Sulfoxides: Potent co-crystal formers

Article abstract of DOI:10.1021/cg1010192

The design of co-crystals requires knowledge of robust supramolecular synthons. The sulfoxide is a potent hydrogen bond acceptor and has been used as a co-crystal former with a range of NH functional groups, via N-H...O S hydrogen bonds. The NH functional group retains favorable hydrogen bond motifs from its own structure in all cases where this is possible, with the sulfoxide interacting in a discrete, capping fashion in four cases and in a bifurcated, bridging fashion in the three other cases presented here.

Full text of DOI:10.1021/cg1010192

DOI: 10.1021/cg1010192
Sulfoxides: Potent Co-Crystal Formers
2010, Vol. 10
4243–4245
Kevin S. Eccles,† Curtis J. Elcoate,† Stephen P. Stokes,† Anita R. Maguire,‡ and
Simon E. Lawrence*,†
†Department of Chemistry, and ‡Department of Chemistry and School of Pharmacy, Analytical and
Biological Chemistry Research Facility, University College Cork, Cork, Ireland
Received August 3, 2010; Revised Manuscript Received August 13, 2010
ABSTRACT: The design of co-crystals requires knowledge of robust supramolecular synthons. The sulfoxide is a potent hydrogen
bond acceptor and has been used as a co-crystal former with a range of NH functional groups, via N-H OdS hydrogen bonds. The
3 3 3
NH functional group retains favorable hydrogen bond motifs from its own structure in all cases where this is possible, with the sulfoxide
interacting in a discrete, capping fashion in four cases and in a bifurcated, bridging fashion in the three other cases presented here.
There is great interest in co-crystals in recent years, especially
within the pharmaceutical arena.1 This is primarily because co-
crystals have the potential to alter and optimize physical proper-
ties such as crystalline form, solubility, and stability of an active
pharmaceutical ingredient (API) without detrimentally affecting
its activity.2 To date, most work has involved hydrogen bonds as
the structure determining feature,1 although recent work has
shown that weaker noncovalent interactions can also be used.3
The design of co-crystals requires knowledge of robust supra-
molecular synthons. The highly polar sulfoxide moiety, a potent
hydrogen bond acceptor,4 attracted our attention as a powerful
co-crystal former (co-former) with hydrogen bond donors. There
are few reports of the sulfoxide functional group specifically as a
co-crystal former: Nangia investigated co-crystallization of trans-
1,4-dithiane-1,4-dioxide,1i and Bernstein noted that diphenyl
sulfoxide, 1a, does not tend to form co-crystals.1e Indeed, a co-
crystal with benzidine was only achieved when water was present
in the lattice, which acted as a bridge between the N-H donor
and the sulfoxide group, with N-H O-H OdS hydrogen
bondorder, anda correspondingdecreasein the ν(SO) frequency.
Similar effects have been seen for dilute solutions of DMSO in a
variety of solvents.10 There is also an increase of 20 cm-1 in the
ν(CO) frequency in 3f, indicative of an increase in the CO bond
order.
Differential scanning calorimetry (DSC) experiments confirm
co-crystal formation with one sharp endotherm evident in all
cases.9 The mp for 3b, 3d-3g is lower than either co-crystal
former. For 3a it is only 3 °C above that of 1b, whereas for 3c it is
effectively the average of the two co-formers. In a study of APIs
Newman showed that 51% of co-crystals have a mp between the
two co-formers, whilst for 39% the mp was below that of either
co-former, 6% were higher than either co-former and 4% the
same as one co-former.11
In the solid state, 2a,9 2b,12 2c,13 2d,14 2e,15 and 2f16 form the
hydrogen-bonded R22 (8) dimers commonly observed in the solid
state for carboxylic acids, primary amides, and thioamides. The
crystalline form of 2g exhibits C(4) chains and R44 (14) tetramers
giving rise to elegant layers parallel to the bc plane.17 Polymorph-
ism is known for 2b, 2c, and 2e, with the R22 (8) dimer motif
present in all cases except the β form of 2b.12b
3 3 3
3 3 3
bonding.1e Related to these reports, research in the area of chiral
resolution has also shown that hydrogen bonding involving the
sulfoxide moiety as acceptor is possible, although the majority of
the examples to date involve alcohols and carboxylic acids.1h,5
There have been reports of dipeptides interacting with the sulf-
oxide group via hydrogen bonding; interestingly, this has in-
volved positively charged ammonium groups.6 In addition,
Kagan showed that p-tolylmethyl-sulfoxide crystallizes with a
chiral secondary amide, although the focus of this work is in
asymmetric synthesis rather than in crystal engineering.7
Hereinwedescribeco-crystalformationofsulfoxides1aand 1b
with a range of N-H containing compounds, Figure 1, and
extension to a broader series is underway, including O-H donors.8
In addition, the sulfoxide group is very poorly basic, and thus
complications due to salt formation, via complete proton transfer
from donor to an acceptor, are avoided.
