Conformational analysis: Crystallographic, NMR, and molecular mechanics studies of flexible sulfonic esters

Article abstract of DOI:10.1021/jo0260342

Two novel X-ray structures of the sulfonic ester derivatives 2-(6-iodo-1,3-benzodioxol-5-yl)ethyl 4-nitrobenzenesulfonate, 3, and 2-(6-iodo-1,3 -benzodioxol-5-yl)ethyl 4-methylbenzenesulfonate, 4, have been obtained in a study aimed at analyzing the struc

Full text of DOI:10.1021/jo0260342

Con for m a tion a l An a lysis: Cr ysta llogr a p h ic, NMR, a n d Molecu la r
Mech a n ics Stu d ies of F lexible Su lfon ic Ester s
Orde Q. Munro,* J ean M. McKenzie, Sandra D. Strydom, and David Gravestock
School of Chemical and Physical Sciences, University of Natal, Pietermaritzburg,
Private Bag X01, Scottsville 3209, South Africa
munroo@nu.ac.za
Received J une 10, 2002
Two novel X-ray structures of the sulfonic ester derivatives 2-(6-iodo-1,3-benzodioxol-5-yl)ethyl
4-nitrobenzenesulfonate, 3, and 2-(6-iodo-1,3-benzodioxol-5-yl)ethyl 4-methylbenzenesulfonate, 4,
have been obtained in a study aimed at analyzing the structures and conformations of sulfonic
ester derivatives that are routinely used in alkaloid syntheses. The crystal structure of 4 is highly
unusual, containing four independent molecules that belong to two distinct conformational types:
(1) a hairpin conformation (stabilized mainly by intramolecular π-stacking) and (2) a stepped
conformation (stabilized mainly by intermolecular π-stacking). Compound 3, on the other hand,
crystallizes exclusively as the hairpin conformer. New MM+ force field parameters for sulfonic
esters have been developed using the X-ray data, empirical rules, and DFT calculations to estimate
the bond dipole parameters. Grid searches of conformational space for 3 and 4 using MM methods
show that there are several gas-phase conformations within 5 kcal/mol of the global minimum and
that the lowest energy conformations (by ∼4.6 kcal/mol) are of the hairpin type. Analysis of the
MM conformational energies suggests that the dominant intramolecular interaction stabilizing the
hairpin conformations of 3 and 4 is van der Waals attraction. Moreover, the lattice energies for
packing the hairpin conformations of 3 and 4 are ∼4 kcal/mol more favorable than for the stepped
conformations. Various intermolecular interactions contribute to the complexity of the observed
crystal structures of 3 and 4, including electrostatic attraction between O and I atoms in neighboring
molecules. Langevin dynamics (LD) simulations at several temperatures (6.0 ns, friction coefficient
) 2.5 ps-1) indicate that the conformational exchange rates are ∼1010-1011 s-1 over the temperature
1
range 213-400 K, accounting for the temperature-independent H NMR spectra of 3 and 4.
In tr od u ction
limited5 conformational studies by NMR spectroscopy of
1 and 2 (with R′ ) sugar). This situation is also reflected
by the lack of parameters to describe sulfonic esters in
established molecular mechanics (MM) force fields such
as MM26 and MM3.7 The conformational behavior and
intra/intermolecular interactions between these ubiqui-
tous groups and other structural components have thus
largely been overlooked until now. In this paper, we have
attempted to delineate the structures and conformational
p-Nosylates 1 and their methyl analogues (p-tosylates
2) are well-known sulfonic ester derivatives that have
been widely used as coupling reagents in a range of
synthetic methodologies due to their inherent lability as
leaving groups. The utility of 1 and 2 is exemplified by
the many natural product,1 alkaloid,2 and calyx(4)arene3
syntheses that have, as a key step, relied on the use of
sulfonic esters.
Despite the important role of sulfonic esters in both
(4) For example: (a) Gabriels, S.; Vanhaver, D.; Vandewalle, M.;
Declercq, P.; Viterbo, D. Eur. J . Org. Chem. 1999, 1803, 3-1809. (b)
Barili, P. L.; Catelani, G.; Dandrea, F.; Derensis, F.; Falcini, P.
Carbohydr. Res. 1997, 298, 75-84. (c) Chen, Z.; Fei, S.; Strauss, H. L.
J . Phys. Chem. A 1997, 101, 1640-1645. (d) Kirby, A. J .; Parker, J .
K.; Raithby, P. R. Acta Crystallogr., Sect. C 1992, 48, 832-834. (e)
Koch, K. R.; Niven, M. L.; Sacht, C. J . Coord. Chem. 1992, 26, 161-
176. (f) Urata, H.; Miyagoshi, H.; Yumoto, T.; Akagi, M. J . Chem. Soc.,
Perkin Trans. 1 1999, 1833-1838. (g) Kanagasabapathy, V. M.;
Sawyer, J . F.; Tidwell, T. T. J . Org. Chem. 1985, 50, 503.
synthesis and applied work,4 there have been only
(1) (a) Kitaori, K.; Furukawa, Y.; Yoshimoto, H.; Otera, J . Tetrahe-
dron 1999, 55, 14381-14390. (b) Kitaori, K.; Furukawa, Y.; Yoshimoto,
H.; Otera, J . Tetrahedron Lett. 1998, 39, 3173-3176. (c) Marshall, J .
A.; Crute, T. D.; His, J . D. J . Org. Chem. 1992, 57, 115-123. (d) Sutton,
P. W.; Bradley, A.; Farras, J .; Romea, P.; Urpi, F.; Vilarrasa, J .
Tetrahedron 2000, 56, 7947-7958.
(2) (a) Nowak, W.; Gerlach, H. Liebigs Ann. Chem. 1993, 153-159.
(b) Pearson, W. H.; Bergmeier, S. C.; Williams, J . P. J . Org. Chem.
1992, 57, 3977-3987. (c) Denmark, S. E.; Cottell, J . J . J . Org. Chem.
2001, 66, 4276-4284.
(3) (a) Ferguson, G.; Gallagher, J . F.; Lough, A. J .; Notti, A.;
Pappalardo, S.; Parisi, M. F. J . Org. Chem. 1999, 64, 5876-5885. (b)
Arduini, A.; Mcgregor, W. M.; Pochini, A.; Secchi, A.; Ugozzoli, F.;
Ungaro, R. J . Org. Chem. 1996, 61, 6881-6887. (c) Blanda, M. T.;
Farmer, D. B.; Brodbelt, J . S.; Goolsby, B. J . J . Am. Chem. Soc. 2000,
122, 1486-1491. (d) Burns, D.; H.; Chan, H. K.; Miller, J . D.; J ayne, C.
L.; Eichhorn, D. M. J . Org. Chem. 2000, 65, 5185-5196.
(5) J uaristi, E.; Antunez, S. Tetrahedron 1992, 48, 5941-5950.
(6) (a) Allinger, N. L. J . Am. Chem. Soc. 1977, 99, 8127-8134. (b)
Sprague, J . T.; Tai, J . C.; Yuh, Y.; Allinger, N. L. J . Comput. Chem.
1987, 8, 581-603. (c) Allinger, N. L.; Kok, R. A.; Imam, M. R. J .
Comput. Chem. 1988, 9, 591-595.
(7) (a) Allinger, N. L.; Yuh, Y. H.; Lii, J .-H. J . Am. Chem. Soc. 1989,
111, 8551-8566. (b) Lii, J .-H.; Allinger, N. L.; J . Am. Chem. Soc. 1989,
111, 8566-8575. (c) Lii, J .-H.; Allinger, N. L.; J . Am. Chem. Soc. 1989,
111, 8576-8582. (d) Allinger, N. L.; Li, F.; Yan, L. J . Comput. Chem.
1990, 11, 848-867. (e) Allinger, N. L.; Li, F.; Yan, L.; Tai, J . C. J .
Comput. Chem. 1990, 11, 868-895.
10.1021/jo0260342 CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/20/2003
2448
J . Org. Chem. 2003, 68, 2448-2459

