Neutron diffraction analysis of deuterium transfer in chiral β-thiolactam formation in the crystalline state

Article abstract of DOI:10.1107/S0108768105038656

Since N,N-dibenzyl-1-cyclohexenecarbothioamide is photoisomerized to the optically active β-thiolactam with the retention of the single-crystal form, the mechanism of chirality induction was identified by X-ray crystal structure analyses during the process of the reaction [Hosoya et al. (2002). Bull. Chem. Soc. Jpn, 75, 2147-2151]. In order to clarify the mechanism of hydrogen transfer in the reaction, the H atoms of the benzyl groups were replaced with deuterium atoms. The crystal structure after photoisomerization was analyzed by neutron diffraction. One of four deuterium atoms of the two benzyl groups is transferred to the C atom of the cyclohexene ring. The absolute configuration of the -C*HD- group (chiral methylene) in the photoproduct β-thiolactam revealed that the deuterium atom occupies the equatorial position. This suggests that the deuterium atom is not transferred from the benzyl group of a neighbouring molecule, but from one of the benzyl groups within the molecule.

Full text of DOI:10.1107/S0108768105038656

research papers
Acta Crystallographica Section B
Structural
Neutron diffraction analysis of deuterium transfer in
Science
chiral b-thiolactam formation in the crystalline state
ISSN 0108-7681
Takaaki Hosoya,a Hidehiro
Uekusa,a Yuji Ohashi,a* Takashi
Ohhara,b Ichiro Tanakac and
Nobuo Niimurac
Since N,N-dibenzyl-1-cyclohexenecarbothioamide is photo-
isomerized to the optically active ꢀ-thiolactam with the
Received 29 July 2005
Accepted 22 November 2005
retention of the single-crystal form, the mechanism of chirality
induction was identified by X-ray crystal structure analyses
during the process of the reaction [Hosoya et al. (2002). Bull.
Chem. Soc. Jpn, 75, 2147–2151]. In order to clarify the
mechanism of hydrogen transfer in the reaction, the H atoms
of the benzyl groups were replaced with deuterium atoms. The
crystal structure after photoisomerization was analyzed by
neutron diffraction. One of four deuterium atoms of the two
benzyl groups is transferred to the C atom of the cyclohexene
ring. The absolute configuration of the —C*HD— group
(chiral methylene) in the photoproduct ꢀ-thiolactam revealed
that the deuterium atom occupies the equatorial position. This
suggests that the deuterium atom is not transferred from the
benzyl group of a neighbouring molecule, but from one of the
benzyl groups within the molecule.
aDepartment of Chemistry and Materials
Science, Tokyo Institute of Technology,
O-okayama, Meguro, Tokyo 152-8551, Japan,
bInstitute of Materials Structure Science, High
Energy Accelerator Research Organization,
Oho, Tsukuba, Ibaraki 305-0801, Japan, and
cNeutron Structural Biology, Advanced Science
Research Center, Japan Atomic Energy Research
Institute, Tokai, Naka, Ibaraki 319-1195, Japan
Correspondence e-mail:
yujiohashi@cms.titech.ac.jp
1. Introduction
A photoinduced absolute asymmetric synthesis from an
achiral N,N-dibenzyl-1-cyclohexenecarbothioamide (1a) to a
chiral ꢀ-thiolactam (2a) in the solid state has been reported
(Sakamoto et al., 1996, 2000), as shown in (I). After many
trials, the photoreaction partly proceeded without the
destruction of the single-crystalline nature. The structural
change from (1a) to (2a) was directly observed by X-ray
crystal structure analysis (Hosoya et al., 2002). It was assumed
from the results mentioned above that an H atom is extracted
from a benzyl group and transferred to one of the photo-
excited alkenyl C atoms of the cyclohexene group at the first
stage, and then a new bond between another alkenyl carbon
and the benzyl carbon radical produced by the hydrogen
abstraction is formed. A four-membered ꢀ-thiolactam ring is
thus produced. Although the reaction mechanism explaining
how the chirality is introduced in the product during four-
membered ring formation in a crystal was made clear by X-ray
analysis, there remained an unsolved problem as to which
hydrogen was transferred, since there are several candidates.
The H-atom or proton transfer is one of the most elemental
phenomena and plays a key role in various reactions, for
example, absolute asymmetric synthesis via intramolecular
(Ihmels & Scheffer, 1999; Sakamoto et al., 2000; Sakamoto,
2005) or intermolecular (Tang et al., 1995; Chang et al., 1987)
H-atom transfer, photochromism (Hadjoudis & Mavridis,
2004; Naumov et al., 2005), asymmetric hydrogen transfer with
catalysis (Zassinovich & Mestroni, 1992; Clapham et al., 2004)
# 2006 International Union of Crystallography
Printed in Great Britain – all rights reserved
Acta Cryst. (2006). B62, 153–160
doi:10.1107/S0108768105038656 153

research papers
and enzymatic reactions (Soda et al., 2001; Pierre & Thomas,
2005).
ples of crystalline-state reactions in cobaloxime complexes
have been reported by neutron diffraction (Ohgo et al.,
1996a,b, 1997; Ohhara et al., 2000, 2001, 2002).
