Table 1 Crystal data and structure refinement for 1 and 2
1
2
Empirical formula
C19H23ClN2O3
362.84
P21
C19H23ClN2O3
362.84
P21
Formula mass/g molϪ1
Space group
a/Å
b/Å
c/Å
5.728(2)
20.671(3)
8.354(1)
107.39(2)
943.9(4)
2
1.277
1.269
0.222
384
14.344(3)
5.474(1)
14.434(2)
114.75(2)
1029.4(3)
2
1.171
—
0.204
384
β/Њ
V/Å3
Z
Dc/g cmϪ3
Dm/g cmϪ3
Absorption coefficient/mmϪ1
F(000)
Data collection (24 ЊC)
Crystal size/mm
0.4 × 0.5 × 0.3
0.3 × 0.3 × 0.3
Range scanned, θ/(Њ)
Index ranges
1.97–24.97
h: ±6; k: Ϫ24; 0; l: 0, 9
1837
1703 [R(int) = 0.0301]
1703/6/253
R1 = 0.0296, wR2 = 0.0938
R1 = 0.0387, wR2 = 0.1075
Ϫ0.12(9)
1.55–24.99
h: ±15; k: 0, 6; l: 0, 17
2110
2026 [R(int) = 0.0383]
2026/0/138
R1 = 0.1258, wR2 = 0.3111
R1 = 0.2244, wR2 = 0.3134
0.0(5)
Reflections collected
Independent reflections
Data/restraints/parameters
Final R indices [I > 2σ(I)]
R indices (all data)
Flack parameter x
Largest diff. peak and hole/e ÅϪ3
0.202 and Ϫ0.236
0.499 and Ϫ0.345
Single crystal X-ray analyses
Resolution of (±)GAM
Single crystals of 1 and 2 were grown from MeOH and MeOH–
acetone (9:1), respectively. Intensity data were measured on an
Enraf-Nonius CAD4 diffractometer at 294 K using graphite-
monochromated Mo-Kα radiation (λ = 0.710 69 Å). Accurate
cell dimensions were obtained by least-squares refinement of 24
accurately measured reflections. Data in the range 1Њ р θ р 25Њ
were collected in the ω–2θ scan mode with a maximum
recording time of 40 s per reflection. Intensity control was per-
formed every hour by means of three reference reflections and
orientation control every 200 reflections. Measured intensities
were corrected by Lorentz–polarisation factors as well as an
empirical absorption correction.14
(±)GAM (18 g, 74.4 mmol) was dissolved in methanol (250 ml)
and warmed to approximately 60 ЊC. SPEA (9.48 ml, 74.4
mmol) was carefully added via a graduated pipette, and the
warm solution stirred for a few minutes. The resulting 1:1
(±)GAM–SPEA solution was cooled to room temperature and
left to stand in the dark for approximately 24 h affording col-
ourless plate-like crystals of 1 (9.86 g; 73%), mp 187–197 ЊC,
[α]2D4 = ϩ5.5 (c 1.0 in MeOH). The crystals were isolated by fil-
tration, dried and dissolved in water (200 ml) by heating to
approximately 90 ЊC. 3 HCl (90.5 ml) was added at this tem-
perature after which the solution was allowed to cool to room
temperature. The precipitate was filtered, washed with water
and vacuum-dried to give RGAM (6.23 g, 95%), mp 173 ЊC
(Found: C, 54.62; H, 4.98; N, 5.72. C11H12ClNO3 requires C,
54.67; H, 4.96; N, 5.79%); [α]2D4 = ϩ9.5 (c 1.0 in MeOH). The
mother liquor of the crystallisation was isolated and concen-
trated by evaporation. Any further precipitate was removed
from the solution by filtration (ca. 2 g salt). The filtrate evapor-
ated to dryness and the contaminated (SPEA)(SGAM) 2 salt
was recrystallised twice from methanol–acetone (9:1) (7.1 g,
53%), mp 185–190 ЊC; [α]2D4 = Ϫ12.3 (c 1.0 in MeOH). This salt,
when acidified in water, yielded SGAM, [α]2D4 = Ϫ9.5 (c 1.0 in
MeOH).
The structures were solved by direct methods using the pro-
gram SHELX-8615 and refined on F2 by full-matrix least-
squares techniques using SHELX-93.16 Polar axis restraints
after Flack and Schwarzenbach 17 were used to define the origin
of the polar space group P21. For 1 all non-hydrogen atoms
were located in difference electron density syntheses and refined
ϩ
anisotropically. All hydrogen atoms, except those of the ᎐NH3
group, were located and placed in geometrically generated posi-
tions and refined with positional parameters riding on the par-
ϩ
ent atom (C᎐H = 1.00 Å). The ᎐NH3 H atoms were refined
with bond length restraints only. Chemically equivalent hydro-
gen atoms were tied to common isotropic thermal parameters
which were allowed to refine. No hydrogen atoms could be
located in 2 and these were therefore omitted from the model.
All phenyl rings were refined as idealised hexagons with com-
mon variable isotropic thermal parameters. Refinement with
the carbon atoms of ring A included gave rise to peaks in the
electron density surrounding the phenyl ring suggesting an
alternative orientation (ring B) inclined at 72Њ to the plane of
ring A (Fig. 4). This was modelled by allowing two possible
orientations of the phenyl ring whose site occupancy factors
were refined to 0.52 and 0.48 for rings A and B, respectively.
Details of the crystal data and refinement parameters appear in
Table 1.
Atomic coordinates, bond lengths and angles, and thermal
parameters have been deposited at the Cambridge Crystallo-
graphic Data Centre (CCDC). For details of the deposition
scheme, see ‘Instructions for Authors’, J. Chem. Soc., Perkin
Trans. 2, 1997, Issue 1. Any request to the CCDC for this
material should quote the full literature citation and the refer-
ence number 188/57.
(R)(Ϫ)-Baclofen
The conversion of RGAM to (R)-(Ϫ)-baclofen was effected by
the Hofmann reaction using the experimental details previously
reported.1 RGAM (6.57 g, 27.21 mmol) was added to a solution
of sodium hydroxide (2.8 g, 69.4 mmol) in water (50 ml) at
24 ЊC. A 10.8% aqueous solution of sodium hypochlorite (28.3
g) was then added over 2.5 h at 0 ЊC via a dropping funnel. The
solution was stirred for a further 12 h at room temperature
(approximately 24 ЊC) after which time it was carefully neutral-
ised (pH 7.5) with dilute hydrochloric acid. The precipitate was
filtered off at the pump and washed with water. The precipitate
was boiled in methanol to remove the last traces of RGAM and
filtered to yield pure (R)-(Ϫ)-baclofen (1.30 g, 22% yield). The
mother liquors were reduced in volume in vacuo to yield more
product which was boiled in methanol and filtered to give
another 2.05 g of (R)-(Ϫ)-baclofen (3.35 g total, 57%), mp 205–
208 ЊC (Found: C, 56.30; H, 5.78; N, 6.52. C10H12ClNO2
requires C, 56.21; H, 5.66; N, 6.56%). The HPLC procedure
described above had a 0.2% detection limit for (S)-(ϩ)-
764
J. Chem. Soc., Perkin Trans. 2, 1997