Polymorphism of 5-(pyridin-2-ylmethylene)-3-phenyl-2-methylthio-3,5- dihydro-4H-imidazole-4-one

Article abstract of DOI:10.1039/c1ce05098b

The title compound, C₁₆H₁₃N₃OS (1), exists in three polymorphic forms. Crystalline 1 undergoes an enantiotropic, first-order, k2 phase transition at 262.9(5) K with ΔH = 0.3(1) kJ mol^-1. Upon cooling below the transition temperature, the high temperature orthorhombic polymorph (Form I, space group Pbcm) transforms into a low temperature orthorhombic polymorph (Form II, space group Pbca) with a unit cell twice the size of that of the Form I. A molten 1 can be cooled in a controlled fashion to generate a monoclinic Form III of 1 with the unit cell size similar to that of Form I. Metastable Form III, once isolated, is indefinitely stable between 100 K and its melting point of 466 K. If crystals of Form III are in contact with seed crystals of Form I, a monotropic t2 first-order Form III → Form I phase transition occurs upon heating with the onset between 420 and 448 K and ΔH = -1.7(4) kJ mol^-1. The most substantial differences among the molecular geometries of 1 in Forms I-III are observed in the position and tilt of the phenyl ring relative to the rest of the molecule. The packing in Form III is very different from those in the other polymorphs. DFT molecular geometry optimizations produce the following order of stable molecule configurations: Form II (most stable), Form I (0.50 kJ mol ^-1), Form III (2.81 kJ mol^-1).

Full text of DOI:10.1039/c1ce05098b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
CrystEngComm
Cite this: CrystEngComm, 2011, 13, 3444
www.rsc.org/crystengcomm
PAPER
Polymorphism of 5-(pyridin-2-ylmethylene)-3-phenyl-2-methylthio-3,5-
dihydro-4H-imidazole-4-one†
a
Ilia A. Guzei, Erica M. Gunn,b Lara C. Spencer,a Jennifer M. Schomakera and Jared W. Rigolia
*
Received 18th January 2011, Accepted 12th February 2011
DOI: 10.1039/c1ce05098b
The title compound, C16H13N3OS (1), exists in three polymorphic forms. Crystalline 1 undergoes an
enantiotropic, first-order, k2 phase transition at 262.9(5) K with DH ¼ 0.3(1) kJ molꢀ1. Upon cooling
below the transition temperature, the high temperature orthorhombic polymorph (Form I, space group
Pbcm) transforms into a low temperature orthorhombic polymorph (Form II, space group Pbca) with
a unit cell twice the size of that of the Form I. A molten 1 can be cooled in a controlled fashion to
generate a monoclinic Form III of 1 with the unit cell size similar to that of Form I. Metastable Form
III, once isolated, is indefinitely stable between 100 K and its melting point of 466 K. If crystals of Form
III are in contact with seed crystals of Form I, a monotropic t2 first-order Form III / Form I phase
transition occurs upon heating with the onset between 420 and 448 K and DH ¼ ꢀ1.7(4) kJ molꢀ1. The
most substantial differences among the molecular geometries of 1 in Forms I–III are observed in the
position and tilt of the phenyl ring relative to the rest of the molecule. The packing in Form III is very
different from those in the other polymorphs. DFT molecular geometry optimizations produce the
following order of stable molecule configurations: Form II (most stable), Form I (0.50 kJ molꢀ1), Form
III (2.81 kJ molꢀ1).
molecular composition was proven to be that of 1, however the
Introduction
previously reported crystal structure of 1 was based on room
The synthesis and study of well-defined catalysts for N2O acti-
temperature data and had different unit cell parameters. The
vation is an important step towards the development of new
room temperature structure was solved in the space group Pbcm
methods that can utilize this mild and selective oxidant for
with a unit cell half the size of that of the low temperature
organic synthesis. Of particular interest are transition metal
polymorph, indicating the possibility of a solid-state phase
complexes supported by ligands that can be readily manipulated
transition between 100 and 300 K.4 Our X-ray single crystal
sterically and electronically. Recently, a cobalt complex bearing
study revealed a phase transition at 263(1) K between the origi-
an easily prepared thiohydantoin ligand was reported to catalyze
nally reported Form I and the low temperature Form II.
the oxidation of triphenylphosphine and the epoxidation of
Evidence of the existence of another polymorph was found
norbornene using N2O at RT.1 Due to our interest in exploring
during the DSC second heating cycle when measuring the Form
thiohydantoins as tunable ligands for catalysts promoting olefin
I/Form II transformation enthalpy. Indeed, a monoclinic Form
epoxidation using nitrous oxide as the terminal oxidant, we
III of 1, metastable between 100 and 466 K, has been subse-
prepared 5-(pyridine-2-ylmethylene)-3-phenyl-2-methylthio-3,5-
quently isolated from the melt and structurally characterized.
dihydro-4H-imidazole-4-one (1, Fig. 1) for further study.