We intended to apply neutron diffraction analysis to the
crystalline-state photoreaction of the ꢁ,ꢀ-unsaturated thioa-
mide mentioned above in order to determine the H-atom
transfer mechanism. The preliminary result has already been
reported (Hosoya et al., 2003).
2. Experimental
2.1. Preparation
N,N-Di(benzyl-d7)-1-cyclohexenecarbothioamide (1b) was
prepared in a similar way to the procedure reported (Hosoya
et al., 2002) for the corresponding ꢁ,ꢀ-unsaturated amide by
the thionation of the carbonyl group with Lawesson’s Reagent
(Thomsen et al., 1984), as shown in (II).
If more than two H atoms participate in isomerization
reactions, the deuterium labeling technique of the target H
atom is a very powerful method to analyze the mechanism of
the H- or D-atom transfer by 1H NMR and IR spectra.
However, peak assignments are very difficult to make in the
case of a solid-state reaction, in which not only intramolecular
H or D atoms, but also intermolecular H or D atoms may be
transferred. The single-crystal X-ray analysis of the crystal-
line-state reaction also cannot determine the hydrogen-
transfer mechanism even if the target H atom is replaced with
the D atom, because X-ray diffraction usually cannot distin-
guish the H atom from the D atom. On the other hand,
neutron diffraction is able to analyze the position and thermal
motion of an H or D atom accurately and to distinguish a D
atom from an H atom. When the target H atoms of the
reactant molecule are replaced with D atoms as indicators and
the reaction proceeds with the retention of the single-crys-
talline form, the transfer of the target D atom in the crystal-
line-state reaction can be observed directly. Since the neutron
scattering length for the D atom is positive, whereas that of the
H atom is negative, the deuterium and hydrogen positions are
seen as positive and negative peaks, respectively, in the
nucleus density map. Therefore, the D-atom transfer in the
crystalline-state reaction can be directly observed by neutron
diffraction analysis since the peak heights of the replaced D
and H atoms are gradually changed in the process of isomer-
ization.
Especially if a chiral —C*HD— methylene is produced
from the achiral CH— group, the position of the D atom, i.e.
the absolute configuration of the chiral methylene, indicates
the source of the D atom. There were some previous reports
about the determination of the absolute configuration of such
a chiral —C*HD— methylene by neutron diffraction (Yuan et
al., 1988, 1994). However, there are some serious problems to
overcome in the neutron diffraction analysis of the crystalline-
state reaction. Owing to the weak intensity of the neutron
beam compared with that of X-rays, it was necessary to use a
large crystal of more than 1 mm3 in order to collect enough
intensity data, although it is not easy to make such large
crystals for most organic compounds. Moreover, the light
should penetrate into the large crystal, because the photo-
reaction should proceed not from the surface of the crystal but
inside the crystal. Owing to the difficulties, only a few exam-
2.1.1. Dibenzylamide-d15. A toluene solution (10 ml) of
benzyl-d7-amine (3.179 g, 28.11 mmol), benzaldehyde-d6
(3.159 g, 28.11 mmol) and a few boiling stones were placed in a
200 ml round-bottomed flask equipped with a Dean–Stack
fractionating column, a Dimroth condenser and a mantle
heater. The mixture was refluxed at ca 403 K with an azeo-
tropic distillation of the water produced. When the generation
of water stopped, the mixture solution was cooled down to
room temperature and the toluene was evaporated. A yellow
oil of N-benzylidenebenzylamine-d13 (3) was obtained.
The imine (3) was dissolved in methanol-d (100 ml) and
reduced with sodium borodeuteride (NaBD4, 1.18 g,
28.11 mmol). After stirring for 2 h, the solvent was evapo-
rated. Dibenzylamine-d15 (4) was extracted by diethyl ether
from the residual colourless liquid and neutralized by DCl
(35 wt %, 3.77 g, 37.3 mmol) solution. The resulting white
powder was washed with acetone and diethyl ether, yielding
ꢀ
154 Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam
Acta Cryst. (2006). B62, 153–160

research papers
6.95 g (27.84 mmol, 99%). Then (4) was regenerated from the
DCl salt by the addition of a D2O solution of NaOD, and then
extracted with diethyl ether.
