Rac-5-Diphenylacetyl-2,2,4-trimethyl-2,3,4,5-tetrahydro-1, 5-benzothia-zepine and rac-5-formyl-2,2,4-tri-methyl-2,3,4,5-tetra-hydro-1,5- benzo-thia-zepine

Article abstract of DOI:10.1107/S0108270109036646

rac-5-Diphenyl-acetyl-2,2,4-trimethyl-2,3,4,5-tetra-hydro-1, 5-benzothia-zepine, C₂₆H₂₇NOS, (I), and rac-5-formyl-2,2,4-tri-methyl-2,3,4,5-tetra-hydro-1,5-benzothia-zepine, C ₁₃H₁₇NOS, (II), are both characteriz

Full text of DOI:10.1107/S0108270109036646

organic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
and 4 positions are potent antibacterial agents (Micheli et al.,
2001). Therefore, the configuration and conformation of 1,5-
benzothiazepines are of interest and importance.
ISSN 0108-2701
rac-5-Diphenylacetyl-2,2,4-trimethyl-
2,3,4,5-tetrahydro-1,5-benzo-
thiazepine and rac-5-formyl-2,2,4-tri-
methyl-2,3,4,5-tetrahydro-1,5-benzo-
thiazepine
Thanikasalam Kanagasabapathy, Panchanatheswaran
Krishnaswamy* and Jeyaraman Ramasubbu
School of Chemistry, Bharathidasan University, Tiruchirappalli, Tamil Nadu
620 024, India
Correspondence e-mail: panch_45@yahoo.co.in
Received 31 July 2009
The conformational effects of substituents in the seven-
membered ring are pronounced, as indicated by the existence
of the chair conformation in 2,2,4-trimethyl-2,3,4,5-tetra-
hydro-1,5-benzothiazepine, (III) (Muthukumar et al., 2004),
and the twist-boat conformation in 2,4-diphenyl-2,3,4,5-tetra-
hydrobenzothiazepine (Laavanya et al., 2002). Subtle elec-
tronic effects can introduce distortions in the ideal
conformations, viz. chair, twist-chair, boat and twist-boat. In
the solid state, the conformation of rac-5-diphenylacetyl-2,2,4-
trimethyl-2,3,4,5-tetrahydro-1,5-benzothiazepine, (I), is a dis-
torted twist-boat and that of rac-5-formyl-2,2,4-trimethyl-
2,3,4,5-tetrahydro-1,5-benzothiazepine, (II), is intermediate
between boat and twist-boat forms. The molecules of
compounds (I)–(III) are all chiral, raising the possibility that
they could crystallize in an enantiopure form in noncen-
trosymmetric space groups, as required for nonlinear optical
properties (Long, 1995). All three compounds were obtained
by synthetic methods and are racemates. However, only (III)
crystallizes in a noncentrosymmetric space group, viz. P21
(Muthukumar et al., 2004).
Accepted 10 September 2009
Online 17 October 2009
rac-5-Diphenylacetyl-2,2,4-trimethyl-2,3,4,5-tetrahydro-1,5-
benzothiazepine, C26H27NOS, (I), and rac-5-formyl-2,2,4-tri-
methyl-2,3,4,5-tetrahydro-1,5-benzothiazepine, C13H17NOS,
(II), are both characterized by a planar configuration around
the heterocyclic N atom. In contrast with the chair conforma-
tion of the parent benzothiazepine, which has no substituents
at the heterocyclic N atom, the seven-membered ring adopts a
boat conformation in (I) and a conformation intermediate
between boat and twist-boat in (II). The molecules lack a
symmetry plane, indicating distortions from the perfect boat
or twist-boat conformations. The supramolecular architectures
are significantly different, depending in (I) on C—Hꢀ ꢀ ꢀO
interactions and intermolecular Sꢀ ꢀ ꢀS contacts, and in (II) on a
single aromatic ꢀ–ꢀ stacking interaction.
