25
a
In this case both ab initio and MM3 calculations predict that the two most stable conformers are T3 and T6 (C2 symmetry as in
Scheme 1), with T1 being only the third (ab initio) or the fourth (MM3) most stable. Ab initio calculations found T3 to be the most stable,
with T6 having a 1.4 kcal mol-1 higher energy (an opposite trend is indicated by MM3 calculations, which predict T6 to be the most
stable, with T3 having a 0.25 kcal mol-1 higher energy).
SCHEME 2
further) predict that the T3-type conformer of Scheme 1
is more stable than all the other five possible conformers
of 1e, and the corresponding computed structure reported
in Figure 4 appears almost identical with the experi-
mental one.
The anisochronous signals for the exocyclic methylene
hydrogens observed at δ 0.92 and 1.84 (Figure 3, trace
at -161 °C) are too broadened to show the fine structure
expected for an ABX-type spectrum. However, the upfield
line is broader than its downfield companion, the half-
height line widths being 36 and 32 Hz, respectively. This
is a consequence of two different 3J vicinal couplings
F IGURE 2. 1H NMR (400 MHz) spectrum of tetrachlorometh-
ylmethane (1c) in CHF2Cl/CHFCl2 at -158 °C, showing signals
for type T1 (D2d symmetry) and type T3 (S4 symmetry)
conformations, i.e., a singlet and an AB multiplet, respectively
(the latter appears in the inset with a 20-fold amplification).
between the methine and diastereotopic methylene pro-
tons. Both experimental (X-ray) and computed (ab initio)
structures indicate that the H-C-C-H dihedral angles
of the CH2-CH moiety (see Scheme 2) are +68° (HA-
3
C-C-HC) and -176° (HB-C-C-HC), so the J vicinal
couplings are expected to be respectively 2.1 and 10.4 Hz
shape simulation at -138 °C, is about 8 kcal mol-1
,
according to a Karplus-type equation.8
higher than in 1a -c in agreement with the prediction
of the MM3 calculations. The differences between the
values for the bromine and chlorine derivatives (1c and
1d , respectively) are a consequence of the larger steric
effects of bromine with respect to chlorine.
Computer simulation with these values and assuming
a geminal J AB ) -12 Hz, with an intrinsic width of about
20 Hz, matched satisfactorily the two experimental line
widths (Figure 3, trace at -161 °C). Accordingly, the
upfield signal corresponds to that of the two methylene
hydrogens having the larger dihedral angle with the CH
hydrogen of the ring. Line shape simulations at various
temperatures (Figure 3) provided a set of rate constants
for the CH2-Cq rotation process, yielding a free energy
Tetr a cyclop r op ylm eth ylm eth a n e
(C[CH2CH-
(CH2)2]4, 1e). The temperature dependence of the 400-
1
MHz H spectrum is shown in Figure 3. The signal of
the two exocyclic methylene hydrogens (which at ambient
temperature is a doublet at 1.4 ppm with a J ) 6.6 Hz
due to the coupling with the CH hydrogen) broadens
considerably below -110 °C and decoalesces into a pair
of equally intense integrated signals, separated by 368
Hz at -161 °C. Such a feature indicates the existence of
a single conformer having diastereotopic methylene
hydrogens, due to the CH2-Cq rotation being slow in the
NMR time scale at such a low temperature. This also
implies that the conformer has S4 symmetry, in agree-
ment with the crystal structure shape,7 whose C2 static
symmetry becomes S4 with fast cyclopropyl-CH2 bond
rotation in solution. Indeed our ab initio calculations (see
of activation of 5.8 ( 0.15 kcal mol-1
.
The S4 symmetry adopted by 1e in solution implies
that, even in the presence of a fast cyclopropyl-CH2 bond
rotation, the four endocyclic methylene hydrogens must
all become anisochronous when the CH2-Cq rotation rate
becomes slow. Indeed the 400-MHz spectrum at -161 °C
shows how both the cis and trans ring proton signals
centered at -0.02 and +0.35 ppm appear to be split, their
separation being 36 and 47 Hz, respectively (Figure 3,
trace at -161 °C). The barrier obtained by computer
simulating these signals (Figure 3) is equal to that
(8) This equation takes the form J ) 7.76 cos2θ - 1.1 cos θ + 1.40,
as in: Haasnoot, C. A. G.; Leeuw, F. A. A. M.; Altona, C. Tetrahedron
1980, 36, 2783.