In this study, co-crystals were prepared by two techniques: (i)
solid-state grinding and (ii) slow growth from the solution phase.
In all cases, a 1:1 stoichiometry of 1 and 2 respectively was
observed, except 3d, which has a 1:2 stoichiometry.9
In all cases, there is a decrease in the ν(SO) symmetric stretch-
ing frequency, from 1031 cm-1 and 1037 cm-1 for 1a and 1b,
respectively.9 Thus, in these cases IR is a viable screening tool for
monitoring co-crystal formation. The largest shift, 33 cm-1, is
seen for 3f and the smallest shift, 4 cm-1, for 3a. This shift can be
explained bythe hydrogen-bonded interaction between the donor
hydrogen and the sulfoxide oxygenleadingtoa decrease inthe SO
Interestingly, the structures of the co-crystals 3a-3f reveal
different motifs, despite the similar R22 (8) dimers observed in the
co-formers mentioned above. The co-crystals can begrouped into
two categories: (i) the R22 (8) dimers of the co-former are retained,
with the sulfoxide capping the dimers and (ii) chains or discrete
entities, with the sulfoxide capping the N-H donor, and the R22
(8) dimer is lost. Specifically, 3a-3d all retain the R22 (8) dimer. In
3a this results in a discrete 2 þ 2 complex with the sulfoxide
capping the free amide hydrogen, Figure 2.
The oxygen-based R22 (8) dimer of the carboxylic acid in 3b is
retained, and, in combination with R24 (8) rings involving the
aniline hydrogens and the sulfoxide oxygens, link the molecules
into one-dimensional chains.9
Thiourea, 2c, is well-known to form channel clathrates,18 and
the structure of 3c is fairly typical of such materials. Thus, the
sulfoxide oxygen forms a bifurcated hydrogen bond with the two
hydrogens, anti to the sulfur atom, within the same molecule,
effectively capping the side of the thiourea molecule opposite the
sulfur atom, and allowing the remaining amide hydrogens to
form a zigzag linear motif of R22 (8) rings in one dimension.9
The co-crystal 3d is the only one in this work which showed a
different stoichiometry (1:2) of sulfoxide to base: the reasons for
this are unclear. The structure shows the sulfoxide is utilizing the
amide hydrogens which are not part of the dimer motif and acting
as a bridge between crystallographically distinct dimer pairs.9
The second category, 3e-3g, involves disruption of the R22 (8)
dimer. Thus, for 3e the only hydrogen-bonded feature that is
retained from the co-former, 2e, is the well-known amide
*To whom correspondence should be addressed. E-mail: s.lawrence@
ucc.ie.
r
2010 American Chemical Society
Published on Web 08/23/2010
pubs.acs.org/crystal

4244 Crystal Growth & Design, Vol. 10, No. 10, 2010
Eccles et al.
have shown sulfoxides acting as a general co-former with a series
of NH compounds ranging from amines, (thio)amides, sulfona-
mides, thiourea etc. Critically, N-H OdS hydrogen bonding
3 3 3
is a key component of these structures, in direct contrast with
Bernstein’s observation that water was required to bridge the two
components.
Interestingly, for six of the seven co-crystals key structure
determining hydrogen bond motifs are retained from the co-
former, while this is not possible for the seventh, 3f. The R22 (8)
dimer is retained in four cases, and C(4) chains are retained in a
further two. The sulfoxide is bifurcated in three cases when the
dimer is retained and acts in a discrete fashion in all other cases.
Further work is underway to gain a fuller understanding of the
influence of the sulfoxide on the final hydrogen bonded motifs
found in the co-crystals.
Figure 1. The co-formers investigated in this work.
In conclusion, the potent hydrogen bonding acceptor ability of
sulfoxides render them excellent co-formers with a wide variety of
N-H donors. Notably, co-crystallization with sulfoxides is not
complicated by proton transfer and salt formation since they are
poorlybasic, asexemplifiedbyformation ofco-crystal 3f contain-
ing the relatively acidic saccharin, which displays a high propen-
sity for salt formation in the solid state. Preliminary results have
been obtained using sulfoxides as co-formers with phenols and
sulfonic acids, among other groups, which will be published in
due course.