To expeditiously confirm the successful outcome of the
synthesis, a single crystal of 1 was mounted on a diffractometer
for a unit cell determination at 100 K. The orthorhombic unit cell
matched a known unit cell neither in the in-house Reciprocal Net
database2 nor in the Cambridge Structural Database (CSD).3
Therefore a complete dataset was acquired and the structure
solved and refined in the orthorhombic space group Pbca. The
aDepartment of Chemistry, University of Wisconsin-Madison, 1101
University Ave, Madison, WI, 53706, USA. E-mail: iguzei@chem.wisc.edu
bSchool of Pharmacy, University of Wisconsin-Madison, 777 Highland
Ave, Madison, WI, 53705, USA
Fig. 1 A molecular diagram of 1 at 100 K shown with 50% probability
ellipsoids and ring designations.8
† CCDC reference numbers 809204 and 809205. For crystallographic
data in CIF or other electronic format see DOI: 10.1039/c1ce05098b
3444 | CrystEngComm, 2011, 13, 3444–3450
This journal is ª The Royal Society of Chemistry 2011

View Article Online
(5Z)-5-(pyridine-2-ylmethylene)-2-thioxoimidazolin-4-one (100 mg,
0.36 mmol, 1.0 equiv) in 5 mL of a 3 : 1 mixture of CH3CN : DMF
was treated with MeI (1.5 equiv) and K2CO3 (50 mg, 0.36 mmol, 1.0
equiv) at room temperature, Scheme 1. The reaction mixture was
stirred vigorously for 5 min, heated briefly to boiling and filtered hot.
The filtrate was cooled slowly to RT and then placed in an ice bath
to complete the crystallization. The resulting solid was collected by
filtration to give the desired product in 38% yield. Alternatively, the
volatiles could be removed under reduced pressure and the residue
purified by flash column chromatography (100% CH2Cl2 followed
1
by 100% CH3CN). Mp 481 K. H NMR (DMSO-d6, 300 MHz)
d 8.9 (d, 1H, J ¼ 8.0 Hz), 8.7 (d, 1H, J ¼ 4.5 Hz), 8.0 (t of d, 1H,
J ¼ 7.8, 1.5 Hz), 7.6–7.3 (m, 6H), 6.9 (s, 1H), 2.7 (s, 3H).
Preparation of Form III of 5-(pyridin-2-ylmethylene)-3-phenyl-
2-methylthio-3,5-dihydro-4H-imidazole-4-one. Crystals of the
monoclinic polymorph (Form III) can be readily grown from an
undercooled melt, prepared by placing a few mg of powder of
Form I on a coverslip and heating it briefly to 491 K on
a microscope hot stage. Heating was limited to roughly
30 seconds to avoid decomposition. The sample was then rapidly
quenched to room temperature for a few seconds and returned to
the hotplate at the desired growth temperature. A thin film
sample grown between two glass coverslips produced spherulitic
crystals whose morphology depends on the growth temperature
(Fig. 2). Upon heating, the spherulites were observed to melt
between 466 and 467 K.
Scheme 1 Synthesis of 5-(pyridine-2-ylmethylene)-3-phenyl-2-methyl-
thio-3,5-dihydro-4H-imidazole-4-one.
Herein we report our syntheses of the polymorphs of 1, phase
transitions observed between them, and discuss their molecular
conformations and stability.
Experimental
Synthesis
Preparation of 3-phenyl-2-thiohydantoin5. Glycine (3.0 g,
40 mmol, 1.0 equiv) was dissolved in a 50% KOH solution (2.2 g,
40 mmol, 1.0 equiv). The solution was diluted with 5.5 mL of
EtOH and placed in an ice bath. Phenyl isothiocyanate (7.6 mL,
40 mmol, 1.0 equiv) was added dropwise over 10 min, Scheme 1.
The reaction mixture was warmed to room temperature and
stirred for an additional 3 h. A solution of 200 mL of 0.1 M HCl
was added cautiously to the reaction mixture and the resulting
solid collected by filtration. The filtrate was cooled to 0 ꢁC and an
additional crop of product was collected. The combined crops
were dissolved in 120 mL of acetone, dried over MgSO4 and
gravity filtered. A 5 mL aliquot of 96% H2SO4 was added care-
fully to the solution and the reaction stirred at room tempera-
ture. When TLC (1 : 1 hexanes/ethyl acetate) indicated
completion of the reaction, the acetone was removed under
reduced pressure, carefully neutralized with saturated aqueous
NaHCO3 and the resulting solid filtered to give the desired
product as a bright yellow solid in 31% yield. The NMR data
were consistent with the values reported in the literature.