Conformational Analysis: Flexible Sulfonic Esters
SCHEME 1
SCHEME 2
F IGURE 1. Schematic diagram illustrating the basic shapes
of the hairpin (π-stacked) and stepped (open-chain) conforma-
tions that are accessible to sulfonic esters 3 and 4.
evident for 3 represent an ideal and intriguing oppor-
tunity to develop and check a molecular mechanics force
field applicable to these (and related) systems. New force
field parameters (MM+) have therefore been developed
for sulfonic esters in general and, of specific interest in
this work, to study the structures and conformations of
3 and 4 by grid searches of conformational space and
Langevin dynamics (LD) techniques. Finally, lattice
calculations have been used to elucidate the role played
by crystal packing interactions in fine-tuning the solid-
state conformations of 3 and 4.
dynamics of the sulfonic ester derivatives 3 and 4, which
are readily available synthons for the 1,3-benzodioxole
group found in the skeleton of important alkaloids such
as cephalotaxine 58,9 and its ester derivative homohar-
ringtonine 6 (an anti-cancer drug),10 primarily to address
the paucity of conformational studies of 1 and 2 with
synthons that have at least some relevance in the general
fields of alkaloid and natural product synthesis (Scheme
1). Since our primary interest is in the structures and
conformations of tosylate and nosylate derivatives that
contain a 1,3-benzodioxole ring system, a straightforward
procedure similar to that reported by Semmelhack8a for
the nosylate derivative 3 has been used to synthesize 4
(Scheme 2). Furthermore, the structures of 3 and 4 have
been determined from single-crystal X-ray diffraction
data for the first time. The crystal structure of 4 is
remarkable in that four independent conformations are
observed in the asymmetric unit. More importantly, two
basic conformational types are evident: (1) a “hairpin”
conformation in which the tolyl group lies in van der
Waals contact with the 1,3-benzodioxole ring (intramo-
lecular aromatic stacking), the sulfonic ester backbone
therefore constituting the hairpin chain, and (2) a
“stepped” conformer with a more open-chain geometry
in which the two terminal ring systems are arranged like
a pair of stepping stones (Figure 1).
Resu lts a n d Discu ssion
Molecu la r Str u ctu r es of 3 a n d 4. Compounds 3 and
4 crystallize in the triclinic space groups P1h and P1,
respectively (Table 1). The unusual asymmetric unit of
tosylate 4 comprises four crystallographically unique
conformations (4A-4D) that belong to the two confor-
mational types illustrated in Figure 1. ORTEP diagrams
of one of the stepped and one of the hairpin conformations
of 4 are shown in Figure 2 (molecules A and D, respec-
tively). The asymmetric unit of the nosylate derivative
3, in contrast, comprises a single molecule with a hairpin
conformation that is similar in structure to 4D (Figure
2 and S1). Since 4B and 4C are approximate mirror
images of 4A and 4D, respectively (vide infra), their
structures are given in the Supporting Information
(Figure S2) along with full crystallographic details for
the X-ray structures of 3 and 4. Selected bond distances,
bond angles, and torsion angles for 3 and 4 are given in
Tables 2-4.
Since an important objective of this work was to
develop MM parameters for sulfonic esters in general,
we have averaged some of the more relevant chemically
unique bond distances, bond angles, and torsion angles
for 3 and 4. Indeed, these data are fundamental to
assessing the performance of any computational method
for these compounds and, moreover, provide a good
estimate of the range of values to be expected since these
particular sulfonic esters exhibit multiple conformations.
From Table 2, the mean O(sp2)-S(sp3), O(sp3)-S(sp3),
C(sp2)-S(sp3), and C(sp2)-I distances measure 1.42(3),
1.57(2), 1.75(2), and 2.09(7) Å, respectively. Similarly,
from Table 3 the average I-C(sp2)-C(sp2), C(sp3)-
The complex structure of 4 showing four independent
conformations and the distinctly different crystal packing
(8) (a) Semmelhack, M. F.; Chong, B. P.; J ones, L. D. J . Am. Chem.
Soc. 1972, 94, 8629. (b) Semmelhack, M. F.; Stauffer, R. D.; Rogerson,
T. D. Tetrahedron Lett. 1973, 45, 4519. (c) Semmelhack, M. F.; Chong,
B. P.; Stauffer, R. D.; Rogerson, T. D.; Chong, A.; J ones, L. D. J . Am.
Chem. Soc. 1975, 97, 2507.
(9) (a) Tietze, L. F.; Schirok, H. J . Am. Chem. Soc. 1999, 121, 10264-
10269. (b) Ikeda, M.; Elbialy, S. A. A.; Hirose, K.; Kotake, M.; Sato,
T.; Bayomi, S. M. M.; Shehata, I. A.; Abdelal, A. M.; Gad, L. M.; Yakura,
T. Chem. Pharm. Bull. 1999, 47, 983-987. (c) Robin, J . P.; Dhal, R.;
Dujardin, G.; Girodier, L.; Mevellec, L.; Poutot, S. Tetrahedron Lett.
1999, 40, 2931-2934. (d) Tietze, L. F.; Schirok, H. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1124-1125.
(10) Witte, R. S.; Lipsitz, S.; Goodman, T. L.; Asbury, R. F.; Wilding,
G.; Strnad, C. M.; Smith, T. J .; Haller, D. G. Invest. New Drugs 1999,
17, 173-177.
J . Org. Chem, Vol. 68, No. 6, 2003 2449

Munro et al.
TABLE 1. Cr ysta llogr a p h ic Da ta a n d Str u ctu r e Refin em en t Deta ils for 3 a n d 4a
4
3
empirical formula
formula wt
T (K)
C
16H15IO5S
C15H12INO7S
477.22
293(2)
446.24
293(2)
wavelength (Å)
space group
a, b, c (Å)
R, â, γ (deg)
V (Å3)
0.71073
P1
7.816(2), 9.818(2), 23.165(4)
82.67(1), 85.47(2), 73.27(2)
0.71073
P1h
7.363(2), 9.482(2), 13.300(8)
76.89(3), 73.99(4), 78.14(2)
859.2(6)
1686.7(6)
Z
4
2
Dc (g cm-3
)
)
1.757
2.044
1.845
2.023
3018; Rint ) 0.0101
2891
µabs (mm-1
unique reflns
data with I > 2σ(I)
goodness-of-fit on F2
R indices (I > 2σ(I))
R indices (all data)
6672; Rint ) 0.0075
6204
1.183b
1.117
R1 ) 0.0394; wR2 ) 0.1304
R1 ) 0.0474; wR2 ) 0.1673
R1 ) 0.0408; wR2 ) 0.1002
R1 ) 0.0422; wR2 ) 0.1010
a
b
The esd’s of the least significant digits are given in parentheses. Absolute structure parameter ) 0.26(4).11
TABLE 2. Selected Cr ysta llogr a p h ic Bon d Dista n ces (Å)
for Com p ou n d s 3 a n d 4a
Tosylate 4
2.136(17) S(1a)-O(4a)
I(1a)-C(1a)
S(1a)-O(5a)
S(1a)-C(10a)
I(1b)-C(1b)
S(1b)-O(4b)
S(1b)-O(3b)
I(1c)-C(1c)
S(1c)-O(4c)
S(1c)-C(10c)
I(1d)-C(1d)
S(1d)-O(4d)
S(1d)-O(3d)
1.416(16)
1.576(11)
1.41(2)
1.412(17)
1.780(16)
1.55(2)
1.395(15)
1.606(13)
1.42(2)
1.467(14)
1.734(14)
2.03(2)
S(1a)-O(3a)
O(3a)-C(9a)
S(1b)-O(5b)
S(1b)-C(10b)
O(3b)-C(9b)
S(1c)-O(5c)
S(1c)-O(3c)
O(3c)-C(9c)
O(3d)-C(9D)
S(1d)-O(5d)
S(1d)-C(10d)
1.358(15)
1.580(14)
2.025(19)
1.419(12)
1.747(18)
2.173(15)
1.418(15)
1.552(12)
1.50(2)
1.441(15)
1.722(16)
Nosylate 3b
I-C(1)
S-O(4)
S-C(10)
2.108(4)
1.421(4)
1.772(4)
S-O(5)
1.418(4)
1.558(3)
1.467(6)
S-O(3)
O(3)-C(9)
a
The esd’s of the least significant digits are given in parenthe-
ses. The letters a-d designate atoms belonging to molecules A-D,
b
respectively. The atom numbering for compound 3 is equivalent
to that for molecule D of 4 for the listed atoms (cf. Figure 2).
The most important torsion angles for the X-ray
conformations of 3 and 4 are those involving the flexible
sulfonic ester chain that links the 1,3-benzodioxole and
benzene ring systems (Table 4). The mean absolute
values of the five torsion angles C(10)-S(1)-O(3)-C(9),
S(1)-O(3)-C(9)-C(8), C(6)-C(8)-C(9)-O(3), C(1)-C(6)-
C(8)-C(9), and O(3)-S(1)-C(10)-C(11) for the two
stepped conformations 4A and 4B measure 65(3),
171(1), 72(6), 78(5), and 86(14)°, respectively. These same
angles average 73(9), 146(2), 66(2), 78(6), and 86(14)°,
respectively, for the three hairpin conformers 4C, 4D, and
3. As noted above, 4A and 4B are approximate mirror
images of one another. This is clearly evident from the
reversal in signs of the torsion angles listed in Table 4.
The same approximate mirror image relationship exists
between 4C and 4D.
F IGURE 2. Selectively labeled ORTEP diagrams (ellipsoids
are drawn at the 30% probability level) of crystallographically
unique molecules A (a) and D (b) within the asymmetric unit
of 4. (c) Overlay of the X-ray conformation of 3 and 4D. The
diagram depicts a least-squares fit of the two 1,3-benzodioxole
rings (rmsd ) 0.083 Å).
Conformations 4A and 4B have distinct stepped ge-
ometries in which the 1,3-benzodioxole ring and tolyl
group planes are near parallel and arranged in stepping
stone fashion about the sulfonic ester backbone. The
dihedral angles between the planes containing the two
ring systems measure 6.7° and 10.3° for 4A and 4B,
respectively. The 1,3-benzodioxole ring is essentially
planar in both stepped conformations; the mean absolute
O(sp3)-S(sp3), O(sp3)-S(sp3)-O(sp2), O(sp3)-S(sp3)-
C(sp2), O(sp2)-S(sp3)-O(sp2), and O(sp2)-S(sp3)-C(sp2)
angles measure 119(4), 117(3), 107(3), 104(1), 120(1), and
109(1)°, respectively. These mean bond distances and
angles have been used as a starting point in the deriva-
tion and optimization of an MM+ parameter set for this
class of compounds (vide infra).
2450 J . Org. Chem., Vol. 68, No. 6, 2003

