Asymmetric ruthenium-catalysed [2+2] cycloadditions between bicyclic alkenes and chiral aryl-substituted acetyl-enic acyl camphorsultam alkynes

Article abstract of DOI:10.1107/S0108270109045284

The regio- and absolute stereochemistry of (7S)-N-[4-(3-thien-yl)tricyclo- [4.2.1.0^2,5]non-3-en-3-ylcarbon-yl]-2,10-camphorsultam tetra-hydro-furan hemisolvate, C₂₄H₂₉NO₃S ₂·0.5C₄H₈

Full text of DOI:10.1107/S0108270109045284

organic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
the acetylenic position (see scheme). In this paper, we present
the results of ruthenium-catalysed [2+2] cycloadditions of
norbornene, (I), with 3-thiophenylacetyleneacylcamphorsul-
tam, (III), and 4-tolylacetyleneacylcamphorsultam, (IV). This
yielded the expected complexes (7S)-N-[4-(3-thienyl)tri-
cyclo[4.2.1.02,5]non-3-en-3-ylcarbonyl]-2,10-camphorsultam
tetrahydrofuran hemisolvate, (VI), and (7S)-N-[4-(4-tolyl)-
tricyclo[4.2.1.02,5]non-3-en-3-ylcarbonyl]-2,10-camphorsultam,
(VII). Two diastereomers of (VI) were obtained in a 20:1 ratio,
ISSN 0108-2701
Asymmetric ruthenium-catalysed
[2+2] cycloadditions between bicyclic
alkenes and chiral aryl-substituted
acetylenic acyl camphorsultam
alkynes
Jordan D. Goodreid,a Michael C. Jenningsb* and William
Tama
aDepartment of Chemistry, University of Guelph, 50 Stone Road East, Guelph,
Ontario, Canada N1G 2W1, and b185 Chelsea Avenue, London, Ontario, Canada
N6J 3J5
Correspondence e-mail: mjennings@teksavvy.com
Received 14 September 2009
Accepted 29 October 2009
Online 21 November 2009
The regio- and absolute stereochemistry of (7S)-N-[4-(3-
thienyl)tricyclo[4.2.1.02,5]non-3-en-3-ylcarbonyl]-2,10-camphor-
sultam tetrahydrofuran hemisolvate, C24H29NO3S2ꢀ0.5C4H8O,
and (7S)-N-[4-(4-tolyl)tricyclo[4.2.1.02,5]non-3-en-3-ylcarbon-
yl]-2,10-camphorsultam, C27H33NO3S, have been established.
One contains a half-occupancy tetrahydrofuran solvent mol-
ecule located on a twofold axis and the other contains two
crystallographically unique molecules which are nearly
identical. The extended structures of both complexes can be
explained via weak C—Hꢀ ꢀ ꢀO interactions, which link the
molecules together into two-dimensional sheets in the ab
plane for the thienyl complex and ultimately into a three-
dimensional structure for the tolyl derivative. The stereo-
chemistry of both structures confirms that [2+2] cycloadditions
of bicyclic alkenes and alkynes catalysed by ruthenium are
exclusively exo.
whereas the major stereoisomer of (VII) was obtained in a
104:1 ratio over the minor isomer. The absolute stereo-
chemistry of the major isomers of (VI) and (VII) was estab-
Comment
Recently, the metal-catalysed [2+2] cycloaddition reaction
between bicyclic alkenes and alkynes has been utilized by
several research groups as a means to generate cyclobutene
rings. In terms of an asymmetric version of this process, both a
chiral rhodium catalyst (Shibata et al., 2006) and a ruthenium-
catalysed reaction (Villeneuve & Tam, 2004) involving a chiral
precursor have produced chiral [2+2] cyclobutene products.
The ruthenium-catalysed [2+2] cycloaddition between nor-
bornene and a phenylacetyleneacyl camphorsultam (chiral)
yielded two diastereomers in a 131:1 ratio. The absolute
stereochemistry of the major diastereomer (V) was estab-
lished by single-crystal X-ray diffraction (Lough et al., 2004).
As an extension of this study, the Tam group is investigating
the selectivity effect of different organic groups attached at
Figure 1
A view of the structure of (VI), showing the crystallographic labelling
scheme. Displacement ellipsoids are drawn at the 50% probablility level.
For the sake of clarity, nonchiral H atoms have been omitted, as has the
partial-occupancy THF molecule of solvation.
Acta Cryst. (2009). C65, o649–o652
doi:10.1107/S0108270109045284
# 2009 International Union of Crystallography o649