(7) Ramana, C. V.; Baquer, S. M.; Gonnade, R G.; Gurjar, M. K.
Chem. Commun. 2002, 614.
6390 J . Org. Chem., Vol. 67, No. 18, 2002

Conformational Studies by Dynamic NMR. 91
1
F IGURE 3. Temperature dependence of the experimental H NMR spectrum (400 MHz) of 1e in CHF2Cl/CHFCl2 (left). On the
right is displayed the simulation obtained with the rate constants shown.
obtained from the exocyclic methylene signals, meaning
that the same process (i.e. CH2-Cq rotation) is respon-
sible for the dynamic features observed for all the
methylene hydrogen signals of 1e.
center. The molecular asymmetry that renders methylene
protons diastereotopic is averaged out for both ortho and
meta positions by fast rotation of the phenyl ring.
The crystal structure (see Experimental Section) of
C(CH2Ph)4 shows that this molecule adopts a single
conformation in the solid state that is matched by the
ab initio computations. This has a C2 static symmetry
(Figure 4) that would become S4 when the Ph-CH2
rotation is fast, thus corresponding, in solution, to the
T3-type conformation of Scheme 1.
Tetr a ben zylm eth a n e (C(CH2P h )4, 1f). In the 300-
MHz 1H NMR spectrum the CH2 singlet significantly
broadens below -120 °C and shows a line width of more
than 70 Hz at about -165 °C. On further cooling the
sample, in the -170 to -175 °C range, the line becomes
so broad as to disappear under the noise, whereas the
lines of the solvents are still only 15 Hz wide. To account
for such changes, the diastereotopic hydrogens must have
a very large shift separation of 300 Hz or more. At even
lower temperature, the compound precipitates so, con-
trary to the case of 1e, we could not detect two individual
signals. In any case, since the above experiment indicates
that the line separation must be similar to that of 1e, a
coalescence temperature in the range -170 to -175 °C
implies5,6 a particularly low free energy of activation
between 4.2 and 4.5 kcal mol-1.9 The exchange process
observed might be, in principle, attributed to slow rota-
tion about the Ph-CH2 bond but no dynamic broadening
of the benzene ring signals was observed, so the process
is concluded to involve the bonds to the quaternary
Ca lcu la tion s. The calculated energies of conforma-
tional types T1 and T3 for compounds 1a -f are shown
in Table 2. Some calculated interconversion barriers
based on successive rotations of CH2X groups are also
included in Table 1. Ab initio calculations at the RHF
6.31G* level10 and empirical Molecular Mechanics (MM3
force field) calculations11 were employed (see Experimen-
tal Section).
Particular care had to be taken with empirical calcula-
tions of the tetracyclopropylmethyl and tetrabenzyl com-
pounds 1e and 1f to ensure that best minima had been
found, since conformational variation about one cyclo-
(10) As in the computer package Titan 1.0.5; Wavefunction Inc.:
Irvine, CA.
(9) Anderson, J . E.; de Meijere, A.; Kozhushkov, S. I.; Lunazzi, L.;
Mazzanti, A. J . Am. Chem. Soc. 2002, 124, 6706.
(11) Allinger, N. L.; Yuh, Y. H.; Lii, J . H. J . Am. Chem. Soc. 1989,
111, 8552.
J . Org. Chem, Vol. 67, No. 18, 2002 6391

Grilli et al.
most stable conformer is the T6 type (C2 symmetrry as
in Scheme 1), with energy 1.4 kcal mol-1 higher than T3.
The prediction of T3 to be the most stable conformer in
1e and 1f agrees well with the X-ray structure (Figure
4) in the solid state as well as with the low-temperature
NMR observations in solution.
Conformational energies calculated with MM3 are
expected to be sensitive to the dielectric constant used,
at least for compounds 1c, 1d , and 1f. In the first two of
these compounds there is, contrary to the experimental
observation, a small calculated preference, 0.31 and 0.18
kcal mol-1, respectively, for the T3 conformation over T1
(with the default dielectric constant of 1.5). This prefer-
ence reduces to 0.07 and 0.03 kcal mol-1, respectively,
when the default dielectric constant of 1.5 is increased
modestly to 5.0. The T1 conformation is eventually
calculated to be more stable than T3 at higher values of
the dielectric constant.
