Conformational Studies by Dynamic NMR. 99. Experimental and Computed Determination of Rotation Barriers in the Crystalline State: The Case of Naphthylphenylsulfoxide

Article abstract of DOI:10.1021/jo049843p

The ¹³C NMR CP-MAS spectrum of 2-naphthylphenylsulfoxide in the solid state displays line broadening effects due to the restricted rotation about the Ph-S bond. Line shape simulation of the temperature-dependent traces allowed the corresponding

Full text of DOI:10.1021/jo049843p

the severe constraints of the crystalline environment.5
When interconversion processes can be observed, the
barriers involved are much higher than in solution,6-11
because of the restrictions to the internal motions
brought about by the surrounding lattice. For these
reasons computations that treat the molecule as an
isolated entity cannot reproduce the values of the barriers
measured in the solid state. There are, however, compu-
tational approaches that take into account the situation
of a molecule in the crystal5a,12-14 and that have been
applied to describe the conformation of polymers in the
solids15 and the dynamics of reorientation within the
lattice.5,12,14-16 However, there are not available examples
where theory and experiment have been both applied to
investigate the interconversion of conformers (or to-
pomers) of organic molecules in the crystalline state. We
thus searched for a case where the barrier for such a
process could be measured in the solids with the purpose
of reproducing this value by theoretical methods.
The title compound (2-naphthylphenylsulfoxide, 1)
turned out to be appropriate for this scope. MM computa-
tions,17 which treat the molecule as an isolated object,
provide the energy map as a function of the rotation
about the phenyl-S and naphthyl-S bonds (x- and
y-axis, respectively, as in Figure 1). The rotation about
the phenyl-S bond (dashed line of Figure 1) corresponds
to a topomerization process that exchanges the ortho,
ortho′ and meta, meta′ positions with a computed barrier
Con for m a tion a l Stu d ies by Dyn a m ic NMR.
99.1 Exp er im en ta l a n d Com p u ted
Deter m in a tion of Rota tion Ba r r ier s in th e
Cr ysta llin e Sta te: Th e Ca se of
Na p h th ylp h en ylsu lfoxid e
Daniele Casarini,*,2a Lodovico Lunazzi,*,2b
Andrea Mazzanti,2b Pierluigi Mercandelli,2c and
Angelo Sironi*,2c
Chemistry Department, University of Basilicata, Potenza,
Italy, Department of Organic Chemistry “A. Mangini”,
University of Bologna, Italy, and Dipartimento di Chimica
Strutturale e Stereochimica Inorganica, University of
Milan and CNR ISTM, Italy
lunazzi@ms.fci.unibo.it
Received J anuary 27, 2004
Ab st r a ct : The 13C NMR CP-MAS spectrum of 2-naph-
thylphenylsulfoxide in the solid state displays line broaden-
ing effects due to the restricted rotation about the Ph-S
bond. Line shape simulation of the temperature-dependent
traces allowed the corresponding barrier to be determined
in the solids (14.7 kcal mol-1). By making use of the
information obtained from single-crystal X-ray diffraction,
this barrier could be satisfactorily reproduced by theoretical
calculations (14.5 kcal mol-1) that take into account the
correlated phenyl motion involving a large set of molecules
in the crystalline state
(5) (a) Gavezzotti, A.; Simonetta, M. Chem Rev. 1982, 82, 1. (b)
Braga, D. Chem Rev. 1992, 92, 633.
(6) (a) Mo¨ller, M.; Gronski, W.; Cantow, H.-J .; Ho¨cker, H. J . Am.
Chem. Soc. 1984, 106, 5093. (b) Emeis, D.; Cantow, H.-J .; Mo¨ller, M.
Polym. Bull. (Berlin) 1984, 12, 557.
(7) Twyman, J . M.; Dobson, C. M. J . Chem. Soc., Chem. Commun.
1988, 786.
