Hydrogen bonded dimers vs. one-dimensional chains in 2-thiooxoimidazolidin- 4-one (thiohydantoin) drug derivatives

Article abstract of DOI:10.1039/b924683e

Hydantoins have been known as medicinally active compounds since the 1940s and thiohydantoin derivatives are currently undergoing clinical trials as potent androgen receptor antagonist drugs. Control of solid state properties including the formation of drug polymorphs is important to the pharmaceutical industry, and frequently results from different H-bonded motifs. N-H thiohydantoins show formation of H-bonded arrays in the solid state. Two novel and five known thiohydantoin derivatives were synthesised via reaction of alkyl isothiocyanates with amino acid methyl esters. X-Ray crystallographic data were obtained for all seven compounds showing that four of the structures contain hydrogen bonded dimeric units linked via N-H...S interactions and three of the structures have N-H...O linked H-bonded chains. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/b924683e

View Article Online / Journal Homepage / Table of Contents for this issue
CrystEngComm
www.rsc.org/crystengcomm
Volume 12 | Number 6 | June 2010 | Pages 1659–1962
HIGHLIGHT
PAPER
Imai and Oaki
Archibald et al.
PAPER
Emergence of helical morphologies
with crystals: twisted growth under
diffusion-limited conditions and
chirality control
Hydrogen bonded dimers vs.
one-dimensional chains in
2-thiooxoimidazolidin-4-one
(thiohydantoin) drug derivatives
Raston et al.
Lanthanoid assisted self assembly
of imidazolium cations in organic
bi-layers

PAPER
www.rsc.org/crystengcomm | CrystEngComm
View Article Online
Hydrogen bonded dimers vs. one-dimensional chains in 2-thiooxoimidazolidin-
4-one (thiohydantoin) drug derivatives†
*
Sushil Jha, Jon D. Silversides, Ross W. Boyle and Stephen J. Archibald
*
Received 25th November 2009, Accepted 9th February 2010
First published as an Advance Article on the web 5th March 2010
DOI: 10.1039/b924683e
Hydantoins have been known as medicinally active compounds since the 1940s and thiohydantoin
derivatives are currently undergoing clinical trials as potent androgen receptor antagonist drugs.
Control of solid state properties including the formation of drug polymorphs is important to the
pharmaceutical industry, and frequently results from different H-bonded motifs. N–H thiohydantoins
show formation of H-bonded arrays in the solid state. Two novel and five known thiohydantoin
derivatives were synthesised via reaction of alkyl isothiocyanates with amino acid methyl esters. X-Ray
crystallographic data were obtained for all seven compounds showing that four of the structures
contain hydrogen bonded dimeric units linked via N–H/S interactions and three of the structures have
N–H/O linked H-bonded chains.
desirable or problematic.19–21 In addition there is considerable
Introduction
general interest in the controlled formation of hydrogen bonded
arrays in the solid state, particularly in relation to the competi-
tion of sulfur and oxygen as H-bond acceptors.22
The hydantoin or imidazolidine-2,4-dione heterocycle is present in
a wide range of biologically active compounds including thera-
peutic drugs for the treatment of seizures and antitumour
compounds.1–4 Thiohydantoins, where one of the oxygens is
replacedbya sulfur, havealsobeenused asanti-convulsant agents5
and are present in fungicides, herbicides and natural products.6,7
Both of these heterocycles are commonly used as templates in
combinatorial chemistry libraries.8,9 However, the overriding
current medicinal interest comes from the recent application of
thiohydantoins as androgen receptor antagonistswhich are nowin
clinical trials for the treatment of hormone independent prostate
cancers.10–13 This worklinks with ourongoingeffortsto investigate
new antagonists of cellular receptors.14–17
The thiohydantoin backbone can easily be modified in an
attempt to probe the factors that induce the preference for one
structural type over another. Groups can be added to provide
steric bulk or offer the potential for further intermolecular
interactions such as p/p stacking. We have incorporated
benzyl, phenyl, propyl and allyl substituents onto the heterocycle
backbone in the R1 and R2 positions. We then crystallised the
compounds to determine whether any preference is shown for the
weaker N–H/S bond over N–H/O bonds and how this relates
to structural type.
Previous studies of 5-substituted hydantoins and thiazolines
have used these units to assemble H-bonded arrays. For example,
monosubstituted hydantoins or thiazolines which can potentially
form triply hydrogen bonded dimers have been studied.23 Thio-
hydantoins have also been used in crystal engineering as spacers in
metallo-organic frameworks.24 In these polymeric metal
complexes the thiohydantoin amine group at the 3-position was
deprotonated and coordinated to caesium along with bridging O
and S donors. In our work we have included a substituent in the
3-position which blocks formation of N–H/O dimeric interac-
tions. On ruling out the most energetically favourable interaction
the remaining choice is formation of either an eight membered
ring dimer with N–H/S]C H-bonds or a 1D H-bonded chain
with N–H/O]C interactions. This should offer a finer energetic
balance between H-bonded structures and increased potential for
drug polymorph formation. In this study, our aim was to gain
insight into interactions that may control the topology.
We have investigated the potential for thiohydantoin
compounds to form H-bonded arrays in the solid state, which is
relevant to crystalline drug morphology.18,19 A fine energetic
balance between different H-bonded arrangements can result in
the isolation of different polymorphs, which can either be
Department of Chemistry, The University of Hull, Cottingham Road, Hull,
UK HU6 7RX. E-mail: s.j.archibald@hull.ac.uk; Fax: +44(0) 1482
466410; Tel: +44(0) 1482 465488
Results and discussion
† Electronic supplementary information (ESI) available: CIF and
platon.lst (CALC ALL) files are included for all X-ray structures.