Discu ssion
Electron diffraction studies of compounds 1a , 1c, and
1d in the gaseous phase have been reported4,12,13 and both
T1 and T3 conformation types are found in all three
compounds. The T1:T3 population ratio for tetraethyl-
methane 1a is 66:344 while for the chloride 1c it is 50:
5012 and for the bromide 1d it is 58:42.13
There have been several crystal structure determina-
tions of acyclic compounds C(CH2 X)4 (1), and only
conformations of type T1 or T3 are found. Those adopting
a T1 conformation are 1, X ) Br,14,15 Cl,14,15 CH2CH2Br,14
or CHdCH2.16 Those adopting a T3 conformation are 1,
X ) CH2Br,14 cyclopropyl,7 and COOH.17 There are in
addition eight examples of Pentaerythritol derivatives 1,
X ) OR, of which six have a T1 structure18 and two have
a T3 structure.19
F IGUR E 4. X-ray diffraction (top) and ab initio computed
(bottom) structures of the most stable conformer T3 of
compounds 1e and 1f. The crystallographic data of 1f are
reported in the Experimental Section, for those of 1e see ref
7.
The characteristic of the other conformational types
T2 and T4-T6 is that each has at least one X,X parallel-
1,3 interaction, see Scheme 1. Such interactions are
destabilizing both for steric and for dipole/dipole reasons
and are plausibly the reasons for the higher energy of
these conformations. Bushweller and his collaborators20,21
have studied some compounds quite similar to our
present ones but with only two or three CH2X groups.
The conformational analysis is dominated by conforma-
propyl-CH2 or Ph-CH2 bond affects the minimum about
all CH2 bonds. Interconversion of conformations is a
much more complicated process in these molecules. The
T1-T3 transformation in 1e, for example, requires -120°
of rotation about two Cq-CH2 bonds, and two cyclopro-
pyl-CH2 bonds, so we made no attempt to simulate this
process.
Ab initio calculations10 for 1a -d indicate that in all
four cases a conformation of type T1 conformation is the
most stable, followed by the type T3 conformation, with
energy differences lower than 1 kcal mol-1. This should
lead to a minor (a few percent) population for the latter,
in agreement with experimental observations.
Molecular Mechanics calculations11 suggest that the
easiest pathway for T1 to T3 interconversion is via T2,
although the last conformation is too high in energy to
be significantly populated. We can thus imagine a suc-
cession of T1-T2-T3-T2′-T1′ interconversions with
molecules spending most of their lifetimes in T1 confor-
mations.
(12) Stølvek, R. Acta Chem. Scand. 1974, 28A, 327.
(13) Rustad, S.; Stølvek, R. Acta Chem. Scand. 1976, 30 A, 209.
(14) Newkome, G. R.; Arai, S.; Fronczek, F. R.; Moorefield, C. N.;
Lin, X.; Weis, C. D. J . Org. Chem. 1993, 58, 898.
(15) Klaeboe, P.; Klewe, B.; Martinsen, K.; Nielsen, G. J .; Powell,
D. L.; Stubbles, D. J . J . Mol. Struct. 1986, 140, 1.
(16) Bunz, U.; Vollhardt, K. P. C.; Ho, J . S. Angew. Chem., Int. Ed.
Engl. 1992, 31, 1648.
(17) Ermer, O.; Eling, A. Angew. Chem., Int. Ed. Engl. 1988, 27,
829.
(18) (a) Chao, L. C. F.; Decken, A.; Britten, J . F.; McGlinchey, M. J .
Can. J . Chem. 1995, 73, 1196. (b) Quast, L.; Nieger, M.; Dotz, K. H.
Organometallics 2000, 19, 2179. (c) Uechi, T.; Fuchita, Y. Bull. Chem.
Soc. J pn. 1992, 65, 2831. (d) Ladd, M. F. C. Acta Crystallogr. Sect. B
1979, 35, 2375. (e) Newkome, G. R.; Weis, C. D.; Xiaofeng, L.; Fronczek,
F. R. Acta Crystallogr. Sect. C 1993, 49, 998. (f) Bormann, L. M.;
Baughman, R. G.; Mumba, P.; Meloan, C. E. Acta Crystallogr. Sect. C
1993, 49, 1982.
(19) (a) Marsh, R. E. Acta Crystallogr. Sect. B 1997, 53, 317. (b)
Cady, H. H.; Larson, A. C. Acta Crystallogr. Sect. B 1975, 31, 1864.