Comment
The molecules of these compounds could, in principle, be
linked in the solid state by one or more noncovalent forces,
viz. C—Hꢀ ꢀ ꢀX, N—Hꢀ ꢀ ꢀX (X = O or S), Sꢀ ꢀ ꢀS, C—Hꢀ ꢀ ꢀꢀ and
ꢀꢀ ꢀ ꢀꢀ interactions. The presence or absence of these inter-
actions can affect the molecular conformation and supra-
molecular structure of these crystals. We report here the
crystal and molecular structures of (I) and (II) and compare
them with those of (III). Analysis of the Cambridge Structural
Database (CSD, Version 5.29; Allen, 2002) reveals 75 reported
crystal structures of 1,5-benzothiazepine derivatives, six of
which belong to the diltiazem family, and 11 1,5-benzothia-
zepines which contain neither ring oxo groups nor extra fused
rings, which are akin to compounds (I)–(III).
1,5-Benzothiazepine is a versatile pharmacophore found in a
number of clinically used drugs. The biological function of
such drugs is quite varied, and ranges from calcium antagonist
activity observed for diltiazem, (+)-cis-3-acetoxy-5-(2,2-
dimethylaminoethyl)-2-(4-methoxyphenyl)-2,3-dihydro-1,5-
benzothiazepin-4(5H)-one, and clentiazem, (+)-cis-3-acetoxy-
8-chloro-5-(2,2-dimethylaminoethyl)-2-(4-methoxyphenyl)-
2,3-dihydro-1,5-benzothiazepin-4(5H)-one, to CNS-depres-
sant behaviour observed for thiazesim, (+)-5-(2,2-dimethyl
aminoethyl)-2-phenyl-2,3-dihydro-1,5-benzothiazepin-4(5H)-
one. Replacement of the 3-acetyl group of diltiazem with a
methyl group gives a more potent analogue of diltiazem.
Further, TA 933, which has an inverted stereochemistry with
respect to the substituents at the 2 and 3 positions of diltiazem,
is a more active vasorelaxant (Bariwal et al., 2008). Dihydro-
1,5-benzothiazepines containing phenyl substituents at the 2
Compounds (I) and (II) crystallize as racemates and for
each compound the reference molecules were selected to have
an R configuration at C4 (Figs. 1 and 2). The bond lengths and
angles of (I) and (II) are unexceptional. As expected, the
Acta Cryst. (2009). C65, o579–o582
doi:10.1107/S0108270109036646
# 2009 International Union of Crystallography o579

organic compounds
C—S bond lengths in (I) are unequal [1.7603 (16) and
discern the conformation of the seven-membered ring for each
of the molecules are given in Table 2. Detailed puckering
analysis of the seven-membered ring on the basis of the four
puckering parameters indicates six conformations derived
from the four fundamental conformations discussed above.
They are contained in three pseudorotational manifolds, viz.
chair–twist-chair, boat–twist-boat and sofa–twist-sofa–sofa–
boat forms. The puckering amplitudes q2 and q3 for (I) and (II)
are close to 1.15 and 0, respectively, which are the values
ascribed to both boat and twist-boat forms (Boessenkool &
Boeyens, 1980) of the seven-membered ring. From the puck-
˚
˚
1.8518 (18) A], and in (II) [1.7577 (15) and 1.8536 (13) A].
The means of the bond angles around the ring N atom are
119.98 (8) and 119.82 (6)ꢁ for (I) and (II), respectively,
signifying planarity of the N atom, as expected. The C4—C3—
C2—C12 and C4—C3—C2—C13 torsion angles of (I) are
ꢂ177.45 (19) and ꢂ55.2 (2)ꢁ, respectively, and the corre-
sponding torsion angles for (II) are ꢂ173.65 (12) and
ꢂ50.89 (17)ꢁ, respectively, indicating in both molecules that
the two methyl groups are not isoclinal but assume equatorial
and axial orientations. The C2—C3—C4—C14 torsion angles
of (I) and (II) are 178.41 (15) and ꢂ179.07 (11)ꢁ, respectively,
confirming the equatorial orientations of the methyl group
bonded to C4. Additional selected torsion angles for (I)–(III)
are listed in Table 1.
In each of (I) and (II), the carbonyl group is exo oriented
with respect to the N5—C6 bond and the benzene ring. The S
and N atoms are coplanar with the benzene ring in molecule
(I), while in (II) and (III), although the N atom is essentially
coplanar with the benzene ring, the S atom deviates signifi-
˚
˚
cantly from the ring plane [0.227 (1) A in (II) and 0.211 (1) A
in (III)]. Table 1 indicates the absence of a mirror plane in all
the molecules and the considerable distortion from the four
distinct conformations of the seven-membered ring. Compar-
ison of the standard values of all the four forms (Hendrickson,
1967) revealed that the conformations of (I) and (II) are
clearly nonchair and that of (III) is chair. A transition from
chair to boat form occurs as a result of N-acylation.