Hirshfeld surfaces and fingerprint plots calculated for the X-ray
structures found in the CSD. CCDC reference numbers
756015–756021. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/b924683e
A series of compounds was synthesised which are substituted in
the 3- and 5-positions, see Fig. 1, with groups of different rigidity
and varying potential to form p interactions. The five membered
thiohydantoin ring can be formed by cyclisation with an
1730 | CrystEngComm, 2010, 12, 1730–1739
This journal is ª The Royal Society of Chemistry 2010

View Article Online
Scheme
1
Synthesis of
a
thiohydantoin derivative from an iso-
thiocyanate compound.
compounds including the Bucherer–Bergs route and a highly
efficient microwave assisted version of the Biltz synthesis.26–28
The thiohydantoin core can also easily be functionalised in the
5-position after cyclisation. However, our selected route was to
exploit the reaction of methyl ester derivatives of amino acids
with alkyl isothiocyanates to give the cyclic derivatives in high
yields. All compounds synthesised from chiral amino acids
produced racemic mixtures of products.
Methyl esters of amino acids were reacted with iso-
thiocyanates, see Scheme 1, to cyclise, forming the five membered
rings in an analogous fashion to the common reaction of isocy-
anates or phosgene to form hydantoin heterocycles. Of the seven
compounds produced (1–7), see Fig. 1, the N-benzyl derivatives 2
and 7 have not previously been synthesised. The remaining
compounds (1 and 3–6) have previously been made using
a variety of synthetic procedures.
The most common route to produce these compounds
involves a similar cyclisation using the amino acid rather than
the methyl ester. This method has been used to synthesise 1, 3, 4
and 5.29,30 Compound 4 has also been made by reaction of the
amino acid with a dithiocarbamate,31 and as a single diaste-
reoisomer from Fmoc protected phenylalanine using a micro-
wave procedure.32,33 Compound 6 has been synthesised by
a number of procedures34,35 including a stepwise process
involving isolation and cyclisation of a thiourea derivative.36
Yields for these processes are variable but, in comparison to the
direct reactions of the amino acids with isothiocyanates, our use
of commercially available methyl esters results in higher yields
Fig. 1 Thiohydantoin derivatives synthesised and structurally charac-
terised in this study.
isothiocyanate derivative. Of the seven compounds selected for
this study, two of the thioimidazolidones have not been reported
previously but none had been crystallographically characterised.
These compounds have the potential for intermolecular
H-bonding interactions via an N–H group with either or both of
the O and S atoms. 5-(2-Pyridylmethylene)-2-thiohydantoin has
previously been used in crystal engineering as a component
potentially offering an ADA hydrogen bonding arrangement.23
The structure showed formation of a 2D ‘tape’ network
composed of hydrogen bonded dimers. There were additional
interactions with intramolecular H-bonding to the pyridyl N and
a p/p interaction with the pyridyl ring.
1
(>90%) in all cases. The compounds were characterised by H
and 13C NMR and all crystals were grown by the same method.
Slow evaporation of a mixture of hexane and ethyl acetate gave
diffraction quality crystals for structure determination. Single
crystal X-ray diffraction studies unambiguously determined the
structures of these compounds.
Dimers with N–H/S H-bonds have also previously been
observed in the solid state for related compounds.25 Thermody-
namically the stronger N–H/O interaction should be preferred
over the N–H/S interaction. However, in the solid state struc-
ture other factors such as accommodation of bulky groups and p
interactions can affect this energetic balance. Combinations of
four substituents (allyl, benzyl, phenyl and isopropyl) were used
with the N–H retained at the 1-position.
Structural studies
The crystallographic details for the seven structures are given in
Tables 1 and 2, and selected bond lengths and angles are listed in
Table 3. Two structural types were identified: H-bonded dimers
and one-dimensional H-bonded chains. The dimeric structures
involve the neighbouring N–H and sulfur group acting as donor
and acceptor respectively to form intermolecular interactions. In
all of these molecules the related O dimer cannot form as the
neighbouring N1 position is substituted. Thus the alternative
possibility is the formation of a chain type structure where the
N–H/O interaction is present.
Synthesis
The imidazolidine-2,4-dione core and its mixed thio–oxo equiv-
alent are common five membered rings that can easily be syn-
thesised. There are a number of routes to produce such
In these molecules where the acceptor : donor ratio is 2 : 1
a direct choice is offered between a sulfur and an oxygen
This journal is ª The Royal Society of Chemistry 2010
CrystEngComm, 2010, 12, 1730–1739 | 1731

View Article Online
Table 1 Crystal data for the single crystal X-ray structures of 1 to 4
1
2
3
4
Formula
Mr
C6H6N2OS
154.19
C13H16N2OS
248.34
C9H8N2OS
192.23
C16H12N2OS
282.35
Crystal system
Space group
Monoclinic
P21/c
Monoclinic
P21/c
Monoclinic
P21/c
12.377(4)
5.6900(12)
12.515(4)
91.27(3)
881.1(5)
4
1.449
0.323
400
3.62 to 34.75
8162
3714 [0.0992]
97.9% (34.75)
1.176
0.1199, 0.3058
0.2063, 0.4126
0.043(6)
0.763 and ꢁ0.820
Monoclinic
C2/c
˚
a/A
9.3380(13)
6.3930(5)
13.1620(18)
108.753(11)
774.03(16)
4
1.376
0.364
150
3.27 to 34.74
7718
3169 [0.0606]
98.5% (34.74)
0.956
0.0382, 0.938
0.0666, 0.1024
0.002(3)
5.0267(9)
11.2601(12)
22.199(4)
94.545(14)
1252.5(3)
4
1.317
0.244
150
2.58 to 34.76
27 326
5327 [0.1370]
98.8% (34.76)
0.853
0.0493, 0.1021
0.1816, 0.1410
0.024(3)
28.380(6)
5.0073(6)
20.811(4)
106.952(16)
2828.9(9)
8
1.326
0.225
150
2.87 to 30.00
11 195
4096 [0.0839]
99.6% (30.0)
1.034
0.1072, 0.3080
0.1901, 0.3591
0.006(2)
˚
b/A
˚
c/A
b/ꢀ
Volume/A
3
˚
Z
Density (calc.)/Mg mꢁ3
mMoKa/mmꢁ1
T/K
q range/ꢀ
Measured reflections
Unique reflections [Rint
Completeness (q)/ꢀ
Goodness-of-fit on F2
R1, wR2 (I > 2s(I))
R1, wR2 (all data)
Extinction coefficient
]
ꢁ3
˚
Largest diff. peak and hole/e A
0.410 and ꢁ0.396
0.495 and ꢁ0.833
1.370 and ꢁ0.644
acceptor. The steric bulk appears to prohibit formation of
multiple interactions.
intermolecular close approaches from C–H groups to the
carbonyl oxygen with the most significant from theꢀH-atom on
˚
one of the allyl carbons (C5–H5/O1 3.315(2) A, 145 ). This may
have some impact on the orientation of the dimeric units in the
structure.