(20) Bushweller, C. H.; Whalon, M. R.; Fleischmann, S. H.; Rithner,
C. D.; Sturges, J . S. J . Am. Chem. Soc. 1979, 101, 4703.
(21) Bushweller, C. H.; Whalon, M. R.; Laurenzi, B. J . Tetrahedron
Lett. 1981, 22, 2945.
Ab initio and MM calculations predict that a different
conformational situation holds for compounds 1e and 1f.
Conformations of type T1 are much less stable than type
T3 for these compounds. Whereas in 1e ab initio com-
putations predict T1 to be the second most stable
conformer (with an energy higher than T3 by 1.3 for
compounds 1e and 1f, as in Table 2), in 1f the second
6392 J . Org. Chem., Vol. 67, No. 18, 2002

Conformational Studies by Dynamic NMR. 91
SCHEME 3a
tions which have no such X-X parallel-1,3 interactions.
In the compound CH3C(CH2Cl)3, for example,21 there are
three conformational types with no such interactions, and
the low-temperature spectrum shows (at -171 °C) three
sets of signals of different intensity.
The spectral appearance, at low temperatures, for 1c
and 1d and, by implication, 1a , namely a singlet and an
AB spectrum for T1 and T3 conformations respectively,
implies that interconversion of T1 with T3 and also
interconversion between degenerate conformations of T3
are slow on the NMR time scale. The modes of confor-
mational interconversion of other molecules of the type
C(CX2Y)4 have recently been analyzed in detail,9 and
suggest that the T1-T3 interconversion is by way of T2.
This implies rotation of about 120° by two appropriate
CX2Y groups, and calculations for CX2Y ) isopropyl
suggested that a certain degree of cooperativity in these
rotations lowers the overall barrier to rotation, so that
the best minimum energy T2 conformation may not be
exactly on the easiest interconversion pathway. Our
present MM3 calculations for 1a to 1d suggest that
rotation of less-branched CH2X groups does not require
significant cooperation from neighboring groups, so a
straightforward T1-T2-T3 pathway seems likely. It is
surprising that MM3 calculations for the barrier in the
two halides give results more than 2 kcal mol-1 higher
than experiment. Use of a higher dielectric constant
lowers these T1-T3 interconversion barriers: the 10.0
kcal mol-1 barrier quoted for 1c in Table 1, for example,
is reduced to 9.3 kcal mol-1 when a dielectric constant
ꢀ ) 10 (rather than the standard ꢀ ) 1.5) is used, a value
nearer to the experimental one, although hardly in good
agreement with it.
MM3 calculations suggest that interconversion of
degenerate versions of the T3 conformation, the only
observed process for 1e and 1f, and the minor process
for 1a -d , takes place via the T6 conformation, rather
than by what is now the relatively high-energy T1
conformation. This T6 is followed by a second T6 con-
formation, or a T2 conformation, i.e., T3-T6-T6′-T3′
or T3-T6-T2-T3′ with only T3 conformations of 1e and
1f detectably populated.
The contrasting preferences for the T1 conformation
in 1a -d and for the T3 conformation in 1e and 1f can
be understood in light of an earlier observation of Alder
and co-workers.3 They found that in tetrapropylmethane,
1b, the second bond from the quaternary center in each
propyl group adopts the anti conformation completely.
a
Structures A and B show one side chain of the T1 conforma-
tion of compound 1b, where the second bond is in an anti or a
gauche conformation, respectively. In the latter, the end of the
chain interferes with another propyl group. Structure C shows
that in compound 1e, where the chain branches at the â-position,
interference with another side chain is inevitable.
phenyl rings interfere with each other so, as with 1e,
there is no particularly stable T1 conformation. Even in
the preferred T3 structure of tetrabenzyl derivative 1f
there is a conformational compromise since the Cq-CH2-
C-Cortho torsion angles are not 90° in the crystal but,
alternately, 96.3° and -107.2° and are calculated to be,
alternately, +80° and -106° (ab initio). A T3 conforma-
tion is perhaps to be favored for C(CH2X)4 when any
substantial group X occurs â to the quaternary center
and cannot avoid interactions with it.