The rotational processes that allow the interconversion
between conformers (or topomers) have been widely in-
vestigated by dynamic NMR spectroscopy in solution.3
These studies are often accompanied by theoretical cal-
culations that usually reproduce quite satisfactorily the
experimental barriers.4 This is because the approxima-
tion of considering an isolated molecule model does not
conflict too severely, in the majority of cases, with the
actual situation. The solvent effects upon the barrier are
relatively modest and, if necessary, can be accounted for
by introducing in the calculations appropriate parameters
related to the dielectric constant of the medium.
(8) (a) Riddell, F. G.; Arumugam, S.; Anderson, J . E. J . Chem. Soc.,
Chem. Commun. 1991, 1525. (b) Barrie, P. J .; Anderson, J . E. J . Chem.
Soc., Perkin Trans. 2 1992, 2031. (c) Anderson, J . E.; Casarini, D.;
Lunazzi, L.; Mazzanti, A. J . Org. Chem. 2000, 65, 1729. (d) Anderson,
J . E.; Casarini, D.; Lunazzi, L.; Mazzanti, A. Eur. J . Org. Chem. 2000,
479. (e) Aliev, A. E.; Anderson, J . E.; Butler, D.; Gonzalez-Outerin˜o,
J .; Lunazzi, L.; Mazzanti, A.; Steed, J . Wilhelm, R. Eur. J . Inorg. Chem.
2002, 133.
(9) (a) Riddell, F. G.; Arumugam, S.; Harris, K. D. M.; Rogerson,
M.; Strange. J . H. J . Am. Chem. Soc. 1993, 115, 1881. (b) Riddell, F.G.;
Bremner, M.; Strange, J . H. Magn. Reson. Chem. 1994, 32, 118. (c)
Riddell, F. G.; Bernath, G.; Fu¨lo¨p, F. J . Am. Chem. Soc. 1995, 117,
2327. (d) Riddell, F. G.; Rogerson, M. J . Chem. Soc., Perkin Trans. 2
1996, 493. (e) Riddell, F. G.; Cameron, S.; Holmes, S. A.; Strange, J .
H. J . Am. Chem. Soc. 1997, 119, 7555.
(10) Barnes, J . C.; Chudek, J . A.; Hunter, G.; Blake, A. J .; Dyson,
P. J .; J ohnson, B. F. G.; Weissensteiner, W. J . Chem. Soc., Faraday
Trans. 1995, 91, 2149.
(11) (a) Casarini, D.; Lunazzi, L.; Mazzanti, A. J . Org. Chem. 1998,
63, 9125. (b) Casarini, D.; Foresti, E.; Lunazzi, L., Mazzanti, A. Chem.
Eur. J . 1999, 5, 3501. (c) Lunazzi, L.; Mazzanti, A.; Casarini, D.; De
Lucchi, O.; Fabris, F. J . Org. Chem. 2000, 65, 883. (d) Casarini, D.;
Lunazzi, L.; Mazzanti, A.; Simon, G. J . Org. Chem. 2000, 65, 3207. (e)
Casarini, D.; Lunazzi, L.; Mazzanti, A. Angew. Chem., Int. Ed. 2001,
40, 2536.
(12) Ku¨mmerlen, J .; Sebald, A. Organometallics 1997, 16, 2971.
(13) Mercandelli, P.; Moret, M.; Sironi, A. Inorg. Chem. 1998, 37,
2563.
(14) Neuman, M. A.; J ohnson, M. R.; Radaelli, P. G.; Trommsdorff,
H. P.; Parker, S. F. J . Chem. Phys. 1999, 110, 516.
(15) Miyoshi, T.; Hayashi, S.; Imashiro, F.; Kaito, A. Macromolecules
2002, 35, 2624.
(16) Galli, S.; Mercandelli, P.; Sironi, A. J . Am. Chem. Soc. 1999,
121, 3767.
(17) MMX force field as implemented in the computer package PC
Model, v 7.5; Serena Software: Bloomington, IN.
The situation is quite different when dealing with
molecules in the solid state, where the corresponding
conformers often cannot even interconvert as a result of
(1) Part 98: Bartoli, G.; Lunazzi, L.; Massaccesi, M.; Mazzanti, A.
