Aripiprazole salts. II. Aripiprazole perchlorate

Article abstract of DOI:10.1107/S0108270112021348

The molecular structure of aripiprazole perchlorate (systematic name: 4-(2,3-dichlorophenyl)-1-{4-[(2-oxo-1,2,3,4-tetrahydroquinolin-7-yl)oxy]butyl} piperazin-1-ium perchlorate), C₂₃H₂₈Cl₂N ₃O_{2 +}

Full text of DOI:10.1107/S0108270112021348

organic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
diagrams are reported they usually only fulfil the role of
fingerprint identifiers. The main source of structural infor-
mation on Arip compounds consists of a paper by Tessler &
Goldberg (2006), complemented by two excellent works by
Braun et al. (2009a,b). In the first of these latter works, Braun
and coworkers report a number of different forms of the Arip
molecule in its free form, included in the Cambridge Struc-
tural Database (CSD; Allen, 2002) with refcodes MELFIT01–
05, while in the second of these latter works they present
different solvates, viz. refcodes MELFEP01 (ethanol solvate),
MELFOZ01 (methanol solvate), MELFUF01 (monohydrate)
and MOXDAF01 (1,2-dichloroethane solvate).
ISSN 0108-2701
Aripiprazole salts. II. Aripiprazole
perchlorate
Eleonora Freire,a,b*‡ Griselda Pollaa and Ricardo Baggioa
a
´
´
Gerencia de Investigacion y Aplicaciones, Centro Atomico Constituyentes,
Comision Nacional de Energıa Atomica, Buenos Aires, Argentina, and bEscuela de
With Arip salts, the situation is slightly different, and even if
some of them have been described in the patent literature, no
structure of an Arip salt had been described until recently,
namely aripiprazole nitrate, (II) (Freire et al., 2012). The
protonated state of the ligand confers interesting properties
on the structure, which prompted us to analyse other AripH+
salts. We report herein a second AripH+ salt, namely ari-
piprazole perchlorate or AripH+ꢀClO4ꢁ, (I).
´
´
´
´
´
Ciencia y Tecnologıa, Universidad Nacional General San Martın, Buenos Aires,
Argentina
Correspondence e-mail: freire@tandar.cnea.gov.ar
Received 20 April 2012
Accepted 10 May 2012
Online 18 May 2012
Coꢁmpound (I) (Fig. 1) consists of an AripH+ cation and a
ClO4 counter-ion, completing the structure and providing
charge balance. The whole AripH+ꢀClO4ꢁ molecular assembly
in (I) is very similar to that in the nitrate counterpart, (II), in
terms of bond lengths, bond angles and torsion angles, and
reference should be made to the detailed description in Freire
et al. (2012). As a measure of these similarities, the least-
squares fit of the cations of both structures leads to a mean
deviation of 0.23ꢂ (Fig. 2), and even the (relatively free)
counter-ions, not involved in the fitting, sit in almost exactly
The molecular structure of aripiprazole perchlorate
(systematic name: 4-(2,3-dichlorophenyl)-1-{4-[(2-oxo-1,2,3,4-
tetrahydroquinolin-7-yl)oxy]butyl}piperazin-1-ium perchlor-
ate), C23H28Cl2N3O2 ꢀClO4ꢁ, does not differ substantially
+
from the recently published structure of aripiprazole nitrate
[Freire, Polla & Baggio (2012). Acta Cryst. C68, o170–o173].
Both compounds have almost identical bond distances, bond
angles and torsion angles. The two different counter-ions
occupy equivalent places in the two structures, giving rise to
very similar first-order ‘packing motifs’. However, these
elemental arrangements interact with each other in different
ways in the two structures, leading to two-dimensional arrays
with quite different organizations.
˚
the same place, their baricentres lying only 0.52 A apart.
Comment
Aripiprazole (Arip, 7-{4-[4-(2,3-dichlorophenyl)-1-piperazin-
yl]butoxy}-3,4-dihydroquinolin-2(1H)-one) is an antipsychotic
drug, perhaps the most relevant representative of a modern
family of atypical antipsychotics (Travis et al., 2005), with a
different therapeutic activity from those of the classical anti-
psychotic drugs in standard use.
The drug crystallizes in a number of polymorphic and
solvatomorphic varieties, described in a large number of
patents, but the structural information provided is far from
complete, and when their X-ray powder diffraction (XRPD)
Figure 1
The asymmetric unit of (I), showing the atom-numbering scheme. Ring
centroids are labelled. Displacement ellipsoids are drawn at the 40%
probability level. The dashed line represents the intramolecular C—
Hꢀ ꢀ ꢀCl interaction.
´ ´
‡ Member of Consejo Nacional de Investigaciones Cientıficas y Tecnicas,
Conicet.
Acta Cryst. (2012). C68, o235–o239
doi:10.1107/S0108270112021348
# 2012 International Union of Crystallography o235