Conformational Analysis: Flexible Sulfonic Esters
TABLE 3. Selected Cr ysta llogr a p h ic Bon d An gles (d eg)
for Com p ou n d s 3 a n d 4a
with 4, the 1,3-benzodioxole ring of 3 is essentially planar
(rms deviation ) 0.009 Å for all 9 ring atoms) with the
largest perpendicular displacement measuring 0.016 Å
for C(7). The phenyl group of the nosylate moiety is
effectively planar (rms deviation ) 0.004 Å). The dihedral
angle between the nitro group, which is exactly planar,
and the nosylate phenyl group is 2.1°.14
Tosylate 4
O(4a)-S(1a)-O(5a) 121.6(11) O(4a)-S(1a)-O(3a) 109.1(8)
O(5a)-S(1a)-O(3a) 103.7(8) O(4a)-S(1a)-C(10a) 107.1(8)
O(5a)-S(1a)-C(10a) 109.0(7) O(3a)-S(1a)-C(10a) 105.2(7)
C(9a)-O(3a)-S(1a) 119.0(10) C(6a)-C(1a)-I(1a)
C(2a)-C(1a)-I(1a) 112.7(13) C(6b)-C(1b)-I(1b)
120.6(12)
121.2(14)
O(4b)-S(1b)-O(5b) 118.3(11) O(4b)-S(1b)-O(3b) 102.4(8)
O(5b)-S(1b)-O(3b) 109.8(8) O(4b)-S(1b)-C(10b) 112.7(9)
O(5b)-S(1b)-C(10b) 109.2(9) O(3b)-S(1b)-C(10b) 103.1(7)
The dihedral angle between the mean plane of the 1,3-
benzodioxole ring and that of the benzene ring measures
6.3°, consistent with a near-parallel arrangement of the
ring systems. Interestingly, the two aromatic rings of 3
are aligned closer to parallel than is the case for 4C and
4D. The average separation of the phenyl ring atoms
from the mean plane of the 1,3-benzodioxole ring is
3.7(1) Å, consistent with a π-stacking interaction between
the two groups. The conformation of the sulfonic ester
backbone of 3 is similar to that observed for 4D (Table
4). However, the C(10)-S-O(3)-C(9) dihedral angle for
3 is 14° larger than that for 4D. This probably reflects
differences in the crystal packing of 3 and 4 (vide infra)
rather than conformational adjustments to minimize
intramolecular steric strain due to the p-CH3 group in 4
since the gas-phase MM-calculated conformations of 3
and 4D differ negligibly (Figure S4).
C(9b)-O(3b)-S(1b) 113.5(9) C(2b)-C(1b)-I(1b)
119.4(13)
104.1(8)
O(5c)-S(1c)-O(4c)
O(4c)-S(1c)-O(3c)
121.2(8) O(5c)-S(1c)-O(3c)
108.1(7) O(5c)-S(1c)-C(10c) 109.3(10)
O(4c)-S(1c)-C(10c) 110.1(8) O(3c)-S(1c)-C(10c) 102.3(7)
C(9c)-O(3c)-S(1c)
C(2c)-C(1c)-I(1c)
121.5(11) C(6c)-C(1c)-I(1c)
125.2(15)
118.5(12) C(6d)-C(1d)-I(1d) 117.3(11)
O(4d)-S(1d)-O(5d) 119.7(9) O(4d)-S(1d)-O(3d) 109.0(9)
O(5d)-S(1d)-O(3d) 103.8(7) O(4d)-S(1d)-C(10d) 109.1(9)
O(5d)-S(1d)-C(10d) 110.0(9) O(3d)-S(1d)-C(10d) 104.0(7)
C(9d)-O(3d)-S(1d) 114.2(11) C(2d)-C(1d)-I(1d) 114.7(13)
Nosylate 3
O(5)-S-O(4)
O(4)-S-O(3)
O(4)-S-C(10)
C(9)-O(3)-S
C(2)-C(1)-I
120.1(2)
104.1(2)
108.6(2)
118.4(3)
115.6(3)
O(5)-S-O(3)
O(5)-S-C(10)
O(3)-S-C(10)
C(6)-C(1)-I
110.9(2)
108.6(2)
103.12(18)
121.4(3)
a
The esd’s of the least significant digits are given in parenthe-
Cr ysta l Str u ctu r es of 3 a n d 4. Compounds 3 and 4
exhibit significant intermolecular π-stacking interactions
in the solid state. As might be expected, the different
space groups and asymmetric unit contents for 3 and 4
lead to different types of intermolecular interaction
depending on the conformation involved. For concision,
and because it straightforwardly illustrates the interest-
ing π-π intermolecular interactions in these systems,
Figure 3 shows only the unit cell packing diagram for
nosylate 3. (The crystal packing for tosylate 4 and a
discussion of the nonbonded interactions for this com-
pound may be found in the Supporting Information,
Figure S5.) Figure 3 depicts a view of the unit cell of 3
approximately parallel to the crystallographic [100]
direction. An immediately striking feature of the crystal
packing is the arrangement of the molecules into columns
of diagonally running hairpin units that are stacked in
a back-to-back fashion. There are a number of close
nonbonded contacts between the carbon atoms of the ring
systems of one hairpin unit and the neighboring units
in the columnar stack that are suggestive of intermo-
lecular π-π interactions.12,15 For example, contacts less
than 3.8 Å occurring between back-to-back 1,3-benzo-
dioxole rings include C(1)‚‚‚C(5)′ (3.678(2) Å), C(4)‚‚‚C(6)′
(3.647(2) Å), C(4)‚‚‚C(8)′ (3.680(2) Å), and C(5)‚‚‚C(6)′
(3.633(2) Å). (The atom-numbering scheme for 3 is
identical to that for 4D in Figure 2.) Furthermore, the
dihedral angle between the stacked 1,3-benzodioxole
rings is exactly zero, consistent with the fact that the
crystallographic center of inversion is located between
ses. The letters a-d designate atoms belonging to molecules A-D,
b
respectively. The atom numbering for compound 3 is equivalent
to that for molecule D of 4 for the listed atoms (cf. Figure 2).
perpendicular deviation of the 9 ring atoms from the
mean plane of the fused ring system is 0.04 Å, with O(1a)
and C(7b) displaying the largest individual displacements
(both 0.09 Å) in molecules A and B, respectively.
As shown in Figure 2, the hairpin conformation of
compound 4 differs markedly from the stepped conforma-
tion. More specifically, the bridging sulfonic ester chain
which links the 1,3-benzodioxole and tolyl rings in 4C
and 4D adopts a twisted conformation that favors
juxtaposition of the two aromatic ring systems and thus
leads to individual molecular conformations with well-
defined hairpin shapes (Figure 1). The dihedral angle
between the planes of the two ring systems in each case
measures 14.6° (4C) and 15.5° (4D). The mean inter-
atomic spacing (shortest contacts) between the atoms of
the tolyl ring and those of the 1,3-benzodioxole benzene
ring measures 3.6(2) Å for both 4C and 4D. This is
sufficiently small (and the two ring planes close enough
to parallel) to suggest the existence of a stabilizing
intramolecular aromatic stacking interaction between the
two six-membered rings in each case.12,13 We have in fact
confirmed this using PM3 calculations (Figure S3), which
indicate that a donor-acceptor π-π interaction between
the two ring systems is likely (the HOMO resides on the
1,3-benzodioxole ring and the LUMO on the tolyl ring).
As noted for 4A and 4B, the 1,3-benzodioxole rings of 4C
and 4D are essentially flat.
(14) The out-of-plane tilt of the nitro group is probably an under-
estimate since the anisotropic temperature factors for the oxygen atoms
indicate the possibility of two alternative orientations relative to the
plane of the phenyl ring, particularly for O(7), which is the oxygen
atom of the nitro group furthest from the reader’s view in Figure 2.
Use of a disorder model for two alternate C(14)-C(13)-N-O(7)
dihedral angles (i.e., O(6) and O(7) positions above and below the plane
of the aryl group) slightly lowered (0.02%) the R indices, but afforded
a poorer description of the geometry of the nitro group. We therefore
elected to model the electron density distribution for the nitro group
with a single conformation/orientation.
Nosylate 3 adopts a hairpin (π-stacked) conformation
with a similar geometry to 4D (Figure 2, Table 4). As
(11) Flack x-parameter: (a) Flack, H. D. Acta Crystallogr., Sect. A
1983, 39, 876-881. (b) Bernardinelli, G.; Flack, H. D. Acta Crystallogr.,
Sect. A 1985, 41, 500-511.
(12) (a) Dahl, T. Acta Chem. Scand. Ser. A 1975, 29, 170. (b)
Gavezzotti, A. Chem. Phys. Lett. 1989, 161, 67. (c) Gavezzotti, A.;
Desiraju, G. R. Acta Crystallogr., Sect. B 1988, 44, 427.
(13) Hunter, C. A.; Lawson, K. R.; Perkins, J .; Urch, C. J . J . Chem.
Soc., Perkin Trans. 2 2001, 651-669.
(15) Lokey, R. S.; Iverson, B. L. Nature 1995, 375, 303.
J . Org. Chem, Vol. 68, No. 6, 2003 2451