J . Org. Chem. 2004, 69, 821.
(2) (a) University of Basilicata. (b) University of Bologna. (c)
University of Milan
(3) (a) Dynamic Nuclear Magnetic Resonance Spectroscopy; J ack-
man, L. M., Cotton, F. A., Eds; Academic Press: New York, 1975. (b)
Mann, B. E. Prog. NMR Spectrosc. 1977, 11, 95. (c) Kaplan, J . I.;
Fraenkel, G. NMR of Chemically Exchanging Systems; Academic
Press: New York, 1980. (d) Sandstro¨m, J . Dynamic NMR Spectroscopy;
Academic Press: London, 1982. (e) Oki, M. Applications of Dynamic
NMR Spectroscopy to Organic Chemistry; VCH: Deerfield Beach, 1985.
(4) (a) Berg, U.; Sandstro¨m, J . Adv. Phys. Org. Chem. 1989, 25, 1.
(b) Anderson, J . E. In The Chemistry of Alkanes and Cycloalkanes;
Patai, S., Rappoport, Z., Eds.; Wiley: Chichester, 1992; Chapter 3.II.D.
(c) Bushweller, C. H. In Acyclic Organonitrogen Stereodynamics;
Lambert, J . B., Takeuchi, Y., Eds.; VCH: New York, 1992; Chapter 1.
(d) J ennings, W. B.; Wilson, V. E. In Acyclic Organonitrogen Stereo-
dynamics; Lambert, J . B., Takeuchi, Y., Eds.; VCH: New York, 1992;
Chapter 6.
10.1021/jo049843p CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/20/2004
3574
J . Org. Chem. 2004, 69, 3574-3577

F IGURE 1. Energy map of 1 (considered as an isolated
molecule) computed as a function of the rotation about the
naphthyl-SO and phenyl-SO dihedral angles. The two lines
represent the pathways connecting the ground states. The
diamonds and the square identify the grounds states of the E
and Z conformers, respectively. The circle and the triangle
identify the two possible transition states (relative energies
1.9 and 1.8 kcal mol-1, respectively).
of 1.9 kcal mol-1. The rotation about the naphthyl-S
bond (solid line of Figure 1) corresponds to the exchange
of the conformer where the C1-C2-S-O dihedral angle is
29.5° (thus Z-conformer) with the nearly equally stable
(within the computing approximations) conformer where
this angle is -150.4° (E-conformer). This second com-
puted barrier (1.8 kcal mol-1 ) is almost equal to the
previous one, as conceivable for the case of an isolated
molecule. According to this model, the two rotation
processes are not correlated but take place independently
of each other, as indicated by the fact that the two lines
of Figure 1 are parallel to the respective axes.18,19
Even allowing for the approximation involved in this
theoretical approach, the computed barriers are predicted
to be too low for an experimental determination by NMR
spectroscopy; even at -160°C, in fact, we could not detect
any evidence of such a motion in solution (see Experi-
mental Section).
F IGURE 2. (Left) Relevant region of the solid-state CP MAS
13C NMR spectrum of 1 (75.5 MHz) as function of temperature.
(Right) Spectra simulated with the rate constants indicated.
displayed in Figure 2, where the relevant portion of the
solid-state spectrum of 1 is reported as a function of
temperature, the two ortho carbons of the phenyl ring
exhibit a single signal (125.0 ppm)20 at +63°C since in
these conditions the Ph-S rotation is fast in the NMR
time scale. This line broadens considerably on cooling and
eventually splits, below -30°C, into a pair of lines at
123.3 and 127.6 ppm, the latter being accidentally
coincident with that of carbon C-1 of the naphthalene ring
(in Figure 2 these two lines are indicated21 as o and o′).