The biased position of the equilibrium in 1a , 1c, and
1d produces very little detail in the exchange-broadened
NMR spectra, so it is not possible to distinguish whether
in these compounds there is any involvement of inter-
conversions of the T3-T6-T6′-T3′ type as well as of
the T3-T2-T1-T2′-T3′ type.
In the earlier work9,23,24 on compounds of the type
C(CX2Y)4 with one hydrogen atom Y and two larger
groups, a type T1 or a type T3 conformation is always
adopted rather than the other four types. Tetracyclo-
hexylmethane,23 tetraisopropylmethane,9,24 and tetracy-
clopropylmethane9,24 have been studied, and once again
the stability of conformational types is governed by
parallel-1,3 interactions of large X-groups. In contrast to
the present study such interactions cannot be absent, but
in conformation types T1 and T3 there are only four such
interactions, while there are five or six in the remaining
four types. Bearing on our present results for compounds
1e and 1f, there is an interesting contrast between
tetraisoproylmethane and tetracyclopropylmethane. The
former adopts a type T1 for preference over the T3, the
population ratio being 93:7 at -110 °C.9 When the flexible
isopropyl group is replaced by a more rigid cyclopropyl
group to give the latter compound, molecules are 100%
in a type T3 conformation, and the T1 conformation is
As Scheme 3 shows the quaternary center requires this
anti arrangement A for the Cq-CH2-CH2-CH3 torsion
angle, rather than the gauche B in which the methyl
group has marked interactions with one other propyl
chain. With the more branched cyclopropylmethyl groups,
such an anti conformation free of interaction with other
groups is not possible (see structure C).
In the tetrabenzyl compound 1f, the preferred confor-
mation for an individual benzyl group would have the
Cq-CH2 bond orthogonal to the plane of the benzene ring
as in neopentyl benzene.22 The consequence of this is that
(22) (a) Andersson, S.; Drakenberg T. J . Magn. Reson. 1983, 21, 730.
(b) Anderson, J . E.; Barkel, D. J . D. J . Chem. Soc., Perkin Trans. 2
1984, 1053. (c) Schaefer, T.; Penner, G. H.; Takeuchi, C. S.; Beaulieu,
C. Can. J . Chem. 1989, 67, 1283.
(23) Columbus, I.; Biali, S. E. J . Org. Chem. 1994, 59, 8132.
(24) Kozhushkov, S. I.; Kostikov, R. R.; Molchanov, A. P.; Boese,
R.; Benet-Buchholz, J .; Schreiner, P. R.; Rinderspacher, C.; Ghiviriga,
I.; De Meijere, A. Angew. Chem., Int. Ed. 2001, 40, 180.
J . Org. Chem, Vol. 67, No. 18, 2002 6393

Grilli et al.
calculated to be at least 10 kcal mol-1 less stable.9
Calculations for 1e and 1f show that here likewise the
T1-type conformation is significantly destabilized.
TABLE 3. Selected Tor sion An gles (d eg) for 1e a n d 1f in
th e P r efer r ed Con for m a tion Typ e T3 As F ou n d in th e
Cr ysta l, by Ab In itio a n d Molecu la r Mech a n ics
Ca lcu la tion s w ith Ca lcu la tion s for th e T1-Typ e
Con for m a tion Sh ow n for Com p a r ison a
Exp er im en ta l Section
conformation T3
conformation T1
Ma ter ia ls: 3,3-Dieth ylp en ta n e (tetr a eth ylm eth a n e, 1a )
was prepared as described in the literature.4
dihedral
crystal ab initio MM3 ab initio mm3
4,4-Dip r op ylh ep ta n e (tetr a p r op ylm eth a n e, 1b) was
prepared25 by reacting dipropyl ketone with propylmagnesium
bromide. The resulting tripropylmethanol was treated with
concentrated HCl to obtain the corresponding chloride (bp 81-
83 °C at 13 mbar), which was subsequently reacted with
allylmagnesium bromide. Reduction of the resulting 4,4-
dipropyl-1-heptene (bp 78-79 °C at 8 mbar) by means of H2
on Pd/C (alternatively ammonium formiate on Pd/C also can
be used) yielded the final compound 1b. 1H NMR (CDCl3, 300
MHz) δ 1.12(t, 12H, J ) 6.5 Hz), 1.30-1.45 (m, 16H). 13C NMR
(CDCl3, 75.45 MHz) δ 14.51 (CH3), 16.21 (CH2), 37.39 (Cq),
39.32 (CH2). Anal. Calcd for C13H28
Found: C, 84.66; H, 15.38.