Munro et al.
TABLE 4. Selected Cr ysta llogr a p h ic Tor sion An gles (d eg) for Com p ou n d s 3 a n d 4a
Tosylate 4
C(10a)-S(1a)-O(3a)-C(9a)
S(1a)-O(3a)-C(9a)-C(8a)
C(6a)-C(8a)-C(9a)-O(3a)
C(1a)-C(6a)-C(8a)-C(9a)
O(3a)-S(1a)-C(10a)-C(11a)
C(10c)-S(1c)-O(3c)-C(9c)
S(1c)-O(3c)-C(9c)-C(8c)
C(6c)-C(8c)-C(9c)-O(3c)
C(1c)-C(6c)-C(8c)-C(9c)
O(3c)-S(1c)-C(10c)-C(11c)
-67(1)
-170(1)
68(2)
C(10b)-S(1b)-O(3b)-C(9b)
S(1b)-O(3b)-C(9b)-C(8b)
C(6b)-C(8b)-C(9b)-O(3b)
C(1b)-C(6b)-C(8b)-C(9b)
O(3b)-S(1b)-C(10b)-C(11b)
C(10d)-S(1d)-O(3d)-C(9d)
S(1d)-O(3d)-C(9d)-C(8d)
C(6d)-C(8d)-C(9d)-O(3d)
C(1d)-C(6d)-C(8d)-C(9d)
O(3d)-S(1d)-C(10d)-C(11d)
63(1)
172(1)
-76(2)
-74(2)
76(2)
81(2)
96(2)
-68(1)
147(1)
-63(2)
-84(2)
102(2)
68(1)
-148(1)
67(2)
72(2)
77(1)
Nosylate 3
C(10)-S-O(3)-C(9)
C(6)-C(8)-C(9)-O(3)
O(3)-S-C(10)-C(11)
82.8(3)
66.9(5)
77.8(4)
S-O(3)-C(9)-C(8)
C(1)-C(6)-C(8)-C(9)
-143.9(3)
79.4(6)
a
The esd’s of the least significant digits are given in parentheses. The letters a-d designate atoms belonging to molecules A-D,
b
respectively. The atom numbering for compound 3 is equivalent to that for molecule D of 4 for the listed atoms (cf. Figure 2).
barriers between the hairpin and stepped conformations
large enough to allow experimental observation of more
than one conformation in solution, e.g., by 1H NMR spec-
troscopy? The remainder of this paper seeks to answer
these questions using a combination of variable-temper-
ature NOE experiments and molecular simulations (MM
and LD) to fully elucidate the conformational behavior
of this class of sulfonic esters.
Va r ia ble-Tem p er a tu r e 1H NMR Stu d ies. In gen-
eral, variable-temperature NMR spectroscopy can be
used to delineate conformational exchange processes
when the barrier between two conformations (the free
energy of activation ∆Gq) is of the order of 5-24 kcal/
mol and, perhaps just as importantly, when the confor-
F IGURE 3. Stereoview of the unit cell and closest four
neighboring molecules for nosylate 3. The molecules are
stacked to form diagonal columns (parallel to the cell diagonal
running from [001] to [010]) in which the hairpin conformers
are slightly offset from the column center and adopt a zigzag
arrangement due to the crystallographically required inversion
symmetry.
mational isomers are individually distinguishable due to
intrinsically different chemical shifts for one or more
protons (or other nuclei) in the molecule.16b,17,18 We found
1
that the H NMR spectra of 3 and 4 recorded at -60 °C
in CDCl3 or -95 °C in acetone-d6 were equivalent to those
recorded at 24 °C (spectra not shown). Moreover, the 1D
ROESY spectrum of 4 recorded at -95 °C (irradiation at
5.95 ppm, O-CH2-O protons) in acetone-d6 showed no
NOE enhancements from the tolyl ring protons at 7.29
and 7.71 ppm, consistent with the fact that the hairpin
and stepped conformations are still in fast exchange on
the NMR time scale even at -95 °C. This suggests that
the barriers between the stepped and hairpin conforma-
tions of 3 and 4 are <5 kcal/mol in magnitude or that
the two conformations are spectroscopically indistin-
guishable even in the slow exchange limit. The former
conclusion is consistent with the detailed conformational
energy maps obtained for 3 and 4 by MM and LD
methods (vide infra). Moreover, from the number of
crystallographically observed conformations of 4 and the
fast conformational exchange rate (k > 103 s-1) implied
each pair of rings in the columnar stack. A similar picture
prevails for the back-to-back pairs of nitrobenzene rings.
Close nonbonded contacts consistent with π-π/aromatic
stacking interactions include C(10)‚‚‚C(11)′ (3.783(2) Å),
C(10)‚‚‚C(12)′ (3.666(2) Å), C(11)‚‚‚C(11)′ (3.731(2) Å), and
C(11)‚‚‚O(7)′ (3.369(1) Å).
The crystal structures of 3 and 4 are the first reported
X-ray structures of this class of compounds (sulfonic ester
derivatives containing 1,3-benzodioxole and tolyl ring
systems). The fact that multiple conformations are
observed in the solid state (compound 4) suggests that
these molecules are flexible and thus readily adopt a
range of conformations at little energy cost. For such a
system, the potential energy surface would likely encom-
pass a number of accessible energy minima with rela-
tively low interconnecting barriers. Multiple conforma-
tions would presumably also be populated in solution, as
is the case for other systems capable of intramolecular
aromatic stacking.13,16 The intriguing crystal structure
of 4 raises several questions. First, how many low-energy
conformations are possible and what are the reaction
coordinates (conformational pathways) that connect the
crystallographically observed structures? Second, what
are the energy differences between the hairpin and
stepped conformations of 4? Finally, are the energy
(16) For example: (a) Pang, Y.-P.; Miller, J . L.; Kollman, P. A. J .
Am. Chem. Soc. 1999, 41, 1717-1725. (b) Rashkin, M. J .; Waters, M.
L. J . Am. Chem. Soc. 2002, 124, 1860-1861. (c) Martin, C. B.; Mulla,
H. R.; Willis, P. G.; Cammers-Goodwin, A. J . Org. Chem. 1999, 64,
7802-7806. (d) Newcomb, L. F.; Gellman, S. H. J . Am. Chem. Soc.
1994, 116, 4993-4994. (e) Newcomb, L. F.; Haque, T. S.; Gellman, S.
H. J . Am. Chem. Soc. 1995, 117, 6509-6519. (f) Sindkhedkar, M. D.;
Mulla, H. R.; Cammers-Goodwin, A. J . Am. Chem. Soc. 2000, 122,
9271-9277.
(17) Harris, R. K. Nuclear Magnetic Resonance Spectroscopy; Long-
man Scientific & Technical: Avon, 1986.
(18) Stærk, D.; Norrby, P.-O.; J aroszewski, J . W. J . Org. Chem. 2001,
66, 2217-2221.
2452 J . Org. Chem., Vol. 68, No. 6, 2003

Conformational Analysis: Flexible Sulfonic Esters
by the variable temperature NMR studies, 3 and 4 are
intrinsically highly flexible compounds.
In Va cu o a n d Solid -Sta te MM Ca lcu la tion s. The
objective of this study was 3-fold: (1) to develop ap-
propriate parameters to calculate the structures of sul-
fonic esters with the MM+/MM2 force field, (2) to map
the full conformational space available to the two esters
3 and 4, and (3) to use the force field as a basis for an
extensive Langevin dynamics study to determine the
rates of conformational exchange for these compounds.
The lack of parameters for sulfonic esters in the MM+
force field necessitated the development of new bond
stretching, angle bending, and torsion angle parameters
for this class of compounds from the available crystal-
lographic data. Two types of geometry optimization were
performed on the crystallographically observed confor-
mations of 3 and 4 to verify the performance of the
augmented MM+ force field: (1) a gas-phase geometry
optimization and (2) a separate refinement of each
conformation within its lattice environment, i.e., includ-
ing all neighboring molecule interactions. Selected mean
bond lengths, bond angles, and torsion angles for the
calculated and observed structures of 4 are compared in
Table S7; the level of agreement between the calculated
gas phase structural parameters and the X-ray data is
good and clearly improves upon inclusion of all neighbor-
ing interactions in the solid state (lattice calculation).19
These results are portrayed graphically in Figure 4,
which shows least-squares fits of the MM calculated and
crystallographic conformations of 3 and 4. The average
rmsd for the conformations calculated in the gas phase,
0.28(5) Å, is roughly twice as large as that calculated for
the conformations within their crystal lattice environ-
ment (0.13(3) Å). This suggests that crystal-packing
interactions modulate the conformations of 3 and 4 to
some extent, as already noted (Figure 3).
The absolute energies of the hairpin and stepped
conformations of 3 and 4 are given in Table 5.20 For the
tosylate derivative, gas-phase conformations 4A and 4B
(stepped geometry) are exact mirror images of one
another and thus have equivalent strain energies; the
same holds true for 4C and 4D (hairpin geometry). From
Table 5, the hairpin conformation of 4 is more stable than
the stepped conformation by 4.29 kcal/mol (gas phase)
and 3.7 kcal/mol (solid state). As expected, the solid-state
conformations are somewhat more strained than their
gas phase counterparts. However, the energy differences
amount to <1.2 kcal/mol, consistent with the relatively
minor conformational adjustments that are required for
optimal packing in the crystal lattice (Figures 3 and 4).
The nosylate 3 shows a similar set of energy changes with
the hairpin conformation being 4.60 kcal/mol more stable
than the stepped conformation in the gas phase. Inspec-
tion of the packing energies for the four crystallographi-
cally unique conformations of 4 indicates that the two
hairpin conformers pack more favorably than the two
stepped conformers (by up to ∼4.5 kcal/mol). A question
F IGURE 4. Least-squares fits of selected MM calculated
(blue) and X-ray (red) conformations of 3 and 4. Diagrams (a)
and (b) are fits for molecules A and D, respectively, of 4.
Diagram (c) shows the fits for the hairpin conformation of 3.
The rmsd values are based on all non-H atoms for both gas
phase (left) and crystal lattice geometry optimizations (right).
TABLE 5. Ca lcu la ted En er gies (MM) for
Geom etr y-Op tim ized Con for m a tion s of 3 a n d 4a
c
e
UT
packing
UT
packing
b
conformer
4A
4B
4C
UT (g) (lattice, IA) straind (lattice) energyf
5.64
5.64
1.35
1.35
5.15
6.17
6.25
2.50
2.51
6.00
0.53
0.61
1.15
1.16
0.85
-46.64 -52.81
-47.17 -53.42
-53.88 -56.39
-54.69 -57.20
-51.63 -57.63
4D
3 hairpin “D”g
3 stepped “A”h 9.75
a
b
All energies are in kcal/mol. UT (g) is the total steric energy
of the refined gas phase conformation. c UT (lattice, IA) is the
single-point energy of the conformation refined in its crystal lattice
site after removal of all neighboring molecules. The value of UT
(lattice, IA) is a measure of the intramolecular steric energy of a
d
given conformation in the crystal lattice. This term is defined
as UT (lattice, IA) - UT (g) and gauges the increase in intramo-
lecular strain energy associated with packing the conformation
in its lattice site. e UT (lattice) is the total steric energy of a given
conformer in its crystal lattice site (intra- plus intermolecular steric
energy). f The packing energy is defined as UT (lattice) - UT
(lattice, IA). This term roughly measures the thermodynamic
driving force that leads to crystallization of a given conformation.
g
h
Conformation with a similar geometry to 4D. Conformation
with a similar geometry to 4A.
that we have yet to unambiguously resolve is why
replacement of the p-NO2 group of 3 with a p-CH3 group
in 4 leads to a significantly more complicated unit cell
comprising an equal ratio of hairpin and stepped confor-
mations. This conformational ratio is unexpected given
the higher steric energy of the two stepped conformations
of 4 (Table 5) and the conformer distribution observed
(19) Munro, O. Q.; Madlala, P. S.; Warby, R. A. F.; Seda, T. S.;
Hearne, G. Inorg. Chem. 1999, 38, 4724-4736.
(20) The absolute energies of each conformation are force field
dependent. In cases where parameters have been developed to fit
experimental heats of formation, the absolute energy of the system
may have some physical significance. A more important parameter
therefore is the relative energy of two conformations since this suffers
less from the choice of parameters used in the force field.
J . Org. Chem, Vol. 68, No. 6, 2003 2453