H-bonded dimers
The remaining three structures all contain phenyl or benzyl
substituents, offering the potential for p/p type interactions.
However, the dimeric units are not aligned to optimise stacking
interactions between the benzene rings but CH/p interactions
are observed. As there may be a fine energetic balance differen-
tiating the type of H-bonded network, these weak interactions
could contribute to the structural type. Steric interactions in
packing the molecules together are also likely to be important.
The key factor remains the formation of strong interactions in
the N–H/S dimeric unit. Structure 3 is most similar to 1 in terms
Four of the structures determined, 1–4, contain the H-bonded
dimer motif held together by two N–H/S bonds with the donor
to acceptor distances in the range from 3.32 to 3.40 A and the D–
˚
H/A angle varying between 158.6 to 177.4ꢀ, see Fig. 2 and Table
4. Crystal packing diagrams are shown in Fig. 2, see 1(c), 2(c),
3(c) and 4(c). The structural parameters are further discussed
later in comparison with the related compounds deposited in the
Cambridge Structural Database.
Compound 1 is not substituted in the C5 (R1) position and has
an allyl substituent at the N1 (R2) position. There are two
Table 2 Crystal data for the single crystal X-ray structures of 5 to 7
5
6
7
Formula
Mr
C9H14N2OS
198.28
C18H10N2OS
206.26
C17H16N2OS
296.38
Crystal system
Space group
Monoclinic
P21/c
Monoclinic
C2/c
Monoclinic
P21/n
˚
a/A
9.5863(11)
9.2632(7)
11.9253(12)
95.960(9)
1053.24(18)
4
1.250
0.272
150
2.79 to 34.77
12 811
4526 [0.0356]
99.0% (34.77)
1.130
0.0501, 01372
0.0675, 0.1486
0.043(7)
21.9470(19)
8.2582(8)
11.4242(9)
115.115(6)
1936.1(3)
4
1.415
0.300
150
2.60 to 34.78
10 104
4002 [0.0406]
95.6% (34.78)
1.377
0.0411, 0.1249
0.0493, 0.1298
0.0020(2)
0.452 and ꢁ0.387
9.0344(11)
17.331(13)
9.5981(11)
101.028(10)
1475.1(3)
4
1.335
0.219
150
2.58 to 34.78
20 656
6254 [0.1200]
97.9% (34.78)
1.042
0.0736, 0.1932
0.1188, 0.2374
0.097(11)
0.744 and ꢁ0.814
˚
b/A
˚
c/A
b/ꢀ
Volume/A
3
˚
Z
Density (calc.)/Mg mꢁ3
mMoKa/mmꢁ1
T/K
q range/ꢀ
Measured reflections
Unique reflections [Rint
Completeness (q)/ꢀ
Goodness-of-fit on F2
R1, wR2 (I > 2s(I))
R1, wR2 (all data)
Extinction coefficient
]
ꢁ3
˚
Largest diff. peak and hole/e A
0.585 and ꢁ0.448
1732 | CrystEngComm, 2010, 12, 1730–1739
This journal is ª The Royal Society of Chemistry 2010

View Article Online
˚
Table 3 Selected bond lengths (A) for structures 1 to 7
1
2
3
4
5
6
7
C(1)–S(1)
C(2)–O(1)
C(1)–N(2)
C(1)–N(1)
C(2)–N(1)
C(2)–C(3)
C(3)–N(2)
1.6700(11)
1.2079(15)
1.3319(15)
1.3827(16)
1.3875(14)
1.5082(18)
1.4535(16)
1.665(2)
1.204(3)
1.330(3)
1.380(3)
1.383(3)
1.513(3)
1.453(3)
1.666(5)
1.214(5)
1.331(5)
1.387(5)
1.391(5)
1.497(6)
1.447(6)
1.630(4)
1.237(6)
1.340(5)
1.383(5)
1.383(7)
1.511(7)
1.476(6)
1.6568(11)
1.2174(14)
1.3329(15)
1.3963(14)
1.3700(15)
1.5164(17)
1.4549(15)
1.6557(10)
1.2190(13)
1.3319(13)
1.3959(12)
1.3716(13)
1.5114(15)
1.4488(14)
1.652(2)
1.213(2)
1.335(2)
1.397(2)
1.372(2)
1.515(2)
1.450(2)
of substitution pattern with close approaches from two of the
phenyl protons to carbonyl oxygens in neighbouring molecules
appearing to influence the twist of the phenyl ring (C6–H6/O1
and C9–H9/O 3.246(5) A/166 and 3.388(5) A/166 respec-
tively). These interactions link dimeric units throughout the
structure. There is also a weak intermolecular CH/p interaction
from H3A on C3.
the chain. There are two intermolecular CH/p interactions
from a C–H on each of the phenyl rings.
ꢀ
ꢀ
˚
˚
Hirshfeld surfaces
Another method for the analysis of intermolecular contacts that
offers a whole-of-the-molecule approach is the use of Hirshfeld
surfaces.37,38 This method was used to generate pictures of the
surfaces in which the significant intermolecular contacts are
highlighted, see Fig. 4A. The dominant interactions between N–
H and S to form the dimeric species can be seen in the Hirshfeld
surface plots as the brightest red areas for 1–4. Fingerprint plots
were produced to show the intermolecular surface contact
distances with the regions highlighted for the O/H and S/H
interactions, Fig. 4B and 4C respectively. The plots are reciprocal
and so both donor and acceptor interactions are included.
Similar patterns are observed for all four compounds with the
importance of CH/O intermolecular contacts particularly
notable for compounds 1 and 3 matching the previous analysis.
This can be observed on both the Hirshfeld surfaces and the
fingerprint plots.