organic compounds
lished by the single-crystal X-ray diffraction analysis described
here.
the extension along the c axis, each of which is rotated by 120ꢁ
with respect to the previous layer. The THF solvent molecule
was refined at half-occupancy, as it is located on a rotational
axis; the O atom links to one of the layers via another weak
interaction (C1—H1Aꢀ ꢀ ꢀO31).
A view of (VI) is shown in Fig. 1. Elucidation of the X-ray
structure establishes the absolute configuration of the
following atoms in Fig. 1: C2-S, C5-R, C8-R, C14-S, C15-R,
C18-S and C19-R. This is the same configuration as that
observed for the phenyl derivative (V) (Lough et al., 2004).
There is a tetrahydrofuran (THF) solvent molecule occluded
in the structure. It was modelled at half-occupancy, as it is
As with (VI), it is the weak C—Hꢀ ꢀ ꢀO interactions that
control the geometry observed in the crystal structure of (VII)
(Table 2). The nine weak interactions lead to a very extensive
set of hydrogen-bonding motifs so only the key graph-set
descriptors are noted here. The asymmetric unit contains two
molecules, which couple together via two weak interactions
(C3—H3Bꢀ ꢀ ꢀO31 and C31—H31Bꢀ ꢀ ꢀO3) to form an 11-
membered ring, and, adjacent to this, a nine-membered ring is
formed via C38—H38Aꢀ ꢀ ꢀO1i and C1—H1Bꢀ ꢀ ꢀO33ii inter-
actions. This quaternary graph set, somewhat similar to that in
(VI), has the motifs C22(10)[R22(11)R22(9)]: a ‘chain of rings’
yielding ribbons along the a axis (Fig. 4). These a-axis ribbons
are connected up and down the c axis according to the binary
1
located about (12, , 0), which lies on a twofold axis along the
2
[110] vector. The THF molecule was refined giving reasonable
displacement parameters; the O atom of this molecule is
involved in the weak interactions discussed below.
There are two chemically identical molecules of (VII) in the
asymmetric unit. They are very similar structurally, with an
˚
r.m.s. fit of 0.207 A (PLATON; Spek, 2009). As expected, both
have the same configuration as that observed for (VI). The
absolute configuration, as determined by anomalous disper-
sion, is as follows: C2-S, C5-R, C8-R, C14-S, C15-R, C18-S and
C19-R for one molecule, and C32-S, C35-R, C38-R, C44-S,
C45-R, C48-S and C49-R for the second. The structure of one
of the molecules of (VII) is shown in Fig. 2. The crystal-
lographic difference in the two molecules is primarily attrib-
uted to their spatially different orientations with respect to
one another (see below). The geometric data for both (VI)
and (VII) are in the normal ranges.
Weak C—Hꢀ ꢀ ꢀO interactions in (VI) (see Table 1 for
geometric parameters and symmetry codes) link the mol-
ecules. Perhaps the simplest method to describe the packing is
by using two ‘classical model’ graph-set descriptors (Bernstein
et al., 1995). Firstly, a ‘chain of rings’ is formed through two
weak interactions (C8—H8Aꢀ ꢀ ꢀO2i and C1—H1Bꢀ ꢀ ꢀO3ii) that
link molecules together into ribbons along the a axis {motif
C22(10)[R22(9)]; Fig. 3}. Secondly, adjacent ribbons are linked
together via reciprocating C14—H14Aꢀ ꢀ ꢀO1iii interactions. It
should be noted that these are the ‘weakest’ of the weak
interactions mentioned herein. There are two further layers in
Figure 3
The packing of (VI), showing the primary chain-forming C—Hꢀ ꢀ ꢀO
interactions as dashed lines. All H atoms, except those attached to atoms
involved in weak interactions, have been omitted for clarity. The view is
approximately down the c axis, showing the ribbons along the a axis.
Figure 4
Figure 2
The packing of (VII), showing the primary chain-forming C—Hꢀ ꢀ ꢀO
interactions as dashed lines. All H atoms, except those involved in the
displayed weak interactions, have been omitted for clarity. The view is
approximately down the c axis, showing the formation of a chain along
the a axis.
The molecular structure of one of the molecules of (VII), showing the
atom-labelling scheme and with displacement ellipsoids drawn at the 50%
probability level. For the sake of clarity, the nonchiral H atoms have been
omitted.
ꢂ
o650 Goodreid et al. C24H29NO3S2ꢀ0.5C4H8O and C27H33NO3S
Acta Cryst. (2009). C65, o649–o652