Munro et al.
TABLE 6. Coor d in a tes a n d Rela tive Ster ic En er gies of
th e F u lly Op tim ized Sym m etr y Un iqu e Min im a (∆UT e 5
k ca l/m ol) of 3 a n d 4a
in the LD simulations of this compound (Table 7, vide
infra). Rather than being purely due to chance, we believe
that part of the explanation for the unique solid-state
structure of 4 is that the p-CH3 group is involved in (1)
fewer and (2) weaker intermolecular interactions than
the p-NO2 group of 3. More specifically, as shown in the
packing diagram of Figure 3 (and in somewhat more
detail in Figure S6), the p-NO2 group of 3 forms no less
than three tight contacts with neighboring atoms in the
crystal lattice. Of particular interest are the short,
electrostatically favored nitro-O‚‚‚I′ and nitro-N‚‚‚O′ (sul-
fonic ester) interactions, which measure 3.262 and 3.047
Å, respectively. (The primed atoms are atoms belonging
to neighboring molecules in the crystal lattice.) These
stabilizing interactions are in fact absent in the tosylate
derivative (Figure S7). Their importance as a contributor
to the overall lattice energy is apparent if the p-NO2
group of 3 is replaced with a p-CH3 group and the packing
energy for the hairpin conformation recalculated. With
the present force field, we find that the packing energy
for the hairpin conformer of 3 is 0.56 kcal/mol less
favorable with the p-CH3 group. Thus, in the case of 4,
which lacks a polar nitro group, cocrystallization of the
energetically less favorable stepped conformer may be
required to stabilize the lattice as a whole by facilitating
at least some favorable intermolecular electrostatic in-
teractions. Indeed, inspection of Figure S7 shows that
the main electrostatic interaction is between the I atom
of 4A (stepped conformer) and a 1,3-benzodioxole group
oxygen atom of 4D (hairpin geometry) at a distance of
3.332 Å. A similar interaction is observed between
molecules 4C and 4B (I‚‚‚O′ ) 3.447 Å), while the two
hairpin conformations 4C and 4D behave much like a
dimer with I‚‚‚O′ (sulfonic ester) interactions measuring
3.096 and 3.111 Å (Figure S8). The factors governing the
exact unit cell contents of 3 and 4 are evidently rather
complex and include both the electrostatic interactions
discussed above and π-π interactions between the ring
systems (vide supra). We therefore surmise that there is
probably an energy balance between a multitude of
intermolecular interactions of varying significance (π-
stacking, London forces, dipole-dipole/electrostatic, and
van der Waals interactions) and the strain energies of
the individual conformers that controls the observed
crystal packing in these compounds.
ψ1
ψ2
ψ3
ψ4
ψ5
∆UT
conformer
(deg) (deg) (deg) (deg) (deg) (kcal/mol)
4A
4D
4E
4F
4G
3A
3D
3E
3F
3G
295
77
177
62
179
295
77
176
63
183
218
184
178
178
183
219
185
178
178
61
65
58
179
181
61
66
57
179
181
86
85
87
87
87
86
84
87
88
87
99
88
95
86
93
98
88
95
86
91
4.29b
0.00
3.09b
3.55b
3.50b
4.60c
0.00
3.35c
3.84c
3.80c
179
a
The symmetry-related minima B (inverse of A), C (inverse of
b
D), E*, and F* have the coordinates 360 - ψn°. Steric energy
relative to 4D. c Steric energy relative to 3D.
the MP2 level of theory at the basis set limit for the
parallel stacked benzene dimer with a 3.8-Å ring-ring
separation.21 The success of the MM calculations in this
case may be attributed to the fact that dispersion and
nonbonded repulsion, which apparently dominate the
π-π interactions in this system (and the aromatic
sulfonic esters 3 and 4), are well accounted for by the
Lennard-J ones potential.
Con for m a tion a l An a lysis. To some extent, the data
in Table 5 account for the crystallographically observed
conformations of 3 and 4. However, a more complete
picture of the conformational space accessible to these
systems, particularly the saddle points connecting the
energy minima, may be obtained by a combination of
dihedral angle (grid) searches of conformational space
and MD or LD calculations at several different temper-
atures. Three of the five sulfonic ester chain torsion
angles of compound 4, ψ1-ψ3, were selected for confor-
mational analysis, although all five were monitored
(Table 6), since they constitute the likely pathway by
which the hairpin conformation interconverts to the
stepped conformation (and any others not observed in the
solid state). Figure 5 shows a plot of the change in steric
energy, ∆UT, as a function of the torsion angles ψ1 and
ψ3 for compound 4. Similar conformational energy plots
for the torsion angle combinations ψ2/ψ3 and ψ1/ψ2 of 4
are given in the Supporting Information (Figures S9 and
S10, respectively). The coordinates of the lowest energy
conformation(s) located from each of these 2D grid
searches of conformational space afforded the following
average values for the backbone dihedral angles for the
symmetry unique global minimum energy conformation
of 4: 75, 215, and 60° (ψ1, ψ2, and ψ3, respectively). These
dihedral angles are consistent with those of gas-phase
structure (b) in Figure 4 (i.e., the hairpin conformation
4D with ψ1, ψ2, ψ3 ) 77, 218, 65°).
Despite the respectable geometric data obtained with
the MM calculations (Figure 4), the question remains as
to how accurately the MM+ force field calculates the
intermolecular interaction energies in the above systems.
Since intermolecular π-stacking interactions are manifest
in the crystal structures of 3 and 4, we chose to test the
MM+ force field against the recent high-level ab initio
calculations on aromatic stacking interactions (benzene‚
‚‚benzene dimer) reported by Tsuzuki et al.21 The in vacuo
MM+ geometry optimized structure of the parallel dimer
for benzene (ring systems exactly eclipsed) had a mean
plane separation of 3.42 Å and interaction energy of
-3.69 kcal/mol. The interaction energy of the benzene
dimer with a 3.8-Å mean plane separation was -3.06
kcal/mol. These values compare favorably with the ab
initio interaction energy (-3.28 kcal/mol) calculated at
As expected, the conformational energy surface of
compound 4 (Figure 5) has a center of inversion at ψ1,
ψ3 ) 180, 180°. Accordingly, there are five symmetry
(21) Tsuzuki, S.; Honda, K.; Uchimaru, T.; Mikami, M.; Tanabe, K.
J . Am. Chem. Soc. 2002, 124, 104-112.
2454 J . Org. Chem., Vol. 68, No. 6, 2003