Compounds 2 and 4 have significantly greater steric bulk as
they are substituted in both the 3- and 5-positions. Compound 2
does not have any p/p interactions between the flexible benzyl
groups. There are only very weak intermolecular interactions in
addition to the dimeric H-bonding with a CH/O distance/angle
ꢀ
˚
of 3.366(3) A, 140 for the C12 methyl group (H12C).
Compound 4 again has no p/p interactions between benzene
rings despite the presence of both phenyl and benzyl substituents.
The most significant additional interactions are probably the
intermolecular ones between the benzene ring CH and neigh-
bouring oxygen or sulfur atoms.
One-dimensional H-bonded chains
The Hirshfeld surfaces of compounds 5–7 show the most
significant interactions (red) on opposite sides of the molecule as
would be expected for chain formation and the fingerprint plots
show shorter O/H interactions and longer S/H distances as
expected.
The other structural type observed in this work shows formation
of 1D chains in the solid state, see Fig. 3. In this case the
hydrogen bond is formed between the ring N–H and the carbonyl
oxygen of a neighbouring thiohydantoin molecule. Three struc-
tures of this type were characterised (compounds 5–7), see
Table 3 for bond lengths and angles, and Table 4 for H-bonding
parameters.
The relative contribution of different interactions to the
Hirshfeld surface was calculated. Due to varying number and
types of carbon containing substituents they are not directly
comparable across the series but offer some insight into the
effects of adding different substituents onto the thiohydantoin
backbone. No significant p/p interactions are observed with
C/C close contacts making up less than 1.1% of the surface area
for each of 1–7. C/H intermolecular contacts vary significantly
with the different substituents present. Where an aryl substituent
is present (in compounds 2, 3, 4, 6, and 7) the relative contri-
bution to the surface area for C/H intermolecular contacts is in
the range of 16.0 to 22.8%.
The 3-allyl-5-propyl thiohydantoin 5 forms one-dimensional
hydrogen bonded chains via a single H-bond between the
carbonyl oxygen and the amine ring N1 position. The steric bulk
of the propyl group appears to inhibit formation of the eight
membered ring NH/S H-bonded dimer. There do not appear to
be any other significant inter- or intramolecular interactions.
The inclusion of a benzyl substituent in the 3-position of 6 may
disfavour formation of an optimised dimeric species to preferen-
tially give one-dimensional chains with N–H/O H-bonds.
Crystal packing may also affect the structural preference in this
case. The benzyl group is twisted to be almost perpendicular to the
thiohydantoin ring plane (N1–C4–C5–C10 torsion angle ¼
106.52(11)ꢀ). The packing places the benzylic groups together in
a layer on either side of the H-bonded chains. No p/p interac-
tions or other significant intermolecular interactions are observed
but the hydrophobic benzyl groups are grouped together.
In 7, both phenyl rings are twisted to accommodate the chain
structure formed by the hydrogen bonding. However, in contrast
to the structures observed for 5 and 6, the orientation of the
hydantoin ring in the hydrogen bonded chain alternates along
Comparison with known thiohydantoin X-ray structures
A search of the Cambridge Structural Database39 (CSD) revealed
that ca. 60 structures had been deposited which contain the thi-
ohydantoin heterocyclic backbone, however, many are not rele-
vant for comparison as the heterocycle is part of a larger
polycyclic structure or has a substitution pattern which does not
allow for H-bonding. Eight structures (with coordinates avail-
able) were identified with both a thiohydantoin core and an N–H
group available for hydrogen bonding at the 1-position. They
This journal is ª The Royal Society of Chemistry 2010
CrystEngComm, 2010, 12, 1730–1739 | 1733

View Article Online
Fig. 2 The X-ray crystal structures of compounds 1 to 4. (a) A labelled ORTEP plot (ellipsoids at 50% probability level) of the molecule in the
asymmetric unit. (b) The H-bonded dimer with H/S interactions represented by a dashed line. (c) Packing of the dimers is shown, aligned to most
clearly indicate how they pack. The unit cell outline is shown in green.
can be divided into four categories: the unsubstituted thio-
hydantoin, a 5 monosubstituted derivative, 3,5-substituted
derivatives and 5-disubstituted derivatives, see Fig. 5. Only one
of these compounds has the same substitution pattern that we
have used in this study (EHXBY).
was also available for H-bonding or other species were present
within the crystal lattice, such as iodine.
The unsubstituted thiohydantoin (THHYDT) has two
H-bonds formed with the O atom and the S does not participate
in H-bonding.40 This is expected for a system where there is no
steric bulk or other factors directing the interactions away from
the most favourable H-bond donor–acceptor pair. An O dimer
forms with H-bonding from the N–H in the 3-position (between
The H-bonding patterns were analysed showing the O chains
and S dimers that have been observed in our work but also some
different H-bonding patterns, particularly where the N3 position
1734 | CrystEngComm, 2010, 12, 1730–1739
This journal is ª The Royal Society of Chemistry 2010

View Article Online
Table 4 H-bonding parameters for structures of 1 to 7 and relevant literature structures
Angle (D–H/A)/ꢀ
˚
D–H/A
˚
˚
D–H/A
Type of H-bonding
H/A/A
D/A/A
1 (N–H/S)
2 (N–H/S)
3 (N–H/S)
4 (N–H/S)
5 (N–H/O)
6 (N–H/O)
7 (N–H/O)
THHYDT
THHYDT
YINGIM
YINGIM
NUTVIH
NUTVIH
KUWDOV
EHXBY
S dimer
S dimer
S dimer
S dimer
O chain
O chain
O chain
O dimer
Other N–H/O
O dimer
S dimer
S dimer
O chain
O dimer
S dimer
O chain
O chain
S dimer
0.88
0.88
0.88
0.910(6)
0.830(2)
0.88
0.88
0.829
0.912
0.848/0.834
0.842/0.792
0.841
0.869
0.809
1.09
0.896
0.824
0.806
2.568
2.52
2.44
2.410(6)
2.042(2)
2.041
1.945
2.097
2.022
2.027/2.012
2.519/2.601
2.556
2.248
2.128
2.355
1.946
2.153
2.569
3.403(1)
3.393(2)
3.308(4)
3.316(4)
2.869(2)
2.914(1)
2.801(2)
2.916
158.60(12)
170.3(2)
168.1(4)
177.4(5)
174.01(18)
171.67(11)
163.9(2)
169.3
ꢁx, 1 ꢁ y, ꢁz
1 ꢁ x, 1 ꢁ y, 1 ꢁ z
ꢁx, 1 ꢁ y, ꢁz
½ ꢁ x, 5/2 ꢁ y, ꢁz
x, ½ ꢁ y, ꢁ½ ꢁ z
x, 1 ꢁ y, ꢁ½ + z
ꢁ½ + x, ½ ꢁ y, ꢁ½ + z
2.928
172.5
2.856/2.836
3.358/3.386
3.394
3.094
2.874
3.396
2.794
2.967
3.373
165.8/169.8
174.2/171.0
174.3
164.6
153.5
159.2
157.2
170.1
175.0
HAFLEF
WUSJID
LEVWAL
Fig. 3 The X-ray crystal structures of compounds 5 to 7. (a) A labelled ORTEP plot (ellipsoids at 50% probability level) of the molecule in the
asymmetric unit. (b) The H-bonded 1D chains with the H/O interactions represented by a dashed line. (c) Packing of the molecules is shown, aligned to
give the clearest view. The unit cell outline is shown in green.