organic compounds
Figure 2
A comparison of the stereodisposition of perchlorate salt (I) (solid lines)
and nitrate salt (II) (dashed lines). Note the slight shift between the
counter-ions.
If the noncovalent interactions defining the spatial
arrangement of (I) (hydrogen bonds are given in Table 1 and ꢀ
bonds are given in Table 2) are compared with those in nitrate
isologue (II), it can be seen that all main interactions are
present, in particular the hydrogen bonds involving the N—H
donors (Table 1, entries 2 and 3). These hydrogen bonds give
rise to similar ‘first-order’ packing motifs [point (a) below],
which in turn act as building blocks of fairly different struc-
tures [point (b) below].
The first entry in Table 1 corresponds to an intramolecular
C—Hꢀ ꢀ ꢀCl interaction characteristic of the dichlorophenyl-
piperazin-1-yl group in all reported Arip variants, being in (I),
as it was in (II), rather unexceptional.
(a) The second entry in Table 1, the strong hydrogen bond
between the amide groups of adjacent Arip molecules, is
characteristic of most of the reported Arip variants, though
leading to different supramolecular synthons, viz. a catemer
[graph-set C(4)] or a diamide R22(8) dimer [this was discussed
in detail in Freire et al. (2012); for graph-set nomenclature, see
Bernstein et al. (1995)].
In the cases of (I) and (II), this hydrogen bond leads to
almost identical catemers (zones labelled A in Fig. 3), running
along the shortest axis in each case [b in (I) and a in (II)].
There is, however, an important structural difference between
the two cases, in that the translation symmetry operations
relating concatenated moieties in these catemers are the Pbca
a-glide plane in (II) but the P21/c 21 axis in (I), while having
Figure 3
Packing diagrams for (a) (I), projected on (401), and (b) (II), projected on
(010). Dashed lines indicate hydrogen bonds. The shaded regions A and B
correspond to different hydrogen-bonding regimes, as discussed in the
Comment. [Symmetry codes: (i) ꢁx + 2, y + , ꢁz + ; (ii) x, y ꢁ 1, z.]
1
2
3
2
though the molecular strips are common building blocks in
most Arip structures, in the present case this second inter-
action strengthens their mutual link and enhances their
internal cohesion, forming a rather different entity (Fig. 3). In
fact, zones B in Fig. 3 show the contribution of both coun-
ter-ions to the stability of the elementary packing units therein
depicted. Both include the second strong N—Hꢀ ꢀ ꢀO synthon
(Table 1, third entry), identical in both structures, and a
second, weak, C—Hꢀ ꢀ ꢀO hydrogen bond which serves as an
effective link between second neighbours in the catemer.
However, these are characteristic of each structure, involving
C11—H11 in (II) and C14—H14 in (I), a difference which is
probably due to the geometric differences between the anions
(planar nitrate versus tetrahedral perchlorate). In spite of
these differences, the ‘first-order’ packing motifs shown in
Fig. 3 are very similar.
˚
very similar translation ‘steps’ of 4.2322 (2) and 4.1552 (3) A,
respectively.
Contrasting with the previous interaction, which is common
to almost all Arip compounds, the second N—Hꢀ ꢀ ꢀO bond is
instead unique to (I) and (II), as some of the main participants
(the perchlorate/nitrate counter-ions as acceptors, and the
extra H1 atom as donor) are present only in these salts. Even
+
ꢁ
ꢃ
o236 Freire et al. C23H28Cl2N3O2 ꢀClO4
Acta Cryst. (2012). C68, o235–o239