Figure 2
Ring-puckering parameters can be used to infer the
conformational preferences of the molecules (Cremer &
Pople, 1975). The four ring-puckering parameters required to
The molecule of (II), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level and H atoms are shown
as small spheres of arbitrary radii.
Figure 1
The molecule of (I), showing the atom-labelling scheme. Displacement
ellipsoids are drawn at the 30% probability level and H atoms are shown
as small spheres of arbitrary radii.
Figure 3
Part of the crystal structure of (I), showing the hydrogen-bonded chain
along [010]. For clarity, H atoms bonded to C atoms have been omitted.
ꢃ
o580 Kanagasabapathy et al.
C26H27NOS and C13H17NOS
Acta Cryst. (2009). C65, o579–o582

organic compounds
Figure 4
Part of the crystal structure of (I), showing the Sꢀ ꢀ ꢀS contacts (green/grey
in the electronic version of the paper) between the hydrogen-bonded
chains. For clarity, most H atoms bonded to C atoms have been omitted.
Figure 6
Part of the crystal structure of (II), showing the formation of a
centrosymmetric ꢀ-stacked dimer. For clarity, all H atoms have been
omitted. The S atom marked with an asterisk (*) is at the symmetry
position (ꢂx + 1, ꢂy + 1, ꢂz).
are linked by an N—Hꢀ ꢀ ꢀS hydrogen bond (Muthukumar et
al., 2004). It is noted that the three molecules are in different
conformations and exhibit differences in intermolecular
interactions. The packing indices (%) follow the order (III)
68.6 > (II) 67.2 > (I) 64.5, signifying a lack of close packing in
(I) compared with the other two. This is substantiated by the
lowest density for (I) despite its high molecular weight; the
density (in Mg mꢂ3) order is (II) 1.255 > (III) 1.219 > (I) 1.182.
The melting points (K), which depend upon crystal packing
forces, follow the order (I) 398 > (III) 358 > (II) 337. The
melting point of (II) is low, much lower than (III), indicating
weaker intermolecular interactions in the former. As shown
by AM1 calculations, the stable conformations of isolated
molecules of (I)–(III) are similar to those established through
X-ray analysis. We conclude that the conformations of the
three benzothiazepines depend upon their molecular struc-
tures and crystal packing effects cause subtle distortions in
conformation. We further conclude that in the structures of
the crystals studied here the packing interactions are deter-
mined by the molecular conformations.
Figure 5
Part of the crystal structure of (I), showing the relative orientation of
bonds around adjacent S atoms. The S atom marked with an asterisk (*) is
at the symmetry position (ꢂx + 1, ꢂy + 1, ꢂz + 1).
ering angles ’2 and ’3 (Table 2), the conformation of (I) is
found to be twist-boat, and that of (II) is intermediate
between the twist-boat and boat forms.
In (I), molecules are linked by a C—Hꢀ ꢀ ꢀO hydrogen bond
(Table 3), forming a chain parallel to the [010] axis (Fig. 3).
˚
There are short Sꢀ ꢀ ꢀS contacts [3.376 (6) A] between adjacent
antiparallel chains, involving the S atoms at (x, y, z) and (1 ꢂ x,
1 ꢂ y, 1 ꢂ z) (Fig. 4). The pair of bonds around the S atom are
uneclipsed with respect to those of the other S atom, signifying
a nonparallel orientation (Fig. 5) of the S-atom p-type lone
pairs, promoting close contacts between the S-atom s-type
lone pairs (Ozturk et al., 1994).
The structure of (II) contains neither Sꢀ ꢀ ꢀS nor C—Hꢀ ꢀ ꢀO
interactions; instead, the molecules are linked by an aromatic
ꢀ–ꢀ interaction. The aryl rings in the molecules at (x, y, z) and
(1 ꢂ x, 1 ꢂ y, ꢂz) are parallel with an interplanar spacing of
Experimental
Compound (I) was obtained by the acetylation of the parent 2,2,4-
trimethyl-2,3,4,5-tetrahydrobenzothiazepine, (III), using diphenyl-
acetyl chloride in a 1:1 molar ratio in the presence of triethylamine in
a benzene medium. The product was isolated as a single crystal by
slow evaporation from a solution in 95% ethanol (yield 54%, m.p.