This journal is ª The Royal Society of Chemistry 2010
CrystEngComm, 2010, 12, 1730–1739 | 1735

View Article Online
Fig. 4 Hirshfeld surface analysis of compounds 1–7. (A) Hirshfeld surface of compounds 1–7 mapped with dnorm which highlights both donor and
acceptor equally. (B) Fingerprint plots for compounds 1–7 resolved into O/H contacts. The full fingerprint appears beneath each decomposed plot in
grey. (C) Fingerprint plots for compounds 1–7 resolved into S/H contacts. The full fingerprint appears beneath each decomposed plot in grey.
1736 | CrystEngComm, 2010, 12, 1730–1739
This journal is ª The Royal Society of Chemistry 2010

View Article Online
the C]S and C]O). The N–H in the 1-position also has an
intermolecular H-bond to O in a neighbouring molecule.
5-Phenylthiohydantoin (YINGIM) has the same potential for
hydrogen bond formation as thiohydantoin (THHYDT) but with
additional steric bulk in the form of a single substituent in the
5-position. As with THHYDT the structure shows the formation
of O dimers with the N–H in the 3-position. But in this case
H-bonded S dimers are also present to give an overall three
dimensional structure.41 There are no p/p interactions and
although there are some reasonable distances for C–H/p inter-
actions, all the angles to the phenyl ring centroid are below 130ꢀ.
The next step in increasing steric bulk is to add a further group at
the5-positiontogive the 5-disubstitutedderivatives(NUTVIHand
KUWDOV). 5-Diphenylthiohydantoin (NUTVIH) shows
a combination of O chains and S dimers with no significant further
interactions (no p/p or strong C–H to p interactions are
observed).42 The 5-dimethylhydantoin (KUWDOV) unfortunately
does not offer an appropriate comparison as it is co-crystallised
with iodine (I2).43 Unsurprisingly the iodine in the crystal lattice
interacts with the sulfur blocking any possibility for its involvement
in H-bonding interactions. This gives discrete O-bonded dimers, as
would be predicted for this co-crystallised mixture.
The 3,5-disubstituted derivative which matches the substitu-
tion pattern in our work is 3-ethyl 5-(4-methoxybenzyl) thio-
hydantoin (EHXBY).44 The two 5-benzyl structures we
synthesised adopt different H-bonding motifs and so do not give
predictive structural information for such derivatives. This
compound forms sulfur H-bonded dimers, as was observed for
compound 4. No significant C–H to p interactions are present
but there are two intermolecular C–H/O hydrogen bonds aꢀt
˚
reasonable donor–acceptor distances and angles (3.53 A/159
ꢀ
˚
and 3.32 A/148 ).
Bulkier derivatives are also known with a 3,5-trisubstituted
pattern where the 5-position is disubstituted. This increase in
steric bulk appears to disfavour the formation of sulfur H-
bonded dimers to give structures with H-bonded chains formed
by interaction with the carbonyl oxygen. The 3-methyl
5-dimethyl thiohydantoin (HAFLEF)45 and the 3-ethyl 5-
diphenyl thiohydantoin (WUSJID)28 both give the chain type
structures. There are no significant additional interactions in the
HAFLEF structure. In the WUSJID structure the molecules are
arranged to maintain the H-bonding and to accommodate the
bulky phenyl substituents. There are two C–H to p interactions
Fig. 5 Thiohydantoin derivatives deposited in the crystallographic
database.
for a structure with multiple H-bonded networks (NUTVIH).
The lowest D–H/A angle is 154ꢀ, but this is for the KUWDOV
structure where iodine is incorporated in the crystal structure and
S/I interactions will be optimised. The remaining angles are in
the range 157 to 177ꢀ.
ꢀ
˚
at ca. 3.7 A with angles over 140 but no further significant
intermolecular H-bonding interactions.
The Hirshfeld surfaces were calculated for the literature
structures along with fingerprint plots in which either O/H or
S/H interactions highlighted (Fig. S1–S3)†. The thiohydantoin
structure THHYDT has over 50% of the relative contribution to
the Hirshfeld surface area from either O/H or S/H close
contacts (25.2% and 29.2% respectively). This contrasts with the
hexyldiphenyl substituted LEVWAL where 60.7% of the surface
area is contributed by H/H contacts, reflecting both the larger
number of alkyl and aryl groups and the reduced potential for
H-bonding interactions with O and S.
The final structure identified is also a 3,5-trisubstituted thio-
hydantoin but in this case with a long alkyl chain in the 3-posi-
tion. For this 3-hexyl 5-diphenyl thiohydantoin (LEVWAL),
H-bonded dimers with the sulfur acceptor are observed.46 This
changeover appears to be due to the packing of the long alkyl
chains which would interfere with O-chain formation as the
layers of S-dimers are formed with the oxygens pointing between
the layers. No other significant interactions were observed.