1,3-Dich lor o-2,2-bis(ch lor om eth yl)p r op a n e (1c) a n d
1,3-d ibr om o-2,2-bis(br om om eth yl)p r op a n e (1d ) were com-
mercially available and were further purified by recristalli-
zation (absolute ethanol).
1eb XA-CA-Cq-CB
XB-CB-Cq-CA
XC-CC-Cq-CD
XD-CD-Cq-CC
Cq-CA-C-H
61.9
179.2
62.6
177.8
51.4
-50.2
54.2
65.8
173.1 177.5
65.7
62.3
175.1
175.1
175.4
175.7
62.6 -175.1 -175.6
173.2 177.7 -175.1 -175.9
54.7
-54.7 -56.2
54.7 56.0
-54.7 -56.1
48.2 63.7
55.8
47.7
47.8
-47.7
-47.7
167.6
167.5
48.4
48.5
Cq-CB-C-H
Cq-CC-C-H
Cq-CD-C-H
1f XA-CA-Cq-CB
XB-CB-Cq-CA
XC-CC-Cq-CD
XD-CD-Cq-CC
-48.7
-48.9
170.4
175.1
169.5
174.1
66.5
-55.3
51.1
179.7 -179.5 175.9
51.1 48.2
179.7 -179.5 173.0 -175.2
66.0 -175.2
: C, 84.69; H, 15.31.
Cq-CA-C-Cortho -86.7 -101.8 -76.1
Cq-CB-C-Cortho 79.5 79.3 76.2
Cq-CC-C-Cortho -86.7 -101.8 -76.4
Cq-CD-C-Cortho 79.5 79.3 76.5
35.9
36.0
83.8
83.9
70.3
64.4
68.4
a
b
Atom labeling for T1 and T3 is indicated in Scheme 1. The
crystal structure data for 1e are taken from ref 7.
1,3-Dic yc lop r op yl-2,2-b is(c yc lop r op ylm e t h yl)p r o-
p a n e (tetr a cyclop r op ylm eth ylm eth a n e, 1e) was prepared
as described in the literature.7
1-(2,2-Diben zyl-3-p h en ylp r op yl)ben zen e (tetr a ben zyl-
m eth a n e, 1f) was prepared as described in the literature.26
1H NMR (CDCl3, 300 MHz) δ 2.75(s, 8H), 7.15-7.35 (m, 20H).
13C NMR (CDCl3, 75.45 MHz) δ 42.06 (CH2), 42.26 (Cq), 126.05
(CH), 127.90 (CH), 131.65 (CH), 139.34 (Cq). Anal. Calcd for
tion of the line shape was performed by a PC version of the
computer program based on DNMR6 routines27 and the best
fit was visually judged by superimposing the plotted and
experimental traces. As often reported in the case of confor-
mational processes,28the free energies of activation were found
essentially independent of the temperature in the examined
range, thus indicating that the ∆Sq values are negligible within
the experimental uncertainty.
C
29H28: C, 92.50; H, 7.50. Found: C, 92.45; H, 7.55.
X-r a y Diffr a ction : Cr ysta l d a ta of 1-(2,2-d iben zyl-3-
Com p u ta tion s. The Molecular Mechanics (MM3 force
field11) approach was used for all the conformational types T1-
T6 of each compound and, for the three conformers with the
p h en ylp r op yl)ben zen e (Tetr a ben zylm eth a n e, 1f): C29H28
(376.51), trigonal, space group P3221, Z ) 3, Z primary ) 0.5,
a ) 9.9127(4) Å, c ) 18.9110(9) Å, V ) 1609.27(12) Å3, Dc )
1.166 g cm-3, F(000) ) 606, µMo) 0.065 mm-1, T ) 293 K,
Crystal size 1.2 × 1.2 × 1.0 mm3. Data were collected by using
a graphite monochromated Mo KR X-radiation (λ ) 0.71073
Å) range 2.37° < θ < 25.00°. Of 14577 reflections collected,
1907 were found to be independent (Rint ) 0.0311), 1846 of
which were considered as observed [I > 2σ(I)], and were used
in the refinement of 133 parameters leading to a final R1 of
0.0330 and a Rall of 0.0342. The structure was solved by direct
method and refined by full-matrix least squares on F2, using
SHELXTL 97 program packages. In refinements were used
weights according to the scheme w ) [σ2(Fo2) + (0.0441P)2 +
0.1825P ]-1, where P ) (Fo2) + 2Fc2)/3. The hydrogen atoms
were located by geometrical calculations and refined by using
a “riding” method. wR2 was equal to 0.0874. The goodness of
fit parameter S was 1.094. The largest differences between
the peak and hole were 0.162 and -0.105 eÅ-3. Crystal-
lographic data (excluding structure factors and including
selected torsion angles) have been deposited with the Cam-
bridge Crystallographic Data Center, CCDC no. 188152.