neutron diffraction measurement were obtained. Four ꢁ-H
atoms of the two benzyl groups in (1b) were successfully
1
deuterated by over 98%, which was confirmed by H NMR
2.1.2. N,N-Di(benzyl-d7)-1-cyclohexenecarbothioamide.
1-Cyclohexene-1-carboxylic acid (2.50 g, 19.8 mmol) was
dissolved in 20 ml of THF and placed in a 100 ml flask
equipped with a dropping funnel. The solution was cooled
with an ice bath. Under stirring, 7.93 ml of thionyl chloride
(SOCl2, 19.8 mmol) was added dropwise over a period of
10 min. A Dimroth condenser was used rather than a dropping
funnel and the mixture was refluxed (oil bath temperature ca
363 K) for 2 h. After removing thionyl chloride and THF by
measurement of the crystals. A small single crystal [0.30 ꢁ 0.10
ꢁ 0.10 mm3, crystal (I)] and a large one [4.50 ꢁ 1.50 ꢁ
1.40 mm3, crystal (II)] were prepared for X-ray and neutron
diffraction analyses, respectively. Crystal (II) was subse-
quently photo-irradiated. A small single crystal for X-ray
diffraction measurement after irradiation [0.20 ꢁ 0.20 ꢁ
0.10 mm3, crystal (II0)] was cut out from the central part of
crystal (II) after carrying out the neutron diffraction
measurement of (II).
distillation, a brownish-yellow liquid, 1-cyclohexene-1-
carbonyl chloride (5), was obtained.
A diethyl ether solution (100 ml) of (4) (4.00 g, 19.8 mmol)
and equimolar triethylamine (6.8 ml, 19.8 mmol) were placed
in a 200 ml round-bottomed flask, and cooled with an ice bath.
Under vigorous stirring, 10 ml of a diethyl ether solution of (5)
was added dropwise through the dropping funnel over a
period of 30 min. A vigorous and exothermic reaction
occurred and a white solid was precipitated. After 2 h stirring,
the mixture was suction-filtered. The filtrate was washed
successively with 2N HCl, saturated sodium hydrogen carbo-
nate (NaHCO3) solution and saturated sodium chloride
solution and dried over anhydrous sodium sulfate (Na2SO4),
and then the solvent was removed with a rotary evaporator. A
white powder, N,N-di(benzyl-d7)-1-cyclohexenecarboamide
(6) was obtained and isolated by silica-gel column chromato-
graphy (n-hexane/ethyl acetate = 8/1).
Lawesson’s reagent (3.03 g, 7.5 mol) and (6) (4.79 g,
15.0 mol) were dissolved in 100 ml of toluene and placed in a
200 ml two-necked flask. The mixture was refluxed at ca 353 K
for 2 h. After cooling and evaporation of the solvent, N,N-
di(benzyl-d7)-1-cyclohexenecarbothioamide (1b) was sepa-
rated from the solution mixture by silica-gel column chro-
matography (n-hexane/ethyl acetate = 5/1), and recrystallized
from the diethyl ether solution. Crystals large enough for
2.2. Photoirradiation
The photoreaction in a single crystal usually occurs from the
surface of the crystal and the crystal easily decomposes into
microcrystalline particles. It is very difficult to isomerize a
crystal large enough for neutron diffraction without the loss of
crystallinity. It is very important for a photoreaction in the
crystalline state to be able to select the irradiation wave-
lengths. Use of wavelengths corresponding to the absorption
tail has been recommended (Enkelmann et al., 1993). Fig. 1
shows the UV–vis reflectance spectra of the powdered crystals
with the spectrophotometer (JASCO V-560). A peak around a
wavelength of 375 nm corresponds to the excitation of a C
C
bond of the cyclohexene group. The wavelengths from a xenon
lamp (SAN-EI SUPER BLIGHT 152S with USHIO xenon
short-arc lamp UXL-151DO) were selected with two filters; a
Y-44 sharp-cut filter (TOSHIBA), cutting off the light shorter
than 440 nm, and a heat-absorbing water filter. The selected
wavelengths are also shown in Fig. 1. Small crystals with a
volume of ca 0.2 mm3 were irradiated with light for 500 h as
mentioned above as a preliminary experiment. Crystal struc-
ture analysis after irradiation showed that more than 40% of
the molecules were isomerized to ꢀ-thiolactam. For crystal
(II), both sides were irradiated under the conditions
mentioned above for 1500 h at 273 K on a cool plate (NISSIN
Cool Plate NCP-2215). Although the crystal after irradiation
did not seem to be decomposed, the rocking curve of each
reflection was expanded. This indicates that a slight increase in
mosaicity may occur due to the photoreaction.
2.3. X-ray diffraction analysis of (1b) before and after
irradiation
X-ray diffraction data on (I) and (II0) were collected at 100
and 173 K using a Rigaku R-AXIS RAPID diffractometer
(Mo Kꢁ radiation) and equipped with a Rigaku nitrogen gas-
flow type cryostat. The data were collected by the !-scan
method with a 5.0ꢂ oscillation width and 200 s exposure time
per frame. The data were integrated, scaled, sorted and
averaged by PROCESS AUTO (Rigaku Corporation, 1998).
The structures were solved by direct methods using
SHELXS97 (Sheldrick, 1997a).