Examination of the hydrogen bonding parameters for the
seven structures from this work and the additional eight known
structures shows reasonable consistency in the N–H/O and N–
H/S distances, see Table 4. The donor to acceptor distances are
Conclusions
˚
˚
in the range 3.3 to 3.4 A for the sulfur acceptor and 2.8 to 3.1 A
for the oxygen acceptor. The longest N to O distance is observed
In our compounds (1–7), we have blocked formation of oxygen
acceptor H-bonded O dimers in thiohydantoins as anticipated.
This journal is ª The Royal Society of Chemistry 2010
CrystEngComm, 2010, 12, 1730–1739 | 1737

View Article Online
This has given a series of compounds which adopt one of two
structural types, either H-bonded dimers or 1D chains.
7.53 (br s, 1H, HNH). 13C NMR (100 MHz, CD2Cl2) d 43.45,
48.93, 118.13, 131.31, 171.65, 185.09.
No clear relationship is obvious between the substituent on the
N-position and the structural type. Both of the N-phenyl
compounds (3 and 4) form dimers but the similarly bulky
N-benzyl compounds (2, 6 and 7) adopt either chains or dimers
with no obvious reasons for structure selection. It may be that the
more rigid phenyl substituents encourage formation of the
rotationally inflexible dimers.
Compound 2 (R1 ¼ i-Pr, R2 ¼ Bn). 1H NMR (400 MHz,
acetone-d6) d 4.36 (s, 2H, HCH2), 7.35–7.37 (m, 2H, HAr),
7.43–7.53 (m, 3H, HAr), 9.11 (br s, 1H, HNH). 13C NMR (100
MHz, acetone-d6) d 49.65, 129.23, 129.38, 129.63, 134.87, 172.38,
185.53.
The 3 and 5 substituents can be accommodated by the
formation of infinite H-bonded chains with this motif allowing
for variation in the relative twist of neighbouring rings. Steric
bulk clearly plays a role. A comparison of 1 and 5 suggests
increased steric bulk in the 5-position favours chain formation. It
is also interesting to see the variation for the LEVWAL structure
which uses hydrophobic interactions to bring N-hexyl substitu-
ents together and forms a sulfur H-bonded dimer where oxygen
H-bonded chains might have been predicted.
Compound 3 (R1 ¼ H, R2 ¼ Ph). 1H NMR (400 MHz, CDCl3)
d 0.77 (d, 3H, J ¼ 7.00 Hz, HCH3), 0.96 (d, 3H, J ¼ 7.00 Hz,
H
CH3), 2.14–2.22 (m, 1H, HCH), 3.91 (dd, 1H, J ¼ 3.64 Hz and
1.12 Hz, HCH), 4.91 (d, 2H, J ¼ 2.52 Hz, HCH2), 7.16–7.24 (m,
3H, HAr), 7.35–7.38 (m, 2H, HAr), 8.16 (br s, 1H, HNH). 13C
NMR (100 MHz, CDCl3) d 16.17, 18.77, 44.39, 64.55, 127.82,
128.41, 128.61, 135.61, 173.52, 184.09.
Compound 4 (R1 ¼ Bn, R2 ¼ Ph). 1H NMR (400 MHz, CD2Cl2)
d 3.04 (dd, 1H, J ¼ 14.08 Hz and 6.44 Hz, HCH2), 3.19 (dd, 1H,
J ¼ 14.08 Hz and 4.24 Hz, HCH2), 4.39–4.42 (m, 1H, HCH), 6.89–
6.93 (m, 2H, HAr), 7.15–7.17 (m, 2H, HAr), 7.22–7.27 (m, 3H,
p/p interactions are not a key driving force in formation of
these structures. However, the weaker C–H to p type interactions
could tip the balance in favour of which structural type is
adopted. Computational energy calculations could be of value in
predicting structures for this class of compounds.
H
Ar), 7.32–7.38 (m, 3H, HAr), 8.26 (br s, 1H, HNH). 13C NMR
(100 MHz, CD2Cl2) d 37.27, 60.76, 127.68, 128.07, 128.76,
We plan to study more variants of these compounds to
determine if further rationalisation is possible. It does appear
that there is a fine energetic balance between the two structural
types. Conceivably both structural types could be observed for
thiohydantoin derivatives with certain combinations of substit-
uents to give crystalline polymorphs. The existence of poly-
morphs is of significant interest due to the potential
pharmaceutical applications of these compounds.
129.04, 129.19, 129.50, 132.33, 133.81, 172.71, 183.51.
1
Compound 5 (R1 ¼ i-Pr, R2 ¼ allyl). H NMR (400 MHz,
CDCl3) d 0.86 (d, 3H, J ¼ 6.76 Hz, HCH3), 1.03 (d, 3H, J ¼ 6.76
Hz, HCH3), 2.24 (m, 1H, HCH), 3.98 (dd, 1H, J ¼ 3.92 Hz and
1.40 Hz, HCH), 4.34 (td, 2H, J ¼ 5.60 Hz and 1.40 Hz, HCH2),
5.12–5.21 (m, 2H, H]CH2), 5.72–5.81 (m, 1H, H]CH), 8.56 (br s,
1H, HNH). 13C NMR (100 MHz, CDCl3) d 16.15, 18.80, 30.72,
42.90, 64.64, 118.39, 130.52, 173.31, 183.81.
Experimental
Compound 6 (R1 ¼ H, R2 ¼ Bn). 1H NMR (400 MHz, acetone-
d6) d 3.73 (s, 2H, HCH2), 4.57 (s, 2H, HCH2), 6.82–6.91 (m, 3H,
All reagents and solvents were purchased from Sigma-Aldrich,
Lancaster or Fisher and were used as supplied unless otherwise
stated.
H
Ar), 7.03–7.06 (m, 2H, HAr), 8.40 (br s, 1H, HNH). 13C NMR
(100 MHz, acetone-d6) d 43.37, 47.84, 126.87, 127.59, 127.84,
1
All H NMR and proton decoupled 13C NMR spectra were
135.48, 171.26, 183.86.
collected on a Jeol JNM-LA400 spectrometer at 400 and 100 MHz
respectively, and referenced against residual solvent signals.