398–400 K). Compound (II) was prepared by the formylation of (III)
in a benzene medium using the same base and acetic formic anhy-
dride as formylating agent, adopting the reported procedure of
˚
˚
3.5390 (6) A. The ring–centroid distance is 3.8597 (9) A.
These parameters are comparable with the corresponding
values reported for similar interactions (Portilla et al., 2005;
Delgado et al., 2006). This interaction generates a centro-
symmetric dimer (Fig. 6). By contrast, the molecules of (III)
ꢃ
Acta Cryst. (2009). C65, o579–o582
Kanagasabapathy et al. C26H27NOS and C13H17NOS o581

organic compounds
Muthukumar (2001). The latter was prepared in situ by warming a
mixture of acetic anhydride and 85% formic acid in a 1:1 molar ratio.
Repeated recrystallization by the slow and spontaneous evaporation
of the solvent at room temperature from a solution in petroleum
ether afforded colourless single crystals (yield 89%, m.p. 337–339 K).
Table 1
Selected torsion angles (ꢁ) for (I)–(III).
(I)
(II)
8.26 (17)
(III)
N5—C6—C7—S1
C6—C7—S1—C2
C7—S1—C2—C3
S1—C2—C3—C4
C2—C3—C4—N5
C3—C4—N5—C6
C4—N5—C6—C7
3.51 (19)
ꢂ61.14 (14)
17.01 (15)
ꢂ6.53 (19)
ꢂ57.55 (12)
62.97 (13)
ꢂ70.33 (13)
63.33 (14)
7.64 (11)
63.58 (19)
68.82 (13)
ꢂ55.25 (15)
ꢂ44.92 (15)
74.34 (15)
Compound (I)
ꢂ57.90 (19)
ꢂ41.69 (18)
78.68 (17)
ꢂ67.82 (17)
90.42 (16)
ꢂ69.68 (18)
Crystal data
3
˚
C26H27NOS
Mr = 401.55
V = 2257.13 (9) A
Z = 4
Monoclinic, P21=c
Mo Kꢂ radiation
ꢃ = 0.16 mmꢂ1
T = 296 K
0.30 ꢄ 0.02 ꢄ 0.02 mm
Table 2
Ring-puckering parameters (A, ) for (I)–(III).
˚
a = 10.2731 (2) A
ꢁ
˚
˚
b = 8.5637 (2) A
˚
c = 25.6968 (6) A
(I)
(II)
(III)
ꢁ = 93.2180 (10)ꢁ
q2
q3
1.1591 (14)
0.0650 (15)
151.43 (8)
168.6 (14)
1.1111 (12)
0.0313 (12)
146.09 (7)
129.0 (2)
0.4783 (14)
0.7083 (14)
332.25 (18)
44.74 (12)
Data collection
’
2
Bruker Kappa APEXII CCD
area-detector diffractometer
Absorption correction: multi-scan
(SADABS; Bruker, 1999)
25865 measured reflections
5164 independent reflections
3691 reflections with I > 2ꢄ(I)
Rint = 0.031
’
3
Tmin = 0.954, Tmax = 0.997
Table 3
Hydrogen-bond geometry (A, ) for (I).
ꢁ
˚
Refinement
R[F2 > 2ꢄ(F2)] = 0.042
wR(F2) = 0.117
S = 1.03
289 parameters
H-atom parameteꢂrs3 constrained
D—Hꢀ ꢀ ꢀA
C10—H10ꢀ ꢀ ꢀO16i
Symmetry code: (i) x; y þ 1; z.
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
˚
0.95
2.43
3.263 (2)
146
Áꢅmax = 0.23 e A
ꢂ3
˚
5164 reflections
Áꢅmin = ꢂ0.31 e A
Compound (II)
PLATON (Spek, 2009); plane calculations: PARST in WinGX
(Farrugia, 1999); software used to prepare material for publication:
PLATON.