Figure 1
(a) UV–vis reflectance spectrum of a crystal of (1b) in MgSO4 (2.0 wt %)
at 297 K. The spectrum was converted from reflectance into Kubelka–
Munk. (b) Light wavelength distribution of xenon lamp through a fiber-
optic waveguide, a Y-44 sharp cut filter (HOYA) and a heat absorbing
water filter.
ꢀ
Acta Cryst. (2006). B62, 153–160
Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam 155

research papers
Table 1
Experimental details.
(I)
(II0)
(II)
Crystal data
Chemical formula
Mr
Cell setting, space group
C21H9D14NS
335.43
Monoclinic, P21
8.5992 (7), 10.1263 (7), 10.0412 (7)
98.364 (4)
865.07 (11)
2
C21H9D14NS
335.43
Monoclinic, P21
8.7075 (5), 10.2068 (6), 10.0908 (9)
99.909 (3)
883.45 (11)
2
C21H9D14NS
335.43
Monoclinic, P21
8.9117 (11), 10.5512 (17), 10.3463 (13)
102.589 (8)
949.5 (2)
2
˚
a, b, c (A)
ꢀ (ꢂ)
3
˚
V (A )
Z
Dx (Mg mꢃ3
Radiation type
)
1.234
1.208
1.173
Mo Kꢁ
0.71073
7397
2.0–25.3
0.19
173 (2)
Prism, pale yellow
0.30 ꢁ 0.10 ꢁ 0.10
Mo Kꢁ
0.71073
8684
2.0–27.5
0.18
173 (2)
Prism, pale yellow
0.20 ꢁ 0.20 ꢁ 0.10
Neutron
1.51000
1880
5.8–49.0
0.21
293 (2)
Prism, pale yellow
4.50 ꢁ 1.50 ꢁ 1.40
˚
Wavelength (A)
No. of reflections for cell parameters
ꢄ range (ꢂ)
ꢅ (mmꢃ1
)
Temperature (K)
Crystal form, colour
Crystal size (mm)
Data collection
Diffractometer
Data collection method
Absorption correction
Rigaku RAXIS-RAPID
! scan
Multi-scan (based on symmetry-
related measurements)
0.946
Rigaku RAXIS-RAPID
! scan
Multi-scan
BIX-III
! scan
Gaussian
Tmin
Tmax
0.964
0.982
8684, 3665, 3573
0.591
0.784
1880, 1301, 1300
0.982
7397, 2850, 2807
No. of measured, independent and
observed reflections
Criterion for observed reflections
Rint
I > 2ꢆ(I)
0.031
25.3
I > 2ꢆ(I)
0.038
27.5
I > 2ꢆ(I)
0.044
49.0
ꢄmax (ꢂ)
Range of h, k, l
ꢃ10 ) h ) 10
ꢃ12 ) k ) 11
ꢃ12 ) l ) 12
ꢃ11 ) h ) 11
ꢃ12 ) k ) 13
ꢃ13 ) l ) 13
ꢃ8 ) h ) 8
ꢃ10 ) k ) 10
ꢃ10 ) l ) 10
Refinement
Refinement on
F2
F2
F2
0.079, 0.197, 1.08
130
297
Mixture of independent and
constrained refinement
R[F2 > 2ꢆ(F2)], wR(F2), S
No. of reflections
No. of parameters
H-atom treatment
0.027, 0.062, 1.07
2850
301
0.043, 0.116, 1.06
3665
270
Independent
Constrained
Weighting scheme
w = 1/[ꢆ2(Fo2) + (0.0352P)2 + 0.1521P], w = 1/[ꢆ2(Fo2) + (0.0758P)2 + 0.1336P], w = 1/[ꢆ2(Fo2) + (0.1086P)2 + 0.1004P],
where P = (Fo2 + 2Fc2)/3
< 0.0001
0.16, ꢃ0.17
SHELXL
0.050 (3)
Flack (1983)
ꢃ0.02 (5)
where P = (Fo2 + 2Fc2)/3
< 0.0001
0.39, ꢃ0.17
SHELXL
0.032 (4)
Flack (1983)
ꢃ0.02 (7)
where P = (Fo2 + 2Fc2)/3
< 0.0001
(Á/ꢆ)max
Áꢇmax, Áꢇmin (e A
Extinction method
Extinction coefficient
Absolute structure
Flack parameter
ꢃ3
˚
)
0.04, ꢃ0.04
None
–
Determined by (II0)
–
Computer programs used: PROCESS-AUTO (Rigaku Corporation, 1998), local program, HKL (HKL Research, Inc.; Otwinowski & Minor, 1997), CrystalStructure3.6 (Rigaku
Corporation, 2004), SIR2004 (Burla et al., 2005), SHELXS97 (Sheldrick, 1999a), initial structure determined by X-rays, SHELXL97 (Sheldrick, 1997b), ORTEP3 for Windows (Farrugia,
1997).