Compound 7 (R1 ¼ Bn, R2 ¼ Bn). 1H NMR (400 MHz, CDCl3)
d 2.83 (dd, 1H, J ¼ 14.04 Hz and 8.16 Hz, HCH ), 3.17 (dd, 1H,
2
J ¼ 14.04 Hz and 4.24 Hz, HCH ), 4.23 (dq, 1H, J ¼ 4.24 Hz and
2
Synthetic procedure to prepare substituted 2-thioxoimidazolidin-
4-one derivatives
1.12 Hz, HCH ), 4.82 (s, 2H, HCH ), 7.04–7.07 (m, 2H, HAr), 7.14–
1
2
7.19 (m, 8H, HAr), 7.75 (br s, 1H, HNH). 13C NMR (100 MHz,
CDCl3) d 37.14, 44.37, 60.38, 127.53, 127.68, 128.34, 128.41,
128.89, 129.14, 134.28, 135.31, 173.12, 183.48.
To a solution of hydrochloride salts of the amino acid methyl
ester and alkyl/aryl isothiocyanate (1 equiv.) in CH2Cl2 (20 mL)
at room temperature was added triethylamine (2 equiv.) and the
reaction mixture was stirred for an hour. The solution was given
an aqueous wash (1 ꢂ 10 mL) and the organic phase dried over
Na2SO4. The solvent was removed under reduced pressure and
the crude product purified by short filtration column (flash silica)
using dichloromethane–ethyl acetate (9 : 1) as an eluent to give
the desired product as a white solid in more than 90% yield.
X-Ray quality crystals were obtained by slow evaporation
using hexane–ethyl acetate (9 : 1) as a solvent for crystallisation.
X-Ray crystallography†
€
Diffraction datasets were collected on a Stoe IPDS-II imaging
˚
plate diffractometer using MoKa radiation (l ¼ 0.71073 A). The
temperature of each crystal was kept at 150 K during data
collection and controlled using the Oxford Cryosystems Cryo-
stream Cooler.47 All structures were solved using direct methods
(SHELXS) and refined against F2 (SHELXL).48,49 H atoms were
either located on the difference map or placed in idealised
Compound 1 (R1 ¼ H, R2 ¼ allyl). 1H NMR (400 MHz,
positions and refined using a riding model with C–H ¼ 0.97 A,
˚
˚
CD2Cl2) d 4.01 (s, 2H, HCH ), 4.33 (td, 2H, J ¼ 5.64 Hz and 1.40
N–H ¼ 0.91 A and Uiso(H) ¼ 1.2 (or 1.5 for some H atoms) times
2
Hz, HCH ), 5.10–5.17 (m, 2H, H]CH ), 5.73–5.82 (m, 1H, H]CH),
Ueq of the carrier atom. The WinGX package was used for
2
2
1738 | CrystEngComm, 2010, 12, 1730–1739
This journal is ª The Royal Society of Chemistry 2010

View Article Online
20 S. Roy and A. J. Matzger, Angew. Chem., Int. Ed., 2009, 48, 8505–
8508.
refinement and production of data tables, and ORTEP-3 used for
structure visualisation.50,51 All ORTEP representations show
ellipsoids at the 50% probability level. Analysis of the H-bonded
and p-interactions was carried out using PLATON for both our
structures and the literature compounds.52 CrystalExplorer was
used to produce Hirshfeld surface plots based on dnorm and
decomposed 2D fingerprint plots to show the relative contribu-
tion for specified contacts to the intermolecular contact surface.53
21 Z. B. Ma and B. Moulton, J. Chem. Crystallogr., 2009, 39, 913–918.
22 R. Custelcean, Chem. Commun., 2008, 295–307.
23 M. M. Chowdhry, D. M. P. Mingos, A. J. P. White and
D. J. Williams, J. Chem. Soc., Perkin Trans. 1, 2000, 3495–3504.
24 M. Arca, F. Demartin, F. A. Devillanova, A. Garau, F. Isaia,
V. Lippolis and G. Verani, Inorg. Chem., 1998, 37, 4164–4165.
25 M. K. Krepps, S. Parkin and D. A. Atwood, Cryst. Growth Des.,
2001, 1, 291–297.
26 E. Ware, Chem. Rev., 1950, 46, 403.
27 R. G. Murray, D. M. Whitehead, F. Le Strat and S. J. Conway, Org.
Biomol. Chem., 2008, 6, 988–991.
Acknowledgements
28 G. G. Muccioli, J. H. Poupaert, J. Wouters, B. Norberg, W. Poppitz,
G. K. E. Scriba and D. M. Lambert, Tetrahedron, 2003, 59, 1301–
1307.
We would like to thank the Leverhulme Trust (F/00181H) and
the Wellcome Trust (069719) for supporting this work.
29 A. I. Khodair, H. I. El-Subbagh and A. M. Al-Obaid, Phosphorus,
Sulfur Silicon Relat. Elem., 1998, 140, 159–181.
30 A. Takahashi, H. Matsuoka, Y. Ozawa and Y. Uda, J. Agric. Food
Chem., 1998, 46, 5037–5042.
31 G. Blotny, Synthesis (Stuttgart), 1983, 391–392.
32 M. J. Lin and C. M. Sun, Tetrahedron Lett., 2003, 44, 8739–8742.
33 G. S. M. Sundaram, C. Venkatesh, H. Ila and H. Junjappa, Synlett,
2007, 251–254.
34 A. A. Atalla, E. A. Bakhite, A. M. Hussein and A. M. K. ElDean,
Phosphorus, Sulfur Silicon Relat. Elem., 1996, 112, 1–6.
35 A. I. Khodair, Nucleosides, Nucleotides Nucleic Acids, 2001, 20, 1735–
1750.
36 D. M. Wolfe and P. R. Schreiner, Synthesis (Stuttgart), 2007, 2002–
2008.
37 J. J. McKinnon, D. Jayatilaka and M. A. Spackman, Chem.
Commun., 2007, 3814–3816.