Figure 2. The R22 (8) dimer in 3a, capped by the sulfoxide.
As the sulfoxide functionality is common in a significant
number of APIs, this fundamental exploration into its ability as
a co-former may well lead to improvements in drug development.
Acknowledgment. This publication has emanated from re-
search conducted with the financial support of Science Founda-
tion Ireland under Grant Nos. 08/RFP/MTR1664 (KE), 07/
SRC/B1158 (CE), and 05/PICA/B802/EC07.
Supporting Information Available: Crystallographic data of 1b,
2a, 3a-3f, Figures showing the hydrogen-bonding in 3b-3d, 3f, and
3g, and experimental details for 3a-3f. This material is available
free of charge via the Internet at http://pubs.acs.org.
Figure 3. The N-H CdO hydrogen bond leading to the C(4)
chains in 3e, with the individual units of the chain capped by the
sulfoxide.
3 3 3
References
N-H OdC C(4) chain;19 see Figure 3. The amide hydrogen
3 3 3
(1) (a) Vishweshwar, P.; McMahon, J. A.; Bis, J. A.; Zaworotko, M. J.
J. Pharm. Sci. 2006, 95, 499–516. (b) Shan, N.; Zaworotko, M. J. Drug
which is not involved in this chain is capped by the sulfoxide.
Notably, the weak hydrogen bonds between the benzylic protons
and the oxygen of dibenzyl sulfoxide, which are present in 1b,9,20
are also retained in this co-crystal. Such interactions have been
shown to be structure directing in organosulfur compounds.20,21
As expected, the R22 (8) dimer in 2f is not retained in 3f as
hydrogen bonding between the single N-H and the strong
€
Discovery Today 2008, 13, 440–446. (c) Aakeroy, C. B.; Desper, J.;
Leonard, B.; Urbina, J. F. Cryst. Growth Des. 2005, 5, 865–873.
(d) Ahn, S.; Kariuki, B. M.; Harris, K. D. M. Cryst. Growth Des. 2001,
1, 107–111. (e) Rafilovich, M.; Bernstein, J.; Hickey, M. B.; Tauber, M.
Cryst. Growth Des. 2007, 7, 1777–1782. (f) Blagden, N.; Berry, D. J.;
Parkin, A.; Javed, H.; Ibrahim, A.; Gavan, P. T.; De Matos, L. L.;
Seaton, C. C. New J. Chem. 2008, 32, 1659–1672. (g) Trask, A. V. Mol.
Pharmaceutics 2007, 4, 301–309. (h) Stahly, G. P. Cryst. Growth Des.
2009, 9, 4212–4229. (i) Kumar, V. S. S.; Nangia, A.; Katz, A. K.;
Carrell, H. L. Cryst. Growth Des. 2002, 2, 313–318.
sulfoxide acceptor overcomes the N-H OdC seen in 2f, and
3 3 3
the sulfoxide caps the N-H in a discrete fashion.9 As the CdO
bond is no longer involved in hydrogen bonding, an increase in
the CdO bond order and ν(CO) frequency is seen as mentioned
above.
(2) McNamara, D. P.; Childs, S. L.; Giordano, J.; Iarriccio, A.;
Cassidy, J.; Shet, M . S.; Mannion, R.; O’Donnell, E.; Park, A.
Pharm. Res. 2006, 23, 1888–1897.
While co-former 2g is the only compound studied which does
not exhibit a dimer motif, interestingly the structure of 3g is
ꢀ
ꢀ
(3) Fabian, L. Cryst. Growth Des. 2009, 9, 1436–1443.
(4) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43, 5310–5324.
(5) (a) Kobayashi, Y.; Soetrisno; Kodama, K.; Saigo, K. Tetrahedron:
Asymmetry 2008, 19, 295–301. (b) Fantin, G.; Fogagnolo, M.;
Bortolini, O.; Masciocchi, N.; Galli, S.; Sironi, A. New J. Chem.
2004, 28, 1295–1300. (c) Toda, F.; Tanaka, K.; Mak, T. C. W. Chem.
Lett. 1984, 2085–2088. (d) Toda, F.; Tanaka, K.; Miyamoto, H.;
Koshima, H.; Miyahara, I.; Hirotsu, K. J. J. Chem. Soc. Perkin Trans.