Such a spectral behavior is completely analogous to that
reported for the intramolecular motions occurring in the
solid state (e.g., bond rotation,7,8a,b,9a-e,11a ring inversion,6a,11e
tautomerism,22): in the cases reported in the mentioned
references these spectra were interpreted by the same
Effects due to a rotation process could be, however,
observed in the crystals of 1 near ambient temperature
by means of CP-MAS NMR spectroscopy. This suggests
that the corresponding barrier has increased in the
crystal with respect to the value expected in solution. As
(20) The ortho and meta carbon signals of the solid-state spectrum
at +63 °C (125.0 and 128.5 ppm, respectively) were assigned by analogy
with the shifts (125.0 and 129.3 ppm, respectively) of the solution
spectrum (see Experimental Section). Such an assignment was further
supported by the observation that these two signals grow much more
than those of all other carbons when a standard single pulse experi-
ment (without cross polarization) is applied with a 16 s delay time.
This is due to the faster relaxation rate that is a consequence of the
rapid rotation of the phenyl ring at high temperature (all other carbon
signals would require a delay at least 10 times longer to display a
significant growth in this single pulse experiment).
(18) J og. P. V.; Brown, R. E.; Bates, D K. J . Org. Chem. 2003, 68,
8240.
(19) (a) Casarini, D.; Grilli, S.; Lunazzi, L.; Mazzanti, A. J . Org.
Chem. 2001, 66, 2757. (b) Grilli, S.; Lunazzi, L.; Mazzanti, A. J . Org.
Chem. 2001, 66, 4444. (c) Grilli, S.; Lunazzi, L.; Mazzanti, A. J . Org.
Chem. 2001, 66, 5853.
J . Org. Chem, Vol. 69, No. 10, 2004 3575

configuration S since compound 1 crystallizes as a con-
glomerate of enantiomorphic crystals. This was ascer-
tained by means of an enantioselective HPLC determi-
nation where a random mixture of a great number of
crystals yielded two distinct peaks, having the same
enantiomer ratio (2:1) as that determined by the 13C
NMR solution spectrum in a chiral environment26 (the
enrichment is due to the use of enantiopure menthol in
the synthetic procedure, as reported in Supporting
Information). On the contrary, examination of a single
crystal in the HPLC experiment yielded a single peak,
corresponding to only one of the two enantiomers. In
addition use was made of the X-ray powder spectrum (see
Supporting Information) to ensure that the bulk phase
employed for the solid-state CP-MAS NMR experiment
is the same as that for which the crystal structure was
derived.
F IGURE 3. Plot of the unit cell content view along the 100
direction.
The data obtained from the X-ray diffraction determi-
nation were instrumental for calculating the Ph-SO
rotation barrier in the crystal. These computations were
performed by employing a local version of the MM3
program, upgraded in order to deal with crystal lattices.13
Minimization was carried out with the Newton-Raphson
algorithm using the standard MM3(92) force field.27
To estimate the rotational barrier of the phenyl group
in the solid state, we used the DRIVE option of the MM3
program to force the phenyl of the reference molecule
(RM) to rotate within the field generated by 256 sur-
rounding molecules (SMs) positioned in accord with the
(time and space) averaged structural parameters ob-
served in the X-ray crystal structure determination. The
SMs have been chosen on the basis of a cutoff radius of
20 Å from the RM. In this first attempt the rotation of
the phenyl was hampered by its clashing with a nearby
molecule.
type of line shape simulation used for the dynamic
exchange in solution.23 The spectra of Figure 2 thus
provide evidence that the rotation rate of the phenyl
group varies with temperature, rendering the ortho
carbons inequivalent, thus anisochronous, at low tem-
perature (the analogous splitting expected for the meta
carbon line could not be observed because of overlapping
with a number of the naphthalene carbon signals). All
of the lines of the naphthyl carbons, on the contrary,
remain sharp, indicating the lack of a detectable rotation
process about the naphthyl-S bond and suggesting,
therefore, that the naphthyl moiety appears to be firmly
packed in the lattice. Line shape simulations24 (Figure
2, right) yield the rate constant for the Ph-S rotation
process at various temperatures. The corresponding ∆Gq
25
value (14.7 ( 0.4 kcal mol-1
)
shows that the barrier in
the solids is, as conceivable, much higher than that
computed for the isolated molecule.