38 J. J. McKinnon, A. S. Mitchell and M. A. Spackman, Chem.–Eur. J.,
1998, 4, 2136–2141.
39 D. A. Fletcher, R. F. McMeeking and D. Parkin, J. Chem. Inf.
Comput. Sci., 1996, 36, 746–749.
40 L. Walker, K. Folting and L. Merritt, Jr, Acta Crystallogr., Sect. B:
Struct. Crystallogr. Cryst. Chem., 1969, 25, 88.
41 T. Ogawa, S. Kitoh, M. Ichitani, A. Kuwae, K. Hanai and
K. Kunimoto, Anal. Sci.: X-Ray Struct. Anal. Online, 2007, 23, x199.
42 A. Roszak and D. Weaver, Acta Crystallogr., Sect. C: Cryst. Struct.
Commun., 1998, 54, 1168.
43 F. Cristiani, F. Demartin, F. Devillanova, F. Isaia, G. Saba and
G. Verani, J. Chem. Soc., Dalton Trans., 1992, 3553.
44 H. Fujiwara and J. M. van der Veen, J. Chem. Soc., Perkin Trans. 2,
1979, 659.
45 F. Demartin, F. Devillanova, A. Diaz, F. Isaia, V. Lippolis and
G. Verani, Phosphorus, Sulfur Silicon Relat. Elem., 1992, 70, 121.
46 C. Michaux, G. G. Muccioli, D. M. Lambert and J. Wouters, Bioorg.
Med. Chem. Lett., 2006, 16, 4772–4776.
References
1 M. Mudit, M. Khanfar, A. Muralidharan, S. Thomas, G. V. Shah,
R. W. M. van Soest and K. A. El Sayed, Bioorg. Med. Chem., 2009,
17, 1731–1738.
2 A. Cavazzoni, R. R. Alfieri, C. Carmi, V. Zuliani, M. Galetti,
C. Fumarola, R. Frazzi, M. Bonelli, F. Bordi, A. Lodola, M. Mor
and P. G. Petronini, Mol. Cancer Ther., 2008, 7, 361–370.
3 C. Carmi, A. Cavazzoni, V. Zuliani, A. Lodola, F. Bordi, P. V. Plazzi,
R. R. Alfieri, P. G. Petronini and M. Mor, Bioorg. Med. Chem. Lett.,
2006, 16, 4021–4025.
4 Q. F. Zhu, Y. H. Pan, Z. X. Xu, R. M. Li, G. G. Qiu, W. J. Xu,
X. B. Ke, L. M. Wu and X. M. Hu, Eur. J. Med. Chem., 2009, 44,
296–302.
5 P. Dang and A. K. Madan, J. Chem. Inf. Comput. Sci., 1994, 34, 1162–
1166.
6 T. Lindel and H. Hoffmann, Tetrahedron Lett., 1997, 38, 8935–8938.
7 L. D. S. Yadav and S. Shukla, J. Agric. Food Chem., 1995, 43, 2526–
2529.
8 E. Colacino, F. Lamaty, J. Martinez and I. Parrot, Tetrahedron Lett.,
2007, 48, 5317–5320.
9 B. A. Dressman, L. A. Spangle and S. W. Kaldor, Tetrahedron Lett.,
1996, 37, 937–940.
10 C. Tran, S. Ouk, N. J. Clegg, Y. Chen, P. A. Watson, V. Arora,
J. Wongvipat, P. M. Smith-Jones, D. Yoo, A. Kwon,
T. Wasielewska, D. Welsbie, C. D. Chen, C. S. Higano, T. M. Beer,
D. T. Hung, H. I. Scher, M. E. Jung and C. L. Sawyers, Science,
2009, 324, 787–790.
11 M. E. Taplin and S. P. Balk, J. Cell. Biochem., 2004, 91, 483–490.
12 C. Harrison, Nat. Rev. Drug Discovery, 2009, 8, 452–452.
13 H. C. Shen and S. P. Balk, Cancer Cell, 2009, 15, 461–463.
14 A. Khan, G. Nicholson, J. Greenman, L. Madden, G. McRobbie,
C. Pannecouque, E. De Clercq, R. Ullom, D. L. Maples,
R. D. Maples, J. D. Silversides, T. J. Hubin and S. J. Archibald,
J. Am. Chem. Soc., 2009, 131, 3416–3417.
15 G. C. Valks, G. McRobbie, E. A. Lewis, T. J. Hubin, T. M. Hunter,
P. J. Sadler, C. Pannecouque, E. De Clercq and S. J. Archibald,
J. Med. Chem., 2006, 49, 6162–6165.
16 A. Khan, J. D. Silversides, L. Madden, J. Greenman and
S. J. Archibald, Chem. Commun., 2007, 416–418.
47 J. Cosier and A. M. Glazer, J. Appl. Crystallogr., 1986, 19, 105–107.
48 G. M. Sheldrick, SHELXS-97 Program for Crystal Structure Solution,
€
€
€
Institut fur Anorganische Chemie der Universitat, Tammanstrasse 4,
D-3400 Gottingen, Germany, 1997.
49 G. M. Sheldrick, SHELXL-97 Program for Crystal Structure
€
€
€
Refinement, Institut fur Anorganische Chemie der Universitat,
Tammanstrasse 4, D-3400 Gottingen, Germany, 1997.
50 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
51 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
52 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7.
17 G. McRobbie, G. C. Valks, C. J. Empson, A. Khan, J. D. Silversides,
C. Pannecouque, E. De Clercq, S. G. Fiddy, A. J. Bridgeman,
N. A. Young and S. J. Archibald, Dalton Trans., 2007, 5008–5018.
18 J. Bernstein, Polymorphism in Molecular Crystals, Oxford University
Press, Oxford, 2002.
53 S. K. Wolff, D. J. Grimwood, J. J. McKinnon, D. Jayatilaka and
M. A. Spackman, CrystalExplorer v2.1, University of Western
Australia, Perth, Australia, 2007.
19 J. Lu and S. Rohani, Curr. Med. Chem., 2009, 16, 884–905.
This journal is ª The Royal Society of Chemistry 2010
CrystEngComm, 2010, 12, 1730–1739 | 1739