Conformational Analysis: Flexible Sulfonic Esters
F IGURE 6. Labeled structures of the symmetry unique low
energy conformations of 4 taken from the potential energy
surface shown in Figure 5.
than that of E. The final low energy structure, conforma-
tion G (ψ1, ψ3 ) 180, 180°; ∆UT ) 3.50 kcal/mol), is
located at the center of inversion in Figure 5. With both
ψ1 and ψ3 fully extended, the sulfonic ester chain of
conformation G has a staggered, or saw tooth, backbone
geometry. Although the 1,3-benzodioxole and tolyl rings
are slightly canted relative to one another (16.8°), the
structure is clearly a more extended version of stepped
conformations A and B. Importantly, Figure 5 confirms
that rotations about ψ1 and ψ3 are responsible for the
crystallographically observed hairpin and stepped con-
formations of 4 shown in Figures 2 and S2.23
The ψ2/ψ3 and ψ1/ψ2 conformational energy surfaces
for 4 are given in the Supporting Information (Figures
S9 and S10). The global minima from these surfaces are
the isoenergetic hairpin conformations C and D, respec-
tively. Moreover, no additional conformations differing
from those already located on the ψ1/ψ3 surface were
found. The dihedral angle coordinates of the symmetry
unique minimum energy conformations of 3 and 4 that
lie within 5 kcal/mol of the global minimum have been
summarized in Table 6. Interestingly, all of the listed
minima exhibit ψ4 and ψ5 values that fall within 10° of
90°. Indeed, plots of ψ5 vs ψ4 as a function of time from
the LD trajectories of 4 (vide infra, Figure S11) indicate
that ψ4 and ψ5 are confined to the values (90° (with a
spread of (30° in each case). Table 6 also shows that ψ2
is limited to a staggered configuration (180 ( 40°) for
those conformations that lie within 5 kcal/mol of the
global minimum. Stable gauche conformations of ψ2 are,
however, observed with ψ2 ) 70-75° (Figure S12). There
are, in fact, four local minimum gauche ψ2 conformers
F IGURE 5. Plots of the variation of ∆UT (kcal/mol) with ψ1
and ψ3 for tosylate 4. The upper plot is a 3D representation of
the potential surface; the lower plot is a 2D contour map
(contour interval ) 0.75 kcal/mol) of the surface with labels
for the global and local minimum energy conformations.
unique energy minima on the ψ1/ψ3 potential surface
(Figure 6). These conformations, marked A-D in Figure
5, are labeled so that they correspond to the X-ray
conformations A-D, both in structural type and in
designation. Because the surface is centrosymmetric,
stepped conformation A (∆UT ) 4.29 kcal/mol) is mirrored
at B and hairpin conformation D (∆UT ) 0 kcal/mol) is
mirrored at C. The local energy minimum E (ψ1, ψ3 )
177, 58°; ∆UT ) 3.09 kcal/mol) is a stable conformation
found along the pathway (reaction coordinate ψ1) between
hairpin conformation D (ψ1, ψ3 ) 77, 65°) and stepped
conformation A (ψ1, ψ3 ) 295, 61°).22 As might be
anticipated by its coordinates in conformational space
relative to A and D, the structure of E is best described
as a half-opened hairpin conformation in which the 1,3-
benzodioxole ring has swung away from the tolyl ring
and is no longer in close van der Waals contact with it.
Conformation F (ψ1, ψ3 ) 62, 179°; ∆UT ) 3.55 kcal/mol)
lies halfway along the reaction coordinate ψ3 between
hairpin conformation D and the inverted stepped con-
formation B (ψ1, ψ3 ) 65, 299°). The structure of F is also
an opened hairpin conformation. However, the two ring
systems are somewhat more splayed than those of E and
the relative steric energy of F is correspondingly higher
(23) It is important to note that the calculated conformations A and
D have isomers B and C which are related through a Ci symmetry
element (center of inversion) at ψ1, ψ3 ) 180, 180°. The crystallo-
graphically observed isomers of 4 do not display exact inversion
symmetry because it is not crystallographically required in the space
group P1.
(22) The symmetry-related counterpart, E*, is equal in energy to
conformation E and lies halfway between conformations B and C.
J . Org. Chem, Vol. 68, No. 6, 2003 2455

Munro et al.
that have ψ4 values of ca. (90° and (180°. However, all
lie >5 kcal/mol above the global minimum conformation
in energy. In the interests of brevity, these have not been
tabulated and are not discussed further. Finally, the
conformational surfaces for 3 exhibit similar shapes and
symmetries to those of 4 (see, for example, Figure S13).
However, as shown in Table 6, most of the stable
conformations for 3 have relative energies that are ∼0.3
kcal/mol higher than those of the tosylate derivative.
Furthermore, as might be expected, the barriers connect-
ing the energy minima for 3 are ∼0.3 kcal/mol larger than
those of 4.
La n gevin Dyn a m ics Sim u la tion s. The conforma-
tional energy maps obtained for 3 and 4 indicate (1) that
the barriers between the lowest energy structures are of
the order of 3-7.5 kcal/mol in the gas phase and (2) that
most of the stable minima are within 5 kcal/mol of one
another in energy. The system is evidently highly flexible
and this raises three interesting questions. First, what
are the relative populations of the minima at equilibri-
um? Second, how large are the rate constants for the
associated barrier crossings? Third, at what temperature
might conformational interconversions of 3 and 4 be slow
enough to allow measurement of the free energy of
activation by NMR spectroscopy? These questions may
be answered with equilibrium LD simulations because
conformational populations are obtained directly from the
trajectory and rate constants may be determined from
the barrier crossing frequencies at the saddle points of
interest for a given temperature.
F IGURE 7. Time-graded scatter plot of the conformational
distribution of 4 (ψ1/ψ3 reaction coordinate) at 300 K. The
distribution is from the equilibrated portion of a 6.015 ns
Langevin dynamics simulation of 4 (time step ) 1.0 fs, friction
coefficient ) 2.5 ps-1, bath relaxation constant ) 0.1 ps, data
reduction factor ) 1/100).
TABLE 7. F r a ction a l P op u la tion s for Con for m a tion s of
4 a s a F u n ction of Tem p er a tu r e
conformation
300 K
350 K
400 K
4A
4B
4C
0.0131
0.00243
0.361
0.0257
0.0272
0.207
0.00266
0.0131
0.327
We originally carried out MD simulations of 3 and 4
at 213, 298, and 328 K using an explicit solvent model
(25-Å3 solvent box, 113 CHCl3 molecules) in an attempt
to generate MD trajectories that would closely mimic the
behavior of these compounds in solution at NMR-acces-
sible temperatures. Contrary to expectation, we found
that the lowest energy conformations (C and D) were not
the most abundantly populated at 298 and 328 K.
Furthermore; only conformations D and E (and not their
mirror images C and E*) were significantly populated at
213 K. We suspect that the origin of this physically
unrealistic conformational distribution with increasing
temperature relates to a problem with the explicit solvent
model used, namely that the solute had to be restrained
to coordinates within the solvent box to avoid fatal system
instabilities that arose when the solute crossed the
periodic boundary. (The problem does not arise for small
solvent molecules, which correctly cross the periodic
boundary with a leaving molecule being instantaneously
replaced by its mirror image.) This artificial restraint on
the system apparently leads to a sizable damping effect
on the conformational exchange process. We therefore
chose to use an unrestrained, computationally more
efficient nonexplicit solvent model (Langevin method) for
all dynamics simulations on compounds 3 and 4 in an
effort to delineate the solution-phase dynamic behavior
of these compounds.
4D
4E
4F
4G
4E*
4F *
0.351
0.104
0.0159
0.0435
0.0681
0.0407
0.359
0.223
0.0855
0.0414
0.134
0.052
0.068
0.0819
0.0364
0.126
0.131
0.0590
in Table 7. (The LD trajectory for 4 at 400 K is shown in
Figure S14 for comparison.) Notwithstanding some sig-
nificant population dispersion effects with increasing
temperature, particularly between the symmetry-related
minima which should have comparable occupancies, we
find that the lowest-lying conformations D and C are
heavily populated at all temperatures (as expected) with
an increasing spread in the values of ψ1 and ψ3 clearly
evident at higher temperatures. This reflects the greater
kinetic energy and thus amplitude of torsional motion
within the system at higher temperatures (cf. Figure S14)
and, consequently, the higher barrier crossing frequencies
for all saddle points on the surface. The minima directly
adjacent to D and C (i.e., E and E*) are also well
populated (though less so than the global minima due to
their somewhat higher relative energies). The time-
graded scatter plot of Figure 7 is informative because it
shows that the system resides between these two minima
for intervals of up to ∼900 ps. This also explains the
population dispersion seen in Table 7sa 6-ns trajectory
is probably not quite long enough for a fully symmetric
distribution to evolve due to the length of time that the
system spends in the lowest energy states.
Figure 7 exemplifies the LD trajectory of 4 at 300 K
for the system at equilibrium (115-6015 ps) as a time-
graded scatter plot of ψ1 vs ψ3. The conformational
distribution is consistent with that expected from the
relative energies of the conformations 4A-4F * mapped
in Figure 5 and listed in Table 6. The fractional occupan-
cies of the ψ1/ψ3 minima at 300, 350, and 400 K are given
The conformational exchanges D T E and C T E*
exhibit the highest barrier crossing frequencies at all
temperatures. To calculate the transition state param-
eters for at least one main conformational exchange
2456 J . Org. Chem., Vol. 68, No. 6, 2003