2.4. Neutron diffraction measurement of (1b) after irradia-
tion
diffractometer and NIP is short in a longitudinal direction, the
data sets were divided by a series of aluminium pin angle
values (ca 180, 135 and 90ꢂ) rather than changing the ꢃ circle
position, and the total oscillation ranges were 180, 90 and 90ꢂ,
respectively. The peaks of the resulting NIP frames were
integrated using the Denzo program with a resolution range of
After photoirradiation crystal (II) was fixed on an alumi-
nium plate pin and mounted on a BIX-III diffractometer
(Tanaka et al., 2002), equipped with a neutron imaging plate,
NIP (Haga et al., 1999), set up at the JRR-3 reactor room of
the Japan Atomic Energy Research Institute (JAERI). The
diffraction data at 293 K using perfect silicon-crystal-mono-
˚
45.0–1.0 A, and then scaled with the Scalepack program
(Otwinowski & Minor, 1997). A numerical absorption
correction by Gaussian integration was carried out using the
face indices determined by a SMART CCD diffractometer
using the ABSG program (Coppens, 1970) included in
˚
chromated neutron radiation (ꢂ = 1.51000 A) were collected
using the ! scan method with a 1.0ꢂ oscillation width and
25 min exposure time per frame. Since BIX-III is a single-axis
ꢀ
156 Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam
Acta Cryst. (2006). B62, 153–160

research papers
3.2. Structure after irradiation
PLATON (Spek, 2003). A total of 1301 independent reflec-
tions were observed.
3.2.1. X-ray diffraction analysis. The structure of crystal
(II0) was refined with the non-hydrogen atomic coordinates of
(I) as the rigid-model starting parameters. Some extra peaks
appeared on the difference electron-density map and the
peaks were assigned to a photoproduct ꢀ-thiolactam molecule
(2b). However, there were no significant peaks corresponding
to the two C atoms of the cyclohexyl ring of (2b), C20 and C70.
Therefore, we assumed that these atoms did not move. Since
the other atoms of the cyclohexyl group were close to their
initial positions, six C—C bond lengths of the group were
3. Results and discussion
Structure refinements were carried out with full-matrix least-
squares using SHELXL97 (Sheldrick, 1997b). Atomic labels
used for the structure analyses are shown in Scheme (III). A
summary of the experimental conditions and refinement
results for crystals (I), (II0) and (II) is presented in Table 1.1
Diagrams of molecular and crystal structures were produced
using ORTEP-3 for Windows (Farrugia, 1997) and some of
these were rendered using POV-Ray 3.6 (The POV-Ray Team,
2004), and the three-dimensional Fourier map was calculated
using the XFit program of XtalView (McRee, 1992) and
rendered by Raster3D (Bacon & Anderson, 1988; Merritt &
Murphy, 1994).
˚
restrained for the 1.535 (8) A listed in International Tables for
Crystallography, Vol. C. Two site-occupancy factors (SOFs) of
atoms which belong to (1b) and (2b) were refined and summed
up to exactly 1.0, except those of C2 and C7, which were fixed
at 1.0. The positional parameters of the H or D atoms were
calculated with the riding model and the isotropic ADPs were
constrained by 1.2 times those of the bonded atoms for the H
or D atoms of the phenyl and methylene groups. The final SOF
of (2b) was 0.1481 (8). Although the appearance of the crystal
did not alter after photo-irradiation, the final R values of the
photo-irradiated crystals were somewhat high, probably due
to the partial decomposition of the crystal. Fig. 3 shows the
molecular structure of (1b) and (2b) observed in crystal (II0).
The crystal structure is nearly the same as that of the non-
deuterated crystal, (1a) and (2a), reported previously (Hosoya
et al., 2002).
3.2.2. Neutron diffraction analysis. The structures of crystal
(II) were refined with SHELXL97 using the non-H-atom
coordinates of (II0) as an initial model. H and D atoms were
determined from the difference-Fourier map. The positional
parameters of the H and D atoms were refined with restraints
of the C—H/D bond distances, except those of the methylene
H atoms of the cyclohexenyl [in (1b)] and cyclohexyl [in (2b)]
groups, which were calculated using the riding model. The
restraints and constraints of the bond distances were carried
out using the values of Table 9.5.1.1 in International Tables for
3.1. Structure before irradiation
The structure of crystal (I) was solved by direct methods
using SHELXS97 and then it was refined with anisotropic
atomic displacement parameters (ADPs) for all the non-H
atoms. The H and D atoms were located in the difference
electron-density maps and all of their parameters including
the positional and isotropic atomic displacement parameters
were refined freely.
The crystal structure is nearly the same as that of the non-
deuterated crystal of (1a) reported previously (Hosoya et al.,
2002). Fig. 2 shows the molecular structure of (I) determined
by X-rays. The D atoms around C3 were D8A, D8B, D8Ai and
D8Bi, and the C3ꢄ ꢄ ꢄD distances were 2.68 (2), 4.09 (2),
˚
4.72 (2) and 4.12 (2) A, respectively [symmetry code (i)
1
2
2 ꢃ x; ꢃ þ y; ꢃz]. The D8A atom among the four D atoms
mentioned above occupies an exceptionally close position to
˚
C3 and the distance 2.68 (2) A is much shorter than the sum of
˚
the van der Waals radii of the C and D atom, 3.07 A.