organic compounds
Figure 4
The same views as in Fig. 3 for (a) (I) and (b) (II), but showing the lateral interactions (dashed lines) between the structures shown therein. The shaded
region C is discussed in the Comment.
(b) Finally, these motifs further interact laterally, via ꢀ–ꢀ
bonds (shaded region C in Fig. 4), but here also there are
significant crystallographic differences. In the case of (II), the
process is achieved through an (x, y, z + 1) translation of the
whole group or, in other words, the interdigitation of aromatic
rings in neighbouring cells, adjacent along c (Fig. 4b). These ꢀ–
ꢀ interactions result in the formation of a continuous thread
along [100], and the two-dimensional structure generated by
juxtaposition of the broad [100] bands ends up being an array
parallel to (010).
In contrast, the ꢀ–ꢀ linkage process in (I) is noticeably
more complicated, generated by the P21/c c glide (which
provides an [0012] translation) in combination with a further
lateral [2,0,0] shift of the resulting image (Fig. 4a and Table 2).
These two translational effects combine to give an alternating
ꢀ–ꢀ approach (Fig. 4a) and the consequence is a two-dimen-
sional structure parallel to (401). In summary, the two-
dimensional repeat pattern consists of two molecular ribbons
parallel to [100] in (II) but four ribbons parallel to [010] in (I).
There are also differences in the planarity of these two two-
dimensional arrays. When viewed in projection, that in (II)
appears noticeably undulating (Fig. 5b; largest deviation from
˚
the mean plane ꢄ5 A), consistent with the fact that the
linkage operator at A is a plane (shown as a dashed line) and
which allows for the ‘mirror-like’ aspect shown therein by the
two-dimensional trace. In contrast, the 21 axis in (I) forces a
linear disposition and thus favours a more planar set-up, with a
˚
maximum deviation from the mean plane of ꢄ2 A (Fig. 5a).
Regarding packing efficiency, the corrugated scheme in nitrate
salt (II) seems to be more efficient than the more planar
scheme in perchlorate salt (I), where a significant increase of
6% in molecular weight leads to a mere increase of 2% in
density.
The remaining interactions presented in Tables 1 and 2 are
noticeably weaker and provide interplanar interactions.
The results presented herein suggest that there are spatial
arrangements in AripH+ salts (viz. those shown in Fig. 3)
which behave as fairly stable motifs. They do not seem to
+
ꢁ
ꢃ
Acta Cryst. (2012). C68, o235–o239
Freire et al.
C23H28Cl2N3O2 ꢀClO4
o237

organic compounds
Figure 5
Packing diagrams for (a) (I) and (b) (II), at right angles to those shown in Fig. 4. The circled areas indicate the symmetry operations giving rise to the
packing linkage.
Data collection
depend on the symmetry or counter-ion, and can act as well
defined building blocks for more complex packing set-ups.
This presumed stability should be confirmed through further
systematic work on aripiprazole salts, a research line on which
we are presently engaged.
Oxford Gemini S Ultra CCD area-
detector diffractometer
Absorption correction: multi-scan
(CrysAlis PRO; Oxford
Diffraction, 2009)
11131 measured reflections
5590 independent reflections
4186 reflections with I > 2ꢄ(I)
Rint = 0.022
Tmin = 0.92, Tmax = 0.94
Refinement
R[F2 > 2ꢄ(F2)] = 0.048
wR(F2) = 0.121
S = 1.03
5590 reflections
324 parameters
H atoms treated by a mixture of
independent and constrained
refinement
Experimental
Aripiprazole (67 mg, 0.15 mmol) was dissolved in a boiling mixture of
methanol (5 ml) and acetone (0.5 ml). When dissolution was com-
plete, an excess of concentrated HClO4 was added dropwise and the
resulting solution left to cool slowly. Excellent crystals of
AripH+ꢀClO4ꢁ, (I), in the form of colourless prisms, appeared within
a few hours and were used as obtained without further recrystalli-
zation.
ꢁ3
˚
Áꢅmax = 0.39 e A
ꢁ3
˚
Áꢅmin = ꢁ0.35 e A
Table 1
Hydrogen-bond geometry (A, ).
ꢂ
˚
Cg2 is the centroid of the C15–C20 benzene ring.
D—Hꢀ ꢀ ꢀA
D—H
Hꢀ ꢀ ꢀA
Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA
Crystal data
+
ꢁ
3
C3—H3Bꢀ ꢀ ꢀCl1
N3—H3ꢀ ꢀ ꢀO1i
C14—H14Aꢀ ꢀ ꢀO23ii
C1—H1Aꢀ ꢀ ꢀO23iii
C4—H4Bꢀ ꢀ ꢀCg2iii
0.97
2.57
3.193 (2)
2.985 (3)
2.875 (3)
3.413 (4)
3.441 (4)
3.541 (2)
122
˚
V = 2466.03 (16) A
Z = 4
Mo Kꢂ radiation
ꢃ = 0.42 mmꢁ1
T = 291 K
C23H28Cl2N3O2 ꢀClO4
0.81 (3)
0.83 (3)
0.97
0.97
0.97
2.20 (3)
2.05 (3)
2.45
2.51
2.81
162 (2)
174 (3)
175
160
133
Mr = 548.83
Monoclinic, P21=c
N1—H1ꢀ ꢀ ꢀO33
˚
a = 14.7348 (5) A
˚
b = 8.3103 (3) A
˚
c = 20.1590 (7) A
0.24 ꢅ 0.16 ꢅ 0.14 mm
ꢁ = 92.557 (3)ꢂ
1
2
3
2
Symmetry codes: (i) ꢁx þ 2; y þ ; ꢁz þ ; (ii) x; y ꢁ 1; z; (iii) ꢁx þ 1; ꢁy þ 1, ꢁz þ 2.
+
ꢁ
ꢃ
o238 Freire et al. C23H28Cl2N3O2 ꢀClO4
Acta Cryst. (2012). C68, o235–o239