Conformational Analysis: Flexible Sulfonic Esters
process in the system, e.g., D T E, we have made use of
the fact that the barrier crossing frequency is propor-
tional to the populations of the minima either side of the
saddle point. The net number of barrier crossings per unit
time can then be weighted according to the relative
populations of the minima involved at each temperature.
This gave the overall rate constants, koverall, 9.6 × 1010,
1.5 × 1011, and 2.6 × 1011 s-1 at 300, 350, and 400 K,
crystallographically independent conformations that fall
into two categories, namely, a hairpin conformer in which
the two ring systems form a tight intramolecular π-π
dimer and a stepped (open chain) conformer which shows
intermolecular π-π interactions with neighboring mol-
ecules in the crystal lattice. The nosylate 3, in contrast,
crystallizes in a higher symmetry space group exclusively
as the hairpin conformer. A parameter set for use with
the MM+ and MM2 force fields has been developed for
the first time for sulfonic esters using the X-ray data in
conjunction with empirical rules and natural atomic
charges to estimate bond dipoles. Molecular mechanics
and Langevin dynamics simulations with the modified
force field show that there are in fact at least five
symmetry-unique conformational types within 5 kcal/mol
of the global minimum for both systems but that the
hairpin conformers are the most stable by up to ∼4.6 kcal/
mol in the gas phase. Overall, 3 and 4 are highly flexible
molecules and exhibit conformational exchange rates of
the order of 1011 s-1sat least 6 orders of magnitude faster
than can be measured by dynamic 1H NMR spectroscopy.
respectively (where koverall ) kforward + kback and kforward
)
kback at equilibrium). Similar exchange rates have been
reported for the trans-gauche isomerization of butane in
the liquid phase.24 An Arrhenius plot (Figure S15) of the
overall rate constants for the transition D T E had a
slope of -1183 ( 143 K-1 and an intercept of 29.3 ( 0.4
(R ) -0.99). The activation energy for the conformational
switch D T E (2.3 ( 0.3 kcal/mol) is roughly 50% of the
gas-phase steric energy barrier (3.75-4.5 kcal/mol) com-
puted by the dihedral angle driving method for compound
4 in Figure 5. As reported by Kollman’s group for a
flexible bis(indole) derivative,16a solvation tends to facili-
tate conformational exchange in such systems by lower-
ing the activation energy barrier through the transfer of
kinetic energy to the solute. Extrapolation of the rate
constant to -60 °C using the Arrhenius parameters for
the system gives a D T E barrier crossing frequency of
1.9 × 1010 s-1. The rate constant is therefore ∼6 orders
of magnitude faster than can be determined by dynamic
1H NMR spectroscopy and helps explain the temperature
Exp er im en ta l Section
Gen er a l In for m a tion . Starting materials were obtained
from commercial suppliers and were used without further
purification, unless otherwise indicated. Radial chromatogra-
phy was carried out with Merck Silica gel 60 PF254. FTIR
spectra were recorded at a resolution of 1.0 cm-1 as KBr mulls
or neat liquids between NaCl plates. 1H and 13C NMR spectra
were recorded on a 500 MHz spectrometer (magnetic field )
11.744 T) using a switchable 5 mm probe. Standard 1H and
13C pulse sequences were used for 1D and 2D spectra. HRMS
data were obtained from the Mass Spectrometry Unit of the
Cape Technikon.
2-(1,3-Benzodioxol-5-yl)-1-ethanol, 8, and 2-(6-iodo-1,3-ben-
zodioxol-5-yl)-1-ethanol, 9, were synthesized by the standard
literature procedure of Semmelhack et al.8a 2-(6-Iodo-1,3-
benzodioxol-5-yl)ethyl 4-nitrobenzenesulfonate, 3, was syn-
thesized from 9 using Tietze’s nosylation procedure.25
2-(6-Iod o-1,3-ben zod ioxol-5-yl)eth yl 4-m eth ylben zen e-
su lfon a te, 4. 2-(6-Iodo-1,3-benzodioxol-5-yl)-1-ethanol 9 (0.512
g, 1.75 mmol) and p-toluenesulfonyl chloride (0.50 g, 2.6 mmol)
were dissolved separately in a minimal amount of diethyl ether
(distilled from sodium benzophenone) and combined. Pyridine
(0.5 mL, dried over KOH and distilled) was added to the
solution and the mixture left overnight to allow the formation
of colorless and yellow crystals. The supernatant liquid was
then removed and left to stand allowing more crystals to form.
The combined batches of crystals and remaining liquid were
partitioned between dichloromethane and water (2:1). The
organic layer was separated, washed with ice-cold HCl (5%,
30 mL), dried over MgSO4, filtered, and concentrated in vacuo
leaving a yellow residue which was purified by radial chro-
matography (15% ethyl acetate/hexane) giving an oil (0.315
g, 39%). Compound 4 crystallized from the oil on standing.
Colorless crystalline product: mp 50-54 °C; νmax (KBr)/cm-1
2900-3000 (CH), 1358 (νa S(dO)2) and 1176 (νs S(dO)2), 1248
and 1096 (C-O), 554 (C-I); δH (500 MHz, CDCl3) 2.44 (3H, s,
Me), 2.98 (2H, t, J 6.9, ArCH2), 4.16 (2H, t, J 6.9, CH2OTs),
5.95 (2H, s, OCH2O), 6.67 (1H, s, CHCCH2), 7.13 (1H, s, CHCI),
7.29 (2H, d, J 8.2, CHCMe), 7.71 (2H, d, J 8.2, CHCS); δC (125
MHz, CDCl3) 21.61 (Me), 39.81 (ArCH2), 69.10 (CH2OTs), 87.78
(CI), 101.67 (OCH2O), 110.42 (CHCCH2), 118.64 (CHCI),
127.87 (CHCS), 129.74 (CHCMe), 131.97 (CCH2), 132.86
(CMe), 144.68 (CS), 147.52 (CCHCCH2), 148.45 (CCHCI); m/z
1
invariant H NMR spectra obtained for 4, which clearly
remains in the fast exchange limit at the temperatures
studied.
The remaining conformations A, B, F, F*, and G have
temperature-dependent populations that are consistent
with their relative energies (Tables 6). The conforma-
tional populations obtained from the LD trajectories are
useful in a second respect, namely that estimates of the
free energy differences between pairs of minima on the
potential energy surface are possible. For example, ∆G
values for the E f D and E* f C transitions at 300 K
measure -0.73 and -0.99 kcal/mol, respectively, consis-
tent with the lower steric energies for the two hairpin
conformers and thus their significantly higher popula-
tions. If we assume that the ∆UT value for this type of
conformational transition listed in Table 6 approximates
∆H for the process, i.e., -3.09 kcal/mol, then it is clear
that the change in entropy is negative for folding the half-
open conformation into the more stable hairpin structure.
(Using an average ∆G value of -0.86 kcal/mol for the
conformational transition, we calculate that ∆S will be
-7.4 cal K-1 mol-1 or -31 J K-1 mol-1.) Finally, the
dynamic behavior of the nosylate 3 is comparable to that
of 4 due to the similarity of the potential energy surfaces
of these two sulfonic esters.
Con clu sion s
Two novel X-ray structures of the sulfonic ester deriva-
tives 2-(6-iodo-1,3-benzodioxol-5-yl)ethyl 4-methylben-
zenesulfonate, 4, and the corresponding nosylate 2-(6-
iodo-1,3-benzodioxol-5-yl)ethyl 4-nitrobenzenesulfonate,
3, have been obtained. The tosylate 4 exhibits four
(24) (a) Edberg, R.; Evans, D. J .; Morriss, G. P. J . Chem. Phys. 1987,
87, 5700-5708. (b) Brown, D.; Clarke, J . H. R. J . Chem. Phys. 1990,
92, 3062-3073.
(25) Tietze, L. F.; Schirok, H.; Wohrmann, M.; Schrader, K. Eur. J .
Org. Chem. 2000, 2433-2444.
J . Org. Chem, Vol. 68, No. 6, 2003 2457