1 Supplementary data for this paper are available from the IUCr electronic
archives (Reference: OG5014). Services for accessing these data are described
at the back of the journal.
Figure 2
Molecular structure (1b) in crystal (I) (before irradiation, X-rays).
Displacement ellipsoids are drawn at the 50% probability level.
ꢀ
Acta Cryst. (2006). B62, 153–160
Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam 157

research papers
to the C30 atom. In neutron diffraction, a Fourier transform of
the diffraction intensities gives the densities of the neutron
scattering lengths, which for H and D atoms are ꢃ3.739 (1)
and 6.671 (4) fm, respectively. Therefore, the H atom is
observed as a negative peak, whereas the D atom is a positive
˚
Crystallography, Vol. C [CPh—D: 1.083 (11) A, Cmethylene—H2/
˚
˚
D2: 1.092 (13) A, C3—H3: 1.077 (12) A, each standard error
was used for the restraints]. The ADPs of all the H atoms and
all the atoms of the product ꢀ-thiolactam (2b) were refined
isotropically and those of the other atoms were treated
anisotropically. As in the structure refinement of (II0), two site
occupancy factors (SOFs) of atoms which belong to (1b) and
(2b) were refined and summed up to exactly 1.0.
0
˚
peak. A blue positive peak, Q10 [Q1—C3 : 1.07 (4) A], and a
˚
red negative peak, Q2 [Q2—C3 : 1.06 (3) A], were located at
the equatorial and the axial positions of the photoproduct
cyclohexyl ring, respectively. It is reasonable to assign Q1 and
Q2 to D30 and H30, respectively. In order to examine the H/D
occupancies, we performed the refinements for three models
of the configuration of the chiral methylene, which are the D-
axial, D-equatorial and disordered models.
Fig. 4 shows a difference-Fourier density map (Fo ꢃ Fc) for
(II) around the C30 atom from neutron diffraction data. The
map was calculated with contours at the 3ꢆ level after the
refinement of all the atoms, except the H or D atoms bonded
D-axial model: This is an ordered model in which D30
occupies an axial position of the cyclohexyl group. Peaks Q1
and Q2 were relabelled as H30 and D30, respectively. The
positional parameters of all the atoms were fixed and only the
ADPs were refined. The result was not satisfactory. Uiso of H30
and D30 were 2.0000 (ceiling value in SHELXL) and 1.5064,
respectively, and the refinement did not converge at all. The
R1 value for this model was 0.0841.
D-equatorial model: This is also an ordered model in which
D30 occupies an equatorial position. The refinement condition
was the same as the D-axial model. The result was satisfactory.
Uiso of H30 and D30 were 0.065 (9) and 0.05 (1), respectively,
and the refinement converged. The R1 value for this model was
0.0792.
Disordered model: This is a static disordered model of the
two models noted above, i.e. D30(H30)—C30—H30(D30). In this
model, the SOFs of H30 and D30 were refined with the restraint
that the sum of the SOFs for the D-equatorial and D-axial
models was equal to the SOF of the atoms of (2b). The ADPs
of H30 and D30 in the D-axial model were fixed to those of D30
and H30 in the D-equatorial model, respectively. The result
was that the SOF for the D-equatorial model was 0.228 (1),
whereas the D-axial model was 0.001 (1). This indicates that
only the D-equatorial model exists predominantly in the
Figure 3
Molecular structure of (1b) and (2b) in crystal (II0) (after irradiation, X-
rays). Displacement ellipsoids are drawn at the 30% probability level.
The reactant (1b) (white bond) and the product (2b) (red bond) are
disordered, and the site occupancy factor of (2b) was 0.1481 (8).
Figure 4
Difference-Fourier map (Fo ꢃFc) around the C30 atom from neutron
diffraction data. The map was calculated with contours at the 3ꢆ level.
Positive and negative densities are shown as blue and red meshes,
respectively, in which peak tops, Q1 (blue) and Q2 (red), are shown at the
distances of 1.07 (4) and 1.06 (3) A from C3 , respectively. The peak
heights of Q1 and Q2 were at the 5 and 4ꢆ levels, respectively.
Figure 5
Molecular structure of (2b) in crystal (II) (after irradiation, neutron
diffraction). Reactant (1b) was removed for clarity. The isotropic
displacement ellipsoids of the H (white) and D (sky blue) atoms are
shown, while those of the non-H atoms are shown as small spheres. The
site-occupancy factor of (2b) was 0.2365 (9).