Indeed, as highlighted in Figures 3 and 4, the structure
can be formally “cut” in layers perpendicular to the c-axis
because both naphthyl and phenyl groups are “segre-
gated”, i.e., they mainly interact with (are close to)
themselves. In a second run, we have followed the phenyl
rotation by allowing the first enclosure shell (ES) of the
RM (i.e., a cluster of nine nearest-neighbors molecules
surrounding the RM; see Figure 4) to relax. This ap-
proximation is obviously valid only if, upon rotation of
the phenyl of the RM, all molecules of the ES return to
their original position and if the set of SMs is large
enough (here 247 molecules) to contain at least a second
shell of rigid next-nearest neighbors to ensure that the
ES is not allowed undue freedom of motion. Within this
approach, on applying the suitable DRIVE option, the
phenyl smoothly rotates about the S-C bond (Figure 5),
the RM and the ES return on themselves, and the
computed barrier (14.5 kcal mol-1) matches that mea-
sured in the solid-state CP MAS NMR experiments (14.7
kcal mol-1). When applied to the naphthyl-SO bond
rotation these calculations showed that this motion had
such an exceedingly high activation energy as to prevent
The structure obtained by single-crystal X-ray diffrac-
tion (Figure 3) corresponds to the E-conformer whose
shape, apparently, fits the crystal cell better than the
essentially isoenergetic Z-conformer. It should be also
outlined that although the chirality of the sulfur atom
implies the existence of a pair of enantiomers, the
examined crystal contains only the enantiomer with the
(21) The assignment of the upfield signal to the carbon syn to the
oxygen of the SO moiety (labeled o′ in Figure 2) is the result of a
B3LYP/6-31+G(2d,p) ab initio calculation, as implemented in the
Gaussian98 A.9 program (courtesy of Prof. D. Nanni, University of
Bologna). Because of the analogous values observed in the case of 1
for the shifts in solution and in the solids, the assumption was made
that the sequence of the two ortho and ortho′ 13C signals, computed
for the isolated molecule, is the same as that occurring in the crystal.
It has to be stressed, however, that such an assignment is immaterial
for the experimental determination of the rotation barrier, since the
opposite assignment would obviously yield the same ∆Gq value.
(22) Limbach, H. H.; Henning, J .; Kendrick, R.; Yannoni, C. S. J .
Am. Chem. Soc. 1984, 106, 4059.
(23) Above +63 °C, the averaged line of the ortho carbons does not
sharpen further, as would occur in the solution spectra, but begins to
broaden because the motional frequency moves from the site exchange
regime to the dipolar broadening regime,7,9b,e a feature typical of NMR
dynamics in the solids (see also: Frey, M. H.; DiVerdi, J . A.; Opella,
S. J . J . Am. Chem. Soc. 1985, 107, 7311. Ripmeester, J . A. J . Inclusion
Phenom. 1988, 6, 31. Ratcliffe, C. I.; Ripmeester, J . A.; Buchanan, G.
W.; Denike, J . K. J . Am. Chem. Soc. 1992, 114, 3204.). Above this
temperature this type of line shape analysis24 could not be applied
anymore.
(26) The 13C spectrum was obtained at 150.9 MHz (CDCl3 as solvent)
in the presence of a nearly equimolecular amount of the enantiopure
(R)-l-1-(9-anthryl)-2,2,2-trifluoroethanol (Pirkle, W. H. J . Am. Chem.
Soc. 1966, 88, 1837).
(27) (a) Allinger, N. L.; Yuh, Y. H.; Lii, J .-H. J . Am. Chem. Soc. 1989,
111, 8551. (b) Lii, J .-H.; Allinger, N. L. J . Am. Chem. Soc. 1989, 111,
8566. (c) Lii, J .-H.; Allinger, N. L. J . Am. Chem. Soc. 1989, 111, 8576.
(24) QCPE program 633; Indiana University: Bloomington, IN.
