T. Hosoya et al.
Bull. Chem. Soc. Jpn. Vol. 79, No. 5 (2006)
695
dered not dynamically, but statically. The non-hydrogen atoms in-
cluding the disordered ones were refined anisotropically. The hy-
drogen atoms were refined isotropically except for the hydrogen
atoms in the disordered 3,4-lutidine moiety, for which the posi-
tional parameters were calculated by the riding model and the iso-
tropic displacement parameters were constrained by 1.2 times of
those of the bonded atoms for the hydrogen atoms of the phenyl
group or 1.5 times for the hydrogen atoms of the methyl group.
A small crystal (0:30 ꢂ 0:10 ꢂ 0:10 mm3, crystal II) was cut
from crystal III after the photoirradiation described above. X-
ray diffraction measurement of II was carried out by a SMART
CCD diffractometer in the same ways as that of I. The structure
was analyzed using the positional parameters of I as an initial
model. Some extra peaks appeared on the difference electron den-
sity map around the 4-cyanobutyl group. The peaks were assigned
to the produced 1-cyanobutyl group. The structure including the
SOFs of the disordered 4- and 1-cyanobutyl groups were refined
with SHELXL-97. The SOF of the disordered 3,4-lutidine was
also refined. All non-hydrogen atoms were refined anisotropically.
Positional and isotropic temperature parameters of the hydrogen
atoms of the diphenylboryl group were refined. The other hydro-
gen atoms were calculated using riding models. The crystal data
and experimental details of I and II are summarized in Table 1.
Single-Crystal Neutron Diffraction Measurement. The pho-
toirradiated crystal of III was fixed on an aluminum pin and
mounted on the BIX-III diffractometer,19,20 equipped with a neu-
tron imaging plate,21,22 set up at the JRR-3M reactor of the Japan
Atomic Energy Research Institute (JAERI). The neutron diffrac-
tion data were collected by the ! scan method (oscillation range
ꢀ! ¼ 1:0ꢁ) at 293 K using perfect-silicon-crystal-monochromat-
Fig. 1. (a) UV–vis reflectance spectrum of I. Vertical axis
is converted from reflectance (%R) into Kubelka–Munk.
(b) The light wavelength distribution of the Xe lamp
through R-68 filter.
diffraction is very difficult to isomerize inside the crystal. If only
the surface of the crystal is isomerized, the light cannot penetrate
into the crystal because of the broken crystallinity of the surface
and the total population of the photoproduct will be very low.
The selection of the wavelengths is very important in the photore-
action of the single crystal. The ultraviolet–visible reflectance
spectra of the powdered crystal were recorded on a spectropho-
tometer (JASCO V-560) in order to optimize the irradiation con-
dition. As shown in Fig. 1, three peaks around the wavelengths of
400, 500, and 700 nm were observed, which were assigned to the
ligand-to-metal charge transfer, the singlet, and triplet d–d transi-
tions of the cobalt atom, respectively.14 In order to keep the crys-
tallinity during the photoreaction, the longer wavelengths around
700 nm due to the triplet d–d transition were selected for photo-
isomerization. A xenon lamp (SAN-EI SUPER BLIGHT 152S
with USHIO xenon short-arc lamp UXL-151DO) was placed at
a distance of 5 cm from the crystals of 1 as the light source and
two filters were inserted between the lamp and the crystals: an
R-68 filter (TOSHIBA) cutting off the light shorter than 680 nm
and a heat absorbing water filter. A large crystal of 1, crystal III,
which has the dimensions of 4:5 ꢂ 1:5 ꢂ 1:4 mm3, was selected.
Both sides of III were irradiated under the above conditions for
2 days at 0 ꢁC on a cool plate (NISSIN Cool Plate NCP-2215).
FT-IR Spectroscopy. The photoreaction was monitored using
an FT-IR spectrometer (BIO-RAD FTS3000 IR spectrometer).
KBr pellets about 1 mm thick were mounted in a pellet holder
and were irradiated with light under the above conditions at room
temperature. At appropriate intervals, successive FT-IR measure-
ments of each pellet were carried out one by one.
Single Crystal X-ray Diffraction Measurement and Struc-
ture Analysis. Single-crystal diffraction measurements were car-
ried out for a crystal of 1 before the photoirradiation. A crystal of
1 (0:30 ꢂ 0:20 ꢂ 0:10 mm3, crystal I) was mounted on a Bruker
SMART CCD X-ray diffractometer. Diffraction data were collect-
ed at 297, 173, and 93 K, using SAINT software.15 Systematic er-
ror corrections, including absorption correction, were applied us-
ing the SADABS program.16 The structure was solved by the direct
method with the SIR-2002 program.17 The structure refinement on
F2 was carried out with the full-matrix least-squares method using
the SHELXL-97 program.18 The coordinated 3,4-lutidine moiety is
disordered and the site occupancy factor (SOF) is constant with
the temperature. This means that the 3,4-lutidine moiety is disor-
˚
ed neutron radiation (ꢃ ¼ 1:51000 A). Since BIX-III is a single-
axis cylindrical diffractometer, there is a large blind region around
the rotation axis. To reduce the blind region, data were collected
by changing the angle values of the aluminum pin (about 180,
135, and 90ꢁ) instead of changing the ꢄ circle position for the or-
dinary X-ray diffractometer. Reflections were integrated with the
Denzo program23 and the data corrections without absorption cor-
rection were carried out with the Scalepack program.23 Numerical
absorption correction was done using the face indices determined
by the SMART CCD diffractometer with the ABSG program24 in
PLATON.25
Structure Refinement Using the Neutron Diffraction Data.
The positional parameters of non-hydrogen atoms were constrain-
ed using the coordinates of II as an initial model. The refinement
was on F2 against all reflections by full-matrix least-squares using
SHELXL-97. Hydrogen and deuterium atoms were observed in
difference Fourier maps (neutron-scattering-length density maps)
as negative and positive peaks.
For Atoms except 1-Cyanobutyl Group: After confirmation
of the significant peaks, the hydrogen atoms except for the 1-cy-
anobutyl group were located at the ideal positions calculated by
riding models in order to reduce the refinement parameters. The
atomic displacement parameters of all the non-hydrogen atoms
of 3,4-lutidine were refined isotropically because the two disor-
dered groups were too close to refine anisotropically, and those
of the other non-hydrogen atoms were treated anisotropically.
For 1-Cyanobutyl Group: Positional and isotropic-atomic-
displacement parameters of non-hydrogen atoms were refined.
In the difference Fourier map, the strong positive peak and some
negative peaks were found, which correspond to a transferred deu-
terium atom bonded to the methine carbon and methylene hydro-
gen atoms of the 1-cyanobutyl groups, respectively. However, the