0
˚
ꢀ
158 Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam

research papers
3.3. Deuterium transfer mechanism
This photoreaction is assumed to be initiated from the ꢈ-
hydrogen abstraction by the C3 atom of the photoexcited
alkene. In the crystal structure of (II) the site-occupancy
factor of D30, 0.25 (1), is close to those of the other atoms of ꢀ-
thiolactam, 0.24 (1). This indicates that one of the four D
atoms around the C30 atom is extracted and transferred to the
D30 position. Although the hydrogen transfer in the Norrish
Type II reaction is considered to occur via an intramolecular
path, transfer from a neighboring molecule may be possible in
the crystalline state. In the initial crystal (I), the D atoms
around C3 are D8A, D8B, D8Ai and D8Bi, the C3ꢄ ꢄ ꢄD
˚
distances being 2.68 (2), 4.09 (2), 4.72 (2) and 4.12 (2) A,
Figure 6
Two adjacent molecules in crystal (I) are shown. One is an original
1
respectively [symmetry code: (i) 2 ꢃ x; ꢃ þ y; ꢃz]. As shown
2
in Fig. 6, the D8A and D8B atoms are on the opposite side of
the plane composed of C2, C3 and C4 from the S1 atom. The
D8Ai and D8Bi atoms, on the other hand, are located on the
same side of the plane. Fig. 7 shows two possible reaction
processes: intra- and intermolecular deuterium transfer. If
D8A or D8B may be transferred to the C3 atom within a
molecule, it should occupy the equatorial position to the
cyclohexyl ring and H3 of the original molecule may be shifted
down and placed at the axial position. On the other hand, if
D8Ai or D8Bi is transferred from the neighbouring molecule,
it may occupy the equatorial and axial positions at random.
The equatorial D atom observed by neutron diffraction
analysis clearly indicates that the photoisomerization
proceeded via intramolecular deuterium transfer, D8A or
molecule and another is a molecule related by the symmetry operation
1
2 ꢃ x; ꢃ þ y; ꢃz. Four D atoms around the C3 atom of the original
2
molecule, D8A, D8B, D8Ai and D8Bi are shown in half transparent
1
spheres [symmetry code: (i) 2 ꢃ x; ꢃ þ y; ꢃz]. The green half-
2
transparent plane shows a plane defined by the C2, C3 and C4 atoms.
˚
The distances from C3 were 2.68 (2), 4.09 (2), 4.72 (2) and 4.12 (2) A,
respectively.
disordered model. Therefore, the disordered model was
rejected.
Finally, the structure for the D-equatorial model was
refined; the results are shown in Fig. 5. The SOF of the
photoproduct ꢀ-thiolactam was 0.2365 (9), which is larger than
the value obtained for (II0) by X-rays, 0.1481 (8). Since the
small crystal (II0) was cut out from the central part of (II), the
difference between the occupancy factors may suggest that the
ꢀ-thiolactam molecules are preferentially formed at the
surface of the large single crystal. The absolute structure of
(II) was determined from the X-ray diffraction analysis of
(II0).
˚
D8B. Since the distance between D8A and C3, 2.68 (2) A, is
significantly shorter than that between D8B and C3,
˚
4.09 (2) A, which is too far for transfer to the C3, it is possible
that the D8A atom is transferred to C3 during the reaction via
the intramolecular path.
Figure 7
Reaction mechanism of the chiral thiolactam formation. Dashed and dotted arrows represent the intra- and intermolecular paths, respectively.
ꢀ
Acta Cryst. (2006). B62, 153–160
Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam 159

research papers
Merritt, E. A. & Murphy, M. E. P. (1994). Acta Cryst. D50, 869–873.
Naumov, P., Sakurai, K., Ishikawa, T., Takahashi, J., Koshihara, S. &
Ohashi, Y. (2005). J. Phys. Chem. A, 109, 7264–7275.
Ohgo, Y., Ohashi, Y., Klooster, W. T. & Koetzle, T. F. (1996a). Chem.
Lett. 25, 445–446.
Ohgo, Y., Ohashi, Y., Klooster, W. T. & Koetzle, T. F. (1996b). Chem.
Lett. 25, 579.
Ohgo, Y., Ohashi, Y., Klooster, W. T. & Koetzle, T. F. (1997).
Enantiomer, 2, 241–248.
Ohhara, T., Harada, J., Ohashi, Y., Tanaka, I., Kumazawa, S. &
Niimura, N. (2000). Acta Cryst. B56, 245–253.
Ohhara, T., Ikeda, S., Imura, H., Uekusa, H., Ohashi, Y., Tanaka, I. &
Niimura, N. (2002). J. Am. Chem. Soc. 124, 14736–14740.
Ohhara, T., Uekusa, H., Ohashi, Y., Tanaka, I., Kumazawa, S. &
Niimura, N. (2001). Acta Cryst. B57, 551–559.
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.
Pierre, J.-L. & Thomas, F. (2005). C. R. Chim. 8, 65–74.
Rigaku Corporation (1998). PROCESS-AUTO. Rigaku Corporation,
Tokyo, Japan.