Crystal data
3
˚
V = 1245.23 (18) A
Z = 4
C13H17NOS
Mr = 235.34
Monoclinic, P21=n
Mo Kꢂ radiation
ꢃ = 0.24 mmꢂ1
T = 296 K
0.29 ꢄ 0.12 ꢄ 0.10 mm
˚
a = 9.6038 (8) A
˚
b = 9.9485 (8) A
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: GD3302). Services for accessing these data are
described at the back of the journal.
˚
c = 13.6818 (11) A
ꢁ = 107.713 (2)ꢁ
Data collection
References
Bruker Kappa APEXII CCD
area-detector diffractometer
Absorption correction: multi-scan
(SADABS; Bruker, 1999)
17043 measured reflections
4056 independent reflections
2895 reflections with I > 2ꢄ(I)
Rint = 0.024
Allen, F. H. (2002). Acta Cryst. B58, 380–388.
Bariwal, J. B., Upadhyay, K. D., Manvar, A. T., Trivedi, J. C., Singh, J. S., Jain,
K. S. & Shah, A. K. (2008). Eur. J. Med. Chem. 43, 2279–2290.
Boessenkool, I. K. & Boeyens, J. C. A. (1980). J. Cryst. Mol. Struct. 10, 11–18.
Bruker (1999). SADABS. Bruker AXS Inc., Madison, Wisconsin, USA.
Bruker (2004). APEX2, SAINT-Plus and XPREP. Bruker AXS Inc., Madison,
Wisconsin, USA.
Cremer, D. & Pople, J. A. (1975). J. Am. Chem. Soc. 97, 1354–1358.
Delgado, P., Quiroga, J., de la Torre, J. M., Cobo, J., Low, J. N. & Glidewell, C.
(2006). Acta Cryst. C62, o382–o385.
Tmin = 0.934, Tmax = 0.977
Refinement
R[F2 > 2ꢄ(F2)] = 0.041
wR(F2) = 0.124
S = 1.06
4056 reflections
165 parameters
H atoms treated by a mixture of
independent and constrained
refinement
ꢂ3
˚
Áꢅmax = 0.36 e A
Farrugia, L. J. (1999). J. Appl. Cryst. 32, 837–838.
ꢂ3
˚
Áꢅmin = ꢂ0.38 e A
Hendrickson, J. B. (1967). J. Am. Chem. Soc. 89, 7043–7046.
Laavanya, P., Panchanatheswaran, K., Muthukumar, M., Jeyaraman, R. &
Bauer, J. A. K. (2002). Acta Cryst. E58, o701–o702.
The formyl H atom in (II) was located in a difference map and then
Long, M. J. (1995). Angew. Chem. Int. Ed. Engl. 34, 21–38.
Micheli, F., Degiorgis, F., Feriani, A., Paio, A., Pozzan, A., Zarantonello, P. &
Seneci, P. (2001). J. Comb. Chem. 3, 224–228.
Muthukumar, M. (2001). PhD thesis, Bharathidasan University, India.
Muthukumar, M., Thanikasalam, K., Mohamed, E. M., Low, J. N. & Glidewell,
C. (2004). Acta Cryst. C60, o346–o349.
˚
freely refined, giving a C—H distance of 1.009 (18) A. All other H
atoms were treated as riding atoms, with C—H = 0.95 (aromatic), 0.98
˚
(CH3), 0.99 (CH2) or 1.00 A (aliphatic CH). Individual isotropic
displacement parameters were refined for all H atoms.
Ozturk, T., Povep, D. C. & Wallis, J. D. (1994). Tetrahedron, 50, 11205–
11212.
Portilla, J., Quiroga, J., Cobo, J., Nogueras, M., Low, J. N. & Glidewell, C.
(2005). Acta Cryst. C61, o398–o403.
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122.
For both compounds, data collection: APEX2 (Bruker, 2004); cell
refinement: APEX2 and SAINT (Bruker, 2004); data reduction:
SAINT and XPREP (Bruker, 2004); program(s) used to solve
structure: SHELXS97 (Sheldrick, 2008); program(s) used to refine
structure: SHELXL97 (Sheldrick, 2008); molecular graphics:
Spek, A. L. (2009). Acta Cryst. D65, 148–155.
ꢃ
o582 Kanagasabapathy et al.
C26H27NOS and C13H17NOS
Acta Cryst. (2009). C65, o579–o582