To grow single crystals of Form III for X-ray diffraction, the
top coverglass was peeled off a spherulitic sample grown using
the above procedure. The bottom coverslip with crystals adhered
to it was placed on the microscope hotstage and heated slowly to
467 K in order to melt all but the largest crystal nuclei. The
temperature was then dropped to 466 K to begin crystallization,
and cycled repeatedly between 466 and 467 K to produce
a sample with only a few surviving nuclei, which were then grown
to sufficient size for analysis. No nucleation of other polymorphs
was observed during this thermal cycling. When a similar
procedure was carried out at 471 K, only the orthorhombic
polymorph of Form I was observed.
Preparation of 3-phenyl-(5Z)-5-(pyridine-2-ylmethylene)-2-thi-
oxoimidazolin-4-one4. A solution of 3-phenyl-2-thiohydantoin
(100 mg, 0.52 mmol, 1 equiv) in 10 mL of EtOH was treated with
piperidine (0.77 mL, 7.8 mmol, 15 equiv). The 2-pyr-
idinecarboxyaldehyde (0.6 mL, 6.2 mmol, 12 equiv) was added
dropwise to the reaction mixture over 5 min, Scheme 1. The
reaction was stirred under N2 at room temperature until
a precipitate formed. The solid was collected by filtration to yield
the product as a yellow solid in 93% yield. 1H NMR (DMSO-d6,
300 MHz) d 11.9 (br s, 1H), 8.8 (d, 1H, J ¼ 4.8 Hz), 8.0 (t, 1H, J ¼
8.3 Hz), 7.8 (d, 1H, J ¼ 8.0 Hz), 7.5 (m, 6H), 6.8 (s, 1H).
Preparation of 5-(pyridine-2-ylmethylene)-3-phenyl-2-methyl-
thio-3,5-dihydro-4H-imidazole-4-one6. A solution of 3-phenyl-
Fig. 2 Morphology of monoclinic spherulites of Form III grown from
an undercooled melt at 443 K.
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 3444–3450 | 3445

View Article Online
X-Ray crystallography
Form II and P21/n for Form III that yielded chemically
reasonable and computationally stable results of refinement.8,9
A successful solution by the direct methods provided most
non-hydrogen atoms from the E-map. The remaining non-
hydrogen atoms were located in an alternating series of least-
squares cycles and difference Fourier maps. All non-hydrogen
atoms were refined with anisotropic displacement coefficients.
All hydrogen atoms were included in the structure factor calcu-
lation at idealized positions and were allowed to ride on the
neighbouring atoms with relative isotropic displacement
coefficients.
Instrumentation. The data acquisitions were conducted on
a Bruker AXS APEXII diffractometer equipped with a sealed
tube Cu Ka radiation source and an Oxford Cryostream
700 cooling device and a Bruker AXS Quazar diffractometer
equipped with an ImS Mo Ka radiation microsource and an
Oxford Cryostream 600 cooling device. The Oxford Cryostream
700 was carefully calibrated in the range 100–400 K with a Si
diode model DT-421-HR-4 L (DT-421 miniature silicon diode)
and a temperature monitor model 211, both from Lakeshore
Cryotronics, Inc. The DSC curves were recorded on a TA
Instruments Q2000 differential scanning calorimeter.
In the case of Form III, two atoms, C14 and C15, had large
anisotropic displacement coefficients. The existence of only two
elongated thermal ellipsoids did not suggest a reasonable
disorder model; a disorder model was nonetheless refined, but the
outcome showed no improvement over the initial refinement; the
non-disordered structure is reported here.
X-Ray data collection. The relevant crystallographic data for
Forms I–III are tabulated in Table 1.
ꢀ
Form II: a full sphere dataset to a resolution of 0.82 A was
collected with Cu Ka radiation on a yellow crystal with
approximate dimensions 0.49 ꢂ 0.29 ꢂ 0.09 mm3 at 100(1) K.
Investigation of the Form I/Form II transition. The unit cell
determinations at 100, 130, 160, 190, 220, 250, 253, 254, 256, 258,
260, 262, 264, 266, 268, 270, 272, 280, 290, and 293 K were each
based on three 25ꢁ runs with 0.60ꢁ scans in u with an exposure
time 5 s per frame. All unit cell data acquisitions were conducted
on one crystal in a non-stop experiment in an automated mode
with Bruker APEXII software that allowed automatic tempera-
ture adjustments and equilibration delays to allow the crystal to
acclimate at each temperature for 20 min.