4. Summary
Although the previous X-ray and other experimental studies
made clear the structural change and the absolute asymmetric
induction in the process of optically active ꢀ-thiolactam
formation from N,N-dibenzyl-1-cyclohexenecarbothioamide,
the hydrogen-transfer path associated with the isomerization
remained uncertain. In order to clarify the hydrogen-transfer
process, the H atoms of the benzyl groups were replaced with
D atoms. The structure after photoisomerization was analyzed
by neutron diffraction, using a large crystal. An absolute
configuration of the photoproduct chiral methylene,
—C*HD—, was directly determined owing to more than 20%
photo-conversion with the retention of the single-crystal form.
This indicates that the D atom was transferred via the intra-
molecular path. The role of the neutron diffraction will be
more and more important to analyze the hydrogen-transfer
mechanism in the crystalline state.
Rigaku Corporation (2004). CrystalStructure, Version 3.6. Rigaku
Corporation, Tokyo, Japan.
This work was supported by Grant-in-Aid for JSPS (Japan
Society for the Promotion of Science) Fellows.
Sakamoto, M. (2005). Top. Curr. Chem. 254, 207–232.
Sakamoto, M., Takahashi, M., Arai, W., Mino, T., Yamaguchi, K.,
Watanabe, S. & Fujita, T. (2000). Tetrahedron, 56, 6795–6804.
Sakamoto, M., Takahashi, M., Kamiya, K., Yamaguchi, K., Fujita, T.
& Watanabe, S. (1996). J. Am. Chem. Soc. 118, 10664–10665.
References
Bacon, D. J. & Anderson, W. F. (1988). J. Mol. Graphics, 6, 219–220.
Burla, M. C., Caliandro, R., Camalli, M., Carrozzini, B., Cascarano,
G. L., De Caro, L., Giacovazzo, C., Polidori, G. & Spagna, R.
(2005). J. Appl. Cryst. 38, 381–388.
¨
Sheldrick, G. M. (1997a). SHELXS97. University of Gottingen,
Germany.
Sheldrick, G. M. (1997b). SHELXL97. University of Gottingen,
¨
Germany.
Soda, K., Yoshimura, T. & Esaki, N. (2001). Chem. Rec. 1, 373–384.
Spek, A. L. (2003). J. Appl. Cryst. 36, 7–13.
Tanaka, I., Kurihara, K., Chatake, T. & Niimura, N. (2002). J. Appl.
Cryst. 35, 34–40.
Tang, C. P., Chang, H. C., Popovitz-Biro, R., Frolow, F., Lahav, M.,
Leiserowitz, L. & McMullan, R. K. (1995). J. Am. Chem. Soc. 107,
4058–4070.
The POV-Ray Team (2004). POV-Ray. The Persistence of Vision
Raytracer, Version 3.6. http://www.povray.org/.
Thomsen, I., Clausen, K., Scheibye, S. & Lawesson, S.-O. (1984). Org.
Synth. 62, 158–164.
Chang, H. C., Popovitz-Biro, R., Lahav, M. & Leiserowitz, L. (1987).
J. Am. Chem. Soc. 109, 3883–3893.
Clapham, S. E., Hadzovic, A. & Morris, R. H. (2004). Coord. Chem.
Rev. 248, 2201–2237.
Coppens, P. (1970). Crystallogr. Comput. pp. 255–270.
Enkelmann, V., Wegner, G., Novak, K. & Wagener, K. B. (1993). J.
Am. Chem. Soc. 115, 10390–10391.
Farrugia, L. J. (1997). J. Appl. Cryst. 30, 565.
Flack, H. D. (1983). Acta Cryst. A39, 876–881.
Hadjoudis, E. & Marridis, I. M. (2004). Chem. Soc. Rev. 33, 579–588.
Haga, Y. K., Kumazawa, S. & Niimura, N. (1999). J. Appl. Cryst. 32,
878–882.
Yuan, H. S. H., Stevens, R. C., Bau, R., Mosher, H. S. & Koetzle, T. F.
(1994). Proc. Natl Acad. Sci. USA, 91, 12872–12876.
Yuan, H. S. H., Stevens, R. C., Fujita, S., Watkins, M. I., Koetzle, T. F.
& Bau, R. (1988). Proc. Natl. Acad. Sci., USA, 85, 2889–2893.
Zassinovich, G. & Mestroni, G. (1992). Chem. Rev. 92, 1051–
1069.
Hosoya, T., Ohhara, T., Uekusa, H. & Ohashi, Y. (2002). Bull. Chem.
Soc. Jpn, 75, 2147–2151.
Hosoya, T., Uekusa, H., Ohashi, Y., Ohhara, T., Kimura, H. & Noda,
Y. (2003). Chem. Lett. 32, 742–743.
Ihmels, H. & Scheffer, J. R. (1999). Tetrahedron, 55, 885–907.
McRee, D. E. (1992). J. Mol. Graphics, 10, 44–46.
ꢀ
160 Takaaki Hosoya et al.
Deuterium transfer in chiral ꢀ-thiolactam
Acta Cryst. (2006). B62, 153–160