organic compounds
graph-set descriptor C22(12) via C50—H50Aꢀ ꢀ ꢀO2iii and C9—
H9Aꢀ ꢀ ꢀO3iv interactions, thus forming parallel ribbons in the
[010] plane. Finally, a full three-dimensional network is
established via C27—H27Aꢀ ꢀ ꢀO32v and C9—H9Aꢀ ꢀ ꢀO3iv
interactions, making C22(25) connections along the b axis. For
completeness, the last two interactions in Table 2 can be
described by motif S(7).
Table 1
Hydrogen-bond geometry (A, ) for (VI).
ꢁ
˚
D—Hꢀ ꢀ ꢀA
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
C8—H8Aꢀ ꢀ ꢀO2i
C1—H1Bꢀ ꢀ ꢀO3ii
C14—H14Aꢀ ꢀ ꢀO1iii
C1—H1Aꢀ ꢀ ꢀO31
1.00
0.99
1.00
0.99
2.59
2.45
2.66
2.36
3.516 (3)
3.422 (3)
3.500 (3)
3.235 (14)
154
168
141
147
1
3
1
3
Symmetry codes: (i) x ꢃ y; ꢃy þ 1; ꢃz þ ; (ii) x ꢃ y þ 1; ꢃy þ 1; ꢃz þ ; (iii) x ꢃ y þ 1,
While there are no classical hydrogen-bonding networks in
either of these structures, the weak interactions do seem to
explain the observed geometry, which is understandable as
even a weak hydrogen bond will be stronger than the van der
Waals interactions. Unlike the previously published phenyl
cycloadduct which shows partial ꢀ–ꢀ stacking between
1
3
ꢃy þ 2; ꢃz þ .
Table 2
Hydrogen-bond geometry (A, ) for (VII).
ꢁ
˚
D—Hꢀ ꢀ ꢀA
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
˚
neighbouring phenyl rings (closest contact = 3.64 A) through
C3—H3Bꢀ ꢀ ꢀO31
C31—H31Bꢀ ꢀ ꢀO3
C38—H38Aꢀ ꢀ ꢀO1i
C1—H1Bꢀ ꢀ ꢀO33ii
C50—H50Aꢀ ꢀ ꢀO2iii
C9—H9Aꢀ ꢀ ꢀO3iv
C27—H27Aꢀ ꢀ ꢀO32v
C22—H22Aꢀ ꢀ ꢀO3
C52—H52Aꢀ ꢀ ꢀO33
0.99
0.99
1.00
0.99
0.99
0.98
0.98
0.95
0.95
2.58
2.45
2.54
2.37
2.53
2.59
2.43
2.31
2.33
3.539 (3)
3.356 (2)
3.412 (2)
3.332 (3)
3.446 (2)
3.570 (3)
3.404 (3)
3.115 (3)
3.129 (3)
163
152
146
164
153
174
174
142
141
secondary orbital overlap interactions (Lough et al., 2004), no
such ꢀ-stacking interactions between either thiophenyl rings
or tolyl rings are observed in these crystal structures.
In the case of the [2+2] cycloaddition, these crystal struc-
tures have unambiguously identified which are the major
diastereomers formed in the asymmetric cycloaddition. The
previous phenyl derivative combined with both the thiophenyl
and the tolyl derivatives herein reaffirm that [2+2] cyclo-
additions of bicyclic alkenes and alkynes catalysed by ruthe-
nium are exclusively exo with respect to their stereochemistry.
1
2
3
2
1
2
Symmetry codes: (i) x ꢃ 1; y; z; (ii) x þ 1; y; z; (iii) ꢃx þ 1; y ꢃ ; ꢃz þ ; (iv) x þ ,
3
2
ꢃy þ ; ꢃz þ 2; (v) x; y þ 1; z.
Data collection
Nonius KappaCCD diffractometer
Absorption correction: multi-scan
(DENZO–SMN; Otwinowski &
Minor, 1997)
8354 measured reflections
4207 independent reflections
3310 reflections with I > 2ꢃ(I)
Rint = 0.041
Experimental
Preparation of compound (VI) was achieved by addition of norbor-
nene (72.0 mg, 765 mmol), (I), and chiral alkyne (III) (52.2 mg,
0.149 mmol) to Cp*Ru(cod)Cl (Cp* is pentamethylcyclopentadienyl
and cod is 1,5-cyclooctadiene; 7.6 mg, 0.020 mmol) in THF (0.6 ml).
The mixture was stirred for 18 h at 298 K, and subsequent column
chromatography of the crude reaction mixture provided a 98% yield
of the two diastereomeric cycloadducts in a 20:1 ratio. Evaporative
recrystallization of the column-purified product in THF provided the
major diastereomer (VI), giving colourless crystals suitable for X-ray
analysis. Similarly, compound (VII) was prepared by addition of
norbornene (83.0 mg, 0.881 mmol), (I), and chiral alkyne (IV)
(63.2 mg, 0.177 mmol) to Cp*Ru(cod)Cl (7.1 mg, 0.018 mmol) in
THF (0.5 ml). The mixture was stirred for 17 h at 298 K, and subse-
quent column chromatography of the crude reaction mixture
provided a 79% yield of the two diastereomeric cycloadducts in a
104:1 ratio. Evaporative recrystallization of the column-purified
product in THF provided major diastereomer (VII), giving colourless
crystals suitable for X-ray analysis. Both diastereomeric ratios were
back-calculated from the corresponding ee values obtained through
cleavage of the chiral auxiliary under reductive conditions (LiAlH4,
AlCl3), giving an enantiomeric mixture of cycloadduct alcohols that
can be resolved using a Chiralcel OJ-H high-performance liquid
chromatography column.
Tmin = 0.916, Tmax = 0.963
Refinement
R[F2 > 2ꢃ(F2)] = 0.038
wR(F2) = 0.084
S = 1.01
4207 reflections
318 parameters
4 restraints
H-atom parameteꢃrs3 constrained
˚
Áꢄmax = 0.19 e A
ꢃ3
˚
Áꢄmin = ꢃ0.31 e A
Absolute structure: Flack (1983),
1788 Friedel pairs
Flack parameter: ꢃ0.03 (7)
Compound (VII)
Crystal data
3
˚
V = 4737.2 (16) A
Z = 8
C27H33NO3S
Mr = 451.60
Orthorhombic, P212121
˚
Mo Kꢁ radiation
ꢂ = 0.17 mmꢃ1
T = 150 K
0.40 ꢄ 0.35 ꢄ 0.17 mm
a = 11.687 (2) A
˚
b = 12.849 (3) A
˚
c = 31.547 (6) A
Data collection
Nonius KappaCCD diffractometer
Absorption correction: multi-scan
(DENZO–SMN; Otwinowski &
Minor, 1997)
44242 measured reflections
9673 independent reflections
7676 reflections with I > 2ꢃ(I)
Rint = 0.068
Tmin = 0.937, Tmax = 0.972
Compound (VI)
Crystal data
Refinement
ꢃ3
R[F2 > 2ꢃ(F2)] = 0.041
wR(F2) = 0.096
S = 1.03
9673 reflections
583 parameters
H-atom parameters constrained
Áꢄmax = 0.19 e A
˚
C24H29NO3S2ꢀ0.5C4H8O
Z = 6
ꢃ3
˚
Mr = 479.65
Trigonal, P3221
Mo Kꢁ radiation
ꢂ = 0.25 mmꢃ1
T = 150 K
0.35 ꢄ 0.22 ꢄ 0.15 mm
Áꢄmin = ꢃ0.31 e A
Absolute structure: Flack (1983),
4291 Friedel pairs
˚
a = 12.0797 (17) A
˚
c = 28.326 (6) A
˚
V = 3579.5 (10) A
Flack parameter: ꢃ0.02 (5)
3
ꢂ
Acta Cryst. (2009). C65, o649–o652
Goodreid et al.
C24H29NO3S2ꢀ0.5C4H8O and C27H33NO3S o651