NMR Mea su r em en ts. The samples for the low-tempera-
ture determinations were prepared by connecting to a vacuum
line the NMR tubes containing the desired compounds and
some C6D6 or CD2Cl2 for locking purpose. The gaseous solvents
(CHF2Cl and CHFCl2 in about a 4:1 v/v proportion) were
subsequently condensed therein by means of liquid nitrogen.
The tubes were then sealed in a vacuum and introduced in
the precooled probe of the spectrometers. The temperature was
calibrated by means of a Ni/Cu thermocouple inserted in the
NMR probe before the measurements. The computer simula-
lower energy, also by ab initio (at the RHF 6.31G* level10
)
methods. For compounds 1c-f, not previously calculated by
Alder and co-workers, the relative MM3 energies of the six
conformational types T1-T6 are the following: 1c 0.31, 3.48,
0.00, 6.86, 9.53, 2.98; 1d 0.18, 3.68, 0.00, 7.38, 10.20, 3.20; 1e
0.77, 1.76, 0.00, 3.58, 6.34, 1.71; and 1f 3.30, 2.41, 0.25, 4.45,
8.08, 0.00. For compounds 1e and 1f various versions of each
conformational type were found, depending on the disposition
about cyclopropyl-CH2 or phenyl-CH2 bonds, respectively, but
only the lowest energy value is reported. Significant torsion
angles found for compounds 1e and 1f which adopt a T3-type
conformation in the crystal are shown in Table 3 along with
ab initio and molecular mechanics (MM3) calculated values
for that conformation and for T1.
Ackn owledgm en t. One of the authors (M.P.) thanks
CINMPIS for a fellowship. L.L. and A.M. received
financial support from MURST, Rome (national project
“Stereoselection in Organic Synthesis”), and from the
University of Bologna (Funds for selected research
topics 2001-2002). A research fund for young research-
ers from the University of Bologna is also acknowledged
by S.G.
J O025984K
(27) QCPE program no. 633, Indiana University: Bloomington, IN.
(28) Hoogasian, S.; Bushweller, C. H.; Anderson, W. G.; Kingsley,
G. J . Phys. Chem. 1976, 80, 643. Lunazzi, L.; Cerioni, G.; Ingold, K.
U. J . Am. Chem. Soc. 1976, 98, 7484. Bernardi, F.; Lunazzi, L.;
Zanirato, P.; Cerioni, G. Tetrahedron 1977, 33, 1337. Lunazzi, L.;
Magagnoli, C.; Guerra, M.; Macciantelli, D. Tetrahedron Lett. 1979,
3031. Anderson, J . E.; Tocher, D. A.; Casarini, D.; Lunazzi, L. J . Org.
Chem. 1991, 56, 1731. Cremonini, M. A.; Lunazzi, L.; Placucci, G.;
Okazaki, R.; Yamamoto, G. J . Am. Chem. Soc. 1992, 114, 6521. Borghi,
R.; Lunazzi, L.; Placucci, G.; Cerioni, G.; Foresti, E.; Plumitallo, A. J .
Org. Chem. 1997, 62, 4923.
(25) Petrov, A. D.; Tschernyschev, E. A. Izv. Akad. Nauk SSSR Ser.
Khim. 1952, 1082; Chem. Abstr. 1954, 48, 565e.
(26) Trotman, E. R. J . Chem. Soc. 1925, 127, 88. For the preparation
of the intermediates see: Hill, G. A.; Nelson, W. C.; Dunnell, R. L.;
Moody, L. S. J . Am. Chem. Soc. 1941, 63, 1367. Rodriguez, G.; Bazan,
G. C. J . Am. Chem. Soc. 1997, 119, 343.
6394 J . Org. Chem., Vol. 67, No. 18, 2002