organic compounds
Table 2
ANPCyT (project No. PME 2006-01113) for the purchase
of the Oxford Gemini CCD diffractometer, and the Spanish
Research Council (CSIC) for the provision of a free-of-charge
licence to the Cambridge Structural Database (Allen,
2002).
ꢂ
˚
ꢀ–ꢀ and halogen–ꢀ contacts (A, ).
Cg1 is the centroid of the C5–C10 benzene ring, ccd is the centroid-to-centroid
distance, cpd is the centroid-to-plane distance and sa is the slippage angle
(angle subtended by the intercentroid vector to the plane normal). For details,
see Janiak (2000).
˚
ccd (A)
˚
cpd (A)
Group 1/Group 2
sa (ꢂ)
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: SK3436). Services for accessing these data are
described at the back of the journal.
Cg1ꢀ ꢀ ꢀCg1i
3.7751 (14)
3.5481 (12)
3.536
3.474
20.48
11.75
Cl1ꢀ ꢀ ꢀCg1i
Symmetry code: (i) ꢁx, ꢁy, ꢁz + 2.
References
All H atoms were found in a difference map. H atoms attached to
˚
N atoms were further refined, giving N—H = 0.81 (3)–0.83 (3) A and
Allen, F. H. (2002). Acta Cryst. B58, 380–388.
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew. Chem.
Int. Ed. Engl. 34, 1555–1573.
Braun, D. E., Gelbrich, T., Kahlenberg, V., Tessadri, R., Wieser, J. & Griesser,
U. (2009a). J. Pharm. Sci. 98, 2010–2026.
Braun, D. E., Gelbrich, T., Kahlenberg, V., Tessadri, R., Wieser, J. & Griesser,
U. (2009b). Cryst. Growth Des. 9, 1054–1065.
Freire, E., Polla, G. & Baggio, R. (2012). Acta Cryst. C68, o170–o173.
Janiak, C. (2000). J. Chem. Soc. Dalton Trans. pp. 3885–3896.
Oxford Diffraction (2009). CrysAlis PRO. Oxford Diffraction Ltd, Yarnton,
Oxfordshire, England.
2
Uiso(H) = 0.037 (7)–0.040 (7) A . H atoms attached to C atoms were
˚
˚
idealized and allowed to ride, with methylene C—H = 0.97 A and
˚
aromatic C—H = 0.93 A, and with Uiso(H) = 1.2Ueq(C).
Data collection: CrysAlis PRO (Oxford Diffraction, 2009); cell
refinement: CrysAlis PRO; data reduction: CrysAlis PRO;
program(s) used to solve structure: SHELXS97 (Sheldrick, 2008);
program(s) used to refine structure: SHELXL97 (Sheldrick, 2008);
molecular graphics: SHELXTL (Sheldrick, 2008); software used to
prepare material for publication: SHELXL97 and PLATON (Spek,
2009).
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122.
Spek, A. L. (2009). Acta Cryst. D65, 148–155.
Tessler, L. & Goldberg, I. (2006). J. Inclusion Phenom. Macrocycl. Chem. 55,
255–261.
Travis, M. J., Burns, J., Dursun, S., Faby, T., Frangou, S., Gray, R., Haddad, P. M.,
Hunter, R., Taylor, D. M. & Young, A. H. (2005). Int. J. Clin. Pract. 59, 485–
495.
The provision of aripiprazole by Laboratorios Maprimed is
gratefully acknowledged. The authors also acknowledge
+
ꢁ
ꢃ
Acta Cryst. (2012). C68, o235–o239
Freire et al.
C23H28Cl2N3O2 ꢀClO4
o239