2 1997, 1877–1885. (e) Ishizuka, K.; Kawanami, T.; Furuno, H.;
Inanaga, J. Acta Crystallogr. 2004, E60, o607–o609.
similar to the other co-crystals. There is a N-H OdS C(4)
3 3 3
chain linking the sulfonamide molecules, which is retained from
the structure of 2e, with the sulfoxide capping the remaining
hydrogen, in a fashion similar to 3e.9
For all structures, there are no significant differences in the
SdO bonddistance between the sulfoxideco-former9,20,22 and the
sulfoxide in the co-crystals, despite the observed changes in the
ν(SO) frequencies.
As mentioned in the introduction, the sulfoxide group has been
shown to interact via hydrogen bonding with alcohols and
carboxylic acids,1h,1i,5 as well as with the positively charged ammo-
nium group.6 In view of the report by Bernstein that 1a does not
tend to form co-crystals,1e it is particularly significant that we
(6) (a) Akazome, M.; Doba, A.; Matsumoto, S.; Ogura, K. J. Org.
Chem. 2010, 75, 660–665. (b) Akazome, M.; Ueno, Y.; Ooiso, H.;
Ogura, K. J. Org. Chem. 2000, 65, 68–76.
~
(7) Kagan, H. B.; Charpin, P.; Dunach, E.; Theobald, F. R. Tetra-
hedron Lett. 1986, 27, 2989–2992. Only unit-cell dimensions are in the
CSD.

Communication
Crystal Growth & Design, Vol. 10, No. 10, 2010 4245
(8) While DMSO solvates are technically co-crystals, the common
usage has been applied in this study and the term “sulfoxide” here
refers to sulfoxides other than DMSO.
(c) Thun, J.; Seyfarth, L.; Senker, J.; Dinnebier, R. E.; Breu, J. Angew.
Chem., Int. Ed. 2007, 46, 6729–6731.
(16) Wardell, J. L.; Low, J. N.; Glidewell, C. Acta Crystallogr. E 2005,
(9) This work, see Supporting Information.
E61, o1944–o1946.
ꢁ
´
ꢁ
ꢂ
´
(10) Fawcett, W. R.; Kloss, A. A. J. Phys. Chem. 1996, 100, 2019–2024.
(11) Schultheiss, N.; Newman, A. Cryst. Growth Des. 2009, 9, 2950–2967.
(12) (a) Athimoolam, S.; Natarajan, S. Acta Crystallogr. C 2007, C63,
o514–o517. (b) Gracin, S.; Fischer, A. Acta Crystallogr. E 2005, E61,
o1242–o1244.
(17) Gowda, B. T.; Nayak, R.; Kozısek, J.; Tokarcık, M.; Fuess, H.
Acta Crystallogr. E 2007, E63, o2967.
(18) Harris, K. D. M. Chem. Soc. Rev. 1997, 26, 279–289.
(19) Jeffrey, G. A. An Introduction to Hydrogen Bonding; OUP Inc.:
New York, 1997; pp 56-78.
(13) (a) Mullen, D.; Hellner, E. Acta Crystallogr. B 1978, B34, 2789–
2794. (b) Takahashi, I.; Onodera, A.; Shiozaki, Y. Acta Crystallogr. B
1990, B46, 661–664.
(14) Altaf, M.; Stoeckli-Evans, H. Acta Crystallogr. E 2009, E65, o1894.
(15) (a) Penfold, B. R.; J. White, C. B. Acta Crystallogr. 1959, 12, 130–
135. (b) David, W. I. F.; Shankland, K.; Pulham, C. R.; Blagden, N.;
Davey, R. J.; Song, M. Angew. Chem., Int. Ed. 2005, 44, 7032–7035.
(20) Fuller, A. L.; Aitken, R. A.; Ryan, B. M.; Slawin, A. M. Z.;
Woollins, J. D. J. Chem. Crystallogr. 2009, 39, 407–415.
(21) Brondel, N.; Moynihan, E. J. A.; Lehane, K. N.; Eccles, K. S.;
Elcoate, C. J.; Coles, S. J.; Lawrence, S. E.; Maguire, A. R.
CrystEngComm 2010, DOI: 10.1039/c000371a.
(22) Casarini, D.; Lunazzi, L.; Mazzanti, A. Angew. Chem., Int. Ed.
2001, 40, 2536–2540.