(25) Within the experimental uncertainty the ∆Gq value was found
to be independent of temperature, indicating that the ∆Sq term must
be quite small.
3576 J . Org. Chem., Vol. 69, No. 10, 2004

Exp er im en ta l Section
2-Na p h th ylp h en ylsu lfoxid e28 (1) has been obtained by
means of the general procedure reported in the literature29 (see
Supporting Information).
NMR Sp ectr oscop y. The 1H and 13C assignments in solution
were obtained by means of DPFGSE-NOE, COSY, DEPT,
gHSQC, and gHMBC sequences. The very low temperature
(-160° C) experiment, which failed to show the effects of re-
stricted rotation in solution, was carried out on a sample pre-
pared by connecting to a vacuum line the NMR tube containing
compound 1 dissolved in C6D6 (for locking purpose) and condens-
ing therein the gaseous solvents CHF2Cl and CHFCl2 (in a 4:1
proportion) by means of liquid nitrogen. The tube was subse-
quently sealed in vacuo and introduced into the precooled probe
of the spectrometer. The temperatures were calibrated by sub-
stituting the sample with a precision Cu/Ni thermocouple before
the measurements. The 13C NMR solid-state CP-MAS measure-
ments have been obtained at 75.45 MHz. The sample was tightly
packed in a 7 mm zirconia rotor and spun about the magic angle
at a speed of 3.8-4.2 kHz. The cross polarization was achieved
using a 1H-90 and a 13C-90 pulse of 4.5 µs, as required by the
Hartmann-Hahn conditions. A contact time of 2 ms, a recycle
delay of 10 s, and 2 K data points were used during the
acquisition, and the number of scans ranged from 512 to 2000.
The temperature in the rotor was allowed to stabilize for at least
15-20 min before the acquisition. The chemical shifts were
calibrated by replacement, with respect to the lower frequency
signal (29.4 ppm) of the adamantane.
F IGUR E 4. Plot of the reference molecule (red) and of its
enclosure shell (9 molecules, gray)
HP LC a n a lysis was performed at +20 °C on an enantiose-
lective column (5 µm), 250 mm × 4.6 mm i.d., UV detected at
254 nm, flow rate 1 mL/min (hexane/i-Pr2O 90:10), retention
times 16.16 and 17.82 min.
Ack n ow led gm en t. The authors thank Prof. R. K.
Harris, University of Durham, U.K. for reading the
manuscript and for useful comments. Financial support
from the University of Bologna (funds for selected
research topics 2001-2002 and RFO) and from MIUR-
COFIN 2001, Rome (national project “Stereoselection
in Organic Synthesis”) was received by L.L. and A.M.
Support from the University of Basilicata and from
MIUR-COFIN 2001, Rome (“New theoretical and ex-
perimental methods for the assignment of the molecular
configuration”) is acknowledged by D.C.
F IGURE 5. Computed relative energy of 1 in the crystalline
state (see text) as function of the Ph-S rotation angle (ω).
the possibility of a rotation process, thus indicating that
the naphthyl ring is firmly locked within the crystal
lattice, in agreement with the experimental NMR obser-
vation.
Su p p or tin g In for m a tion Ava ila ble: Synthetic methods,
crystallographic data, X-ray powder spectrum, and molecular
mechanics parameters for 1. This material is available free of
charge via the Internet at http://pubs.acs.org.
In conclusion, these results indicate that it is possible,
in suitable conditions, to interpret satisfactorily the
conformational processes occurring in the solid state by
means of an appropriate theoretical approach. The bar-
rier involved in the present case has been, in fact,
reproduced with an excellent accuracy.
J O049843P
(28) Karaulova, E. N.; Bardina, T. A.; Gal’pern, G. D.; Bobruiskaya,
T. S Neftekhimiya 1966, 6, 480; Chem. Abstr. 1966, 65, 10452g.
(29) Rebiere, F.; Samuel, O.; Ricard, L.; Kagan, H. B. J . Org. Chem.
1991, 56, 5991.
J . Org. Chem, Vol. 69, No. 10, 2004 3577