Munro et al.
δA δB
µD
)
×
(1.602 × 10-19)(dAB × 10-10)/(3.336 ×
xn
m
10-30) (1)
where δA and δB are the partial atomic charges on atoms A
and B, respectively, n and m are the number of bonds to atoms
A and B, respectively, and dAB is the distance between atoms
A and B (in Å). The factor 3.336 × 10-30 converts the dipole
units from C m to debye. The sign of µD is, by definition,
positive for the bond A-B when atom B is the negative end of
the dipole. Bond dipoles derived with eq 1 are given below as
part of the new MM+ force field parameter set for sulfonic
esters (vide infra). The natural atomic charges and selected
interatomic distances determined at the B3LYP/6-31G* level
of theory for methyl benzenesulfonate are given in Figure 8
for completeness. The utility of eq 1 for garnering estimates
of bond dipoles for standard S-containing (or indeed any other)
functional groups was tested for several simple thioethers,
sulfoxides, and sulfones; the values were found to be in
agreement with those listed for these functional groups in both
the MM2 and MM3 force fields (Supporting Information).
Molecu la r Mech a n ics Ca lcu la tion s. HyperChem 6.0333
(MM+ force field) was used for all calculations. A root-mean-
square gradient termination cutoff of 0.01 kcal Å-1 mol-1 was
used for geometry optimization with the Polak-Ribiere con-
jugate gradient algorithm (dielectric constant ) 1.5 D). New
bond stretching, angle bending, and dihedral angle parameters
were developed for the MM+ force field to specifically handle
sulfonic esters. Initial estimates of strain free bond lengths,
R0, and angles, θ0, were obtained by averaging the crystal-
lographic values from the conformations of 3 and 4. Bond
stretching force constants, ks, were derived using Allinger’s
method.34 Bond dipoles were estimated using eq 1 (vide supra).
Angle bending force constants, kθ, and torsional parameters,
Vn (n ) 1-3), were estimated from existing MM+ parameters
for similar atom sequences. The new parameters were adjusted
iteratively until suitable fits of the calculated and observed
structures of 3 and 4 were obtained. The final parameter set
added to the standard MM+ parameter files was as follows.
Bond stretching (bond, ks/mdyn Å-1, R0/Å, µD/D): C(sp2)-I,
2.620, 2.090, 1.400; O(sp3)-S(sp3), 4.620, 1.580, -3.073; O(sp2)-
S(sp3), 8.412, 1.450, -3.027; and C(sp2)-S(sp3), 3.260, 1.750,
-1.563. Angle bending (angle, kθ/mdyn Å rad-2, θ0/deg):
C(sp3)-O(sp3)-S(sp3), 0.770, 110.0; C(sp2)-C(sp2)-S(sp3), 0.550,
121.4; O(sp3)-S(sp3)-C(sp2), 0.700, 103.6; O(sp3)-S(sp3)-
O(sp2), 0.560, 106.2; and C(sp2)-S(sp3)-O(sp2), 0.650, 109.6.
Torsion angle rotation (dihedral angle, V1/kcal mol-1, V2/kcal
mol-1, V3/kcal mol-1): C(sp3)-C(sp3)-O(sp3)-S(sp3), 0.400,
0.520, 0.467; H-C(sp3)-O(sp3)-S(sp3), 0.000, 0.000, 0.530;
H-C(sp3)-O(sp3)-S(sp3), 0.000, 0.000, 0.403; C(sp3)-O(sp3)-
S(sp3)-O(sp2), 0.000, 0.000, 0.530; C(sp2)-C(sp2)-C(sp2)-
S(sp3), -0.270, 9.000, 0.000; C(sp2)-O(sp3)-C(sp3)-O(sp3),
0.000, 5.000, 0.000; C(sp2)-C(sp2)-S(sp3)-O(sp3), 0.000, 0.000,
0.000; C(sp2)-C(sp2)-S(sp3)-O(sp2), 1.243, 1.445, -1.243; and
H-C(sp2)-C(sp2)-S(sp3), 0.000, 9.000, 0.000.
F IGURE 8. Natural atomic charges for methyl benzene-
sulfonate calculated at the B3LYP/6-31G* level of theory. Bond
distances for the unique bonds not involving H atoms are given
in boldface type.
(EI) 446 (19, M+), 274 (100, M+ - HOTs), 261 (29, M+
-
CH2OTs) (found M+ 455.9682, C16H15O5SI requires 455.9685).
Cr ysta llogr a p h y. Single-crystal X-ray data were collected
at 295(2) K (Mo KR radiation). The data were reduced with
the program XCAD (Oscail V826) using Lorentz and polariza-
tion correction factors. A post-refinement numerical absorption
correction (DIFABS27) was applied to the data for 4. The
structures were solved with the direct methods program
SHELXS-97.28 The E-maps led to the location of all non-
hydrogen atoms; these were refined anisotropically with the
program SHELXL-97.28 All hydrogens were included as ideal-
ized contributors in the least-squares process with standard
SHELXL-97 idealization parameters. The final models were
plotted with the program ORTEP.29 Complete crystallographic
details, fractional atomic coordinates for all non-hydrogen
atoms, anisotropic thermal parameters, fixed hydrogen atom
coordinates, bond lengths, bond angles, and dihedral angles
for 3 and 4 are available in the Supporting Information.
Experimental lattice constants and SHELXL-97 refinement
parameters for 3 and 4 are given in Table 1.
DF T Ca lcu la tion s. These were used primarily to obtain
estimates of bond dipoles for the new/modified MM+ force field
parameter set that we have developed for sulfonic esters. A
“parent” derivative, methyl benzenesulfonate (Figure 8), was
chosen since it is representative of this class of compounds.
Full geometry optimization was effected at the B3LYP30
/
6-31G* level of theory with the program Titan 1.0831 using
the following convergence criteria: rms of gradient elements
) 3.0 × 10-4 Hartrees/Bohr; rms of nuclear displacement
elements ) 1.2 × 10-3 Bohr; and energy difference between
geometry optimization cycles <5.0 × 10-5 Hartrees. Accurate
partial atomic charges for the sulfonic ester group were
obtained via NBO32 analysis of the converged DFT wave
function. (The NBO method of partial charge assignment
affords basis set independent charge distributions.)
Bond dipoles, µD (in debye), were estimated from the natural
atomic charge distribution with eq 1
Con for m a tion a l An a lysis. Conformational analyses using
a standard dihedral angle driving algorithm were carried out
by rotating a pair of torsion angles from 0° to 360° in 5°
increments with optimization of all remaining internal degrees
of freedom at each grid coordinate. Three sulfonic ester chain
torsion angles were used to map out the interconversion of
the stepped and hairpin conformations of 3 and 4; ψ1, C(sp2)-
S(sp3)-O(sp3)-C(sp3); ψ2, S(sp3)-O(sp3)-C(sp3)-C(sp3); and
ψ3, O(sp3)-C(sp3)-C(sp3)-C(sp2). (X-ray structures were used
as the starting conformations for these calculations.)
(26) Oscail Version 8: McArdle, P.; Crystallography Center, Chem-
istry Department, NUI Galway, Ireland. McArdle, P. J . Appl. Crys-
tallogr. 1995, 28, 65-65.
(27) DIFABS: Walker, N. P.; Stuart, D. Acta Crystallogr., Sect. A
1983, A39, 158-166.
(28) SHELXL-97 and SHELXS-97: Sheldrick, G. M. University of
Gottingen. (a) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, A46,
467-473. (b) Sheldrick, G. M. Acta Crystallogr., Sect. D 1993, D49,
18-23. (c) Sheldrick, G. M.; Schneider, T. R. Methods Enzymol. 1997,
277, 319-343.
(29) ORTEP3 for Windows: Farrugia, L. J . Department of Chem-
istry, University of Glasgow, Glasgow G12 8QQ, Scotland, 2001. (a)
Farrugia, L. J . J . Appl. Crystallogr. 1997, 30, 565. (b) ORTEP III:
Burnett, M. N.; J ohnson, C. K. Oak Ridge National Laboratory report
ORNL-6895, 1996.
(32) Weinhold, F. 1999. NBO 4M: Theoretical Chemistry Institute,
University of Wisconsin, Madison, WI.
(30) Becke, A. D. J . Chem. Phys. 1993, 98, 5648.
(33) HyperChem 6.03: Hypercube, Inc., 1115 NW 4th St., Gainsville,
FL 32601-4256.
(34) Allinger, N. L.; Zhou, X.; Bergsma, J . THEOCHEM 1994, 312,
69-83.
(31) (a) Wavefunction, Inc.: 18401 Von Karman Ave., Suite 370,
Irvine, CA 92612. (b) Schro¨dinger, Inc.: 1500 SW First Ave., Suite
1180, Portland, OR 97201.
2458 J . Org. Chem., Vol. 68, No. 6, 2003

Conformational Analysis: Flexible Sulfonic Esters
La n gevin Dyn a m ics Ca lcu la tion s. Simulations of the
dynamic conformational behavior of 3 and 4 were carried out
using a nonexplicit solvent model (Langevin equation) at 300,
350, and 400 K for 6 ns with a 1.0 fs time step at a constant
temperature (bath relaxation time ) 0.1 ps, friction coefficient
) 2.5 ps-1). Each simulation was started at 0 K and then
heated to the required temperature over 15 ps in 10 K
increments. Selected torsion angles were monitored during
each simulation. The trajectories were analyzed with a devel-
opmental version of the 32-bit program MDcalc 1.0 that we
are currently in the process of writing for population and rate
constant analyses of molecular dynamics trajectories.
MOs of the hairpin conformer of 4 (Figure S3); root-mean-
square fit of the MM-calculated hairpin conformations of 3 and
4 (Figure S4); crystal packing diagram for 4 (Figure S5); crystal
packing diagrams highlighting short intermolecular contacts
for 3 and 4 (Figures S6-S8); 2D maps of the ψ2/ψ3 and ψ1/ψ2
conformational energy surfaces of 4 (Figures S9 and S10,
respectively); scatter plot of the ψ5/ψ4 reaction coordinate from
the LD trajectory of 4 at 300 K (Figure S11); minimum energy
conformations of the lowest-lying gauche ψ2 conformers of 4
(Figure S12); plots of the ψ1/ψ3 conformational energy surface
of 3 (Figure S13); scatter plot of the ψ1/ψ3 reaction coordinate
from the LD trajectory of 4 at 400 K (Figure S14); Arrhenius
plot for the E T D conformational exchange of 4 (Figure S15);
calculations of bond dipole parameters from NBO charge
distributions; tables of crystallographic data for compounds 4
(Table S1-S6) and 3 (Table S8-S13); comparisons of MM and
X-ray structural data for tosylate 4 (Table S7); complete
synthetic details for nosylate 3; a ZIP file containing PDB files
of each energy minimized conformation of 3 and 4; and a CIF
file containing the X-ray structural data for 3 and 4. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from the University of Natal (URF) and
the National Research Foundation (Pretoria). We would
also like to thank Martin C. Watson for his assistance
with recording some of our NMR spectra and J ames
Ryan for selecting crystals of 3 and setting up the X-ray
data collection for this compound.
Su p p or tin g In for m a tion Ava ila ble: ORTEP diagrams
of the X-ray structures of each independent molecule of 3 and
4 (Figures S1 and S2, respectively); PM3-calculated frontier
J O0260342
J . Org. Chem, Vol. 68, No. 6, 2003 2459