organic compounds
Both the O—C bonds and the C—C bonds of the disordered
solvent (THF) were restrained, respectively, to be identical. H atoms
were positioned geometrically and refined using a riding model, with
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: GA3134). Services for accessing these data are
described at the back of the journal.
˚
C—H distances of 0.98 (CH3), 0.99 (CH2), 1.00 (CH) or 0.95 A
(aromatic CH). Uiso(H) values were set at 1.2 times Ueq(C) for all
except the methyl H atoms, for which Uiso(H) = 1.5Ueq(C).
For both compounds, data collection: COLLECT (Nonius, 1997);
cell refinement: DENZO–SMN (Otwinowski & Minor, 1997); data
reduction: DENZO–SMN; program(s) used to solve structure:
SHELXS97 (Sheldrick, 2008); program(s) used to refine structure:
SHELXL97 (Sheldrick, 2008); molecular graphics: SHELXTL/PC
(Sheldrick, 2008); software used to prepare material for publication:
SHELXTL/PC.
References
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew. Chem.
Int. Ed. 34, 1555–1573.
Flack, H. D. (1983). Acta Cryst. A39, 876–881.
Lough, A. J., Villeneuve, K. & Tam, W. (2004). Acta Cryst. E60, o1566–o1567.
Nonius (1997). KappaCCD Server Software. Windows 3.11 Version. Nonius
BV, Delft, The Netherlands.
Otwinowski, Z. & Minor, W. (1997). Methods in Enzymology, Vol. 276,
Macromolecular Crystallography, Part A, edited by C. W. Carter Jr & R. M.
Sweet, pp. 307–326. New York: Academic Press.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122.
Shibata, T., Takami, K. & Kawashi, A. (2006). Org. Lett. 8, 1343–1345.
Spek, A. L. (2009). Acta Cryst. D65, 148–155.
Villeneuve, K. & Tam, W. (2004). Angew. Chem. Int. Ed. 43, 610–613.
The authors acknowledge NSERC Canada, the University
of Western Ontario and the University of Guelph for funding.
ꢂ
o652 Goodreid et al. C24H29NO3S2ꢀ0.5C4H8O and C27H33NO3S
Acta Cryst. (2009). C65, o649–o652