ꢀ
Form III: a full sphere dataset to a resolution of 0.65 A was
collected with Mo Ka radiation on a yellow crystal with
approximate dimensions 0.41 ꢂ 0.22 ꢂ 0.21 mm3 at 100(1) K.
Data processing and refinement. The data collections on crys-
tals of Form II and III were performed similarly, in a routine
fashion. The acquired scans were integrated using SAINT and
the highly redundant final dataset was corrected for Lorentz and
polarization effects.7 The absorption correction was based on
fitting a function to the empirical transmission surface as
sampled by multiple equivalent measurements using SADABS.7
The systematic absences in the diffraction data and the
E-statistics were uniquely consistent for the space group Pbca for
Powder X-ray diffraction. Powder X-ray diffraction data were
collected in Bragg-Brentano geometry on a Bruker D8 Advance
powder diffractometer equipped with a LynxEye detector.
Samples were mounted on a zero-background silicon holder, and
Table 1 Crystallographic data for the two polymorphs of 1
Polymorph
Form I (ref. 4)
Form II (this work)
Form III (this work)
Empirical formula
Formula weight
Temperature
C16H13N3OS
295.35
293 K
0.71073 A
Orthorhombic
Pbcm
C16H13N3OS
295.35
100(1) K
1.54178 A
Orthorhombic
Pbca
C16H13N3OS
295.35
100(1) K
0.71073 A
Monoclinic
P21/n
ꢀ
ꢀ
ꢀ
Wavelength
Crystal system
Space group
Unit cell dimensions
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
a ¼ 8.486(2) A
a ¼ 17.0269(4) A
a ¼ 10.1716(2) A
b ¼ 23.474(5) A
b ¼ 23.4679(6) A
b ¼ 12.1281(3) A
ꢀ
c ¼ 7.2250(10) A
b ¼ 90.0
c ¼ 6.9207(2) A
c ¼ 12.4038(3) A
ꢀ
b ¼ 90.0
b ¼ 111.8165(10) A
3
3
3
ꢀ
1439.2(5) A
ꢀ
ꢀ
1420.57(6) A
Volume
Z (Z0)
Density (calculated)
Absorption coefficient
F(000)
Crystal size
q Range for data collection
Index ranges
2765.41(12) A
8 (1)
4 (0.5)
4 (1)
1.363 Mg mꢀ3
0.227 mmꢀ1
616
1.419 Mg mꢀ3
2.094 mmꢀ1
1232
1.381 Mg mꢀ3
0.230 mmꢀ1
616
0.72 ꢂ 0.35 ꢂ 0.08 mm3
0.49 ꢂ 0.29 ꢂ 0.09 mm3
0.41 ꢂ 0.22 ꢂ 0.21 mm3
2.40–24.96ꢁ
0 # h # 10, 0 # k # 27, 0 # l # 8
3.77–71.85ꢁ
4.88–66.7ꢁ
ꢀ20 # h # 20, ꢀ28 # k # 27, ꢀ8 #
l # 8
3 8391
2682 [R(int) ¼ 0.0250]
99.7% (q ¼ 67.00ꢁ)
2682/0/191
ꢀ15 # h # 15, ꢀ18 # k # 18, ꢀ18
# l # 19
4 5264
5464 [R(int) ¼ 0.0215]
99.6% (q ¼ 66.7ꢁ)
5464/0/191
Reflections collected
Independent reflections
Completeness to q
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I > 2S(I)]
R Indices (all data)
840
840
N/A
839/0/155
1.047
R1 ¼ 0.0300, wR2 ¼ 0.0809
1.027
R1 ¼ 0.0308, wR2 ¼ 0.0880
1.052
R1 ¼ 0.0387, wR2 ¼ 0.1041
R1 ¼ 0.0300, wR2 ¼ 0.08
R1 ¼ 0.0317, wR2 ¼ 0.0892
R1 ¼ 0.0425, wR2 ¼ 0.1076
ꢀ3
ꢀ3
ꢀ3
ꢀ
ꢀ
ꢀ
Largest diff. peak and hole
0.149 and ꢀ0.158 e A
0.281 and ꢀ0.413 e A
0.561 and ꢀ0.713 e A
3446 | CrystEngComm, 2011, 13, 3444–3450
This journal is ª The Royal Society of Chemistry 2011

View Article Online
Table 2 Transformations of the polymorphs of 1
The onset of the Form III / Form I transformation in the
presence of seed crystals of Form I was observed to commence at
temperatures ranging between 420 and 448 K in our three
experiments carried out with heating rates of 10ꢁ per min.
Transformation
Temperature (onset)/K
Enthalpy/kJ molꢀ1
Form I fusion
480.7(5)
466.3–468.1
262.9(5)
37.9(2)
33.8(2)
0.3(1)
ꢀ1.7(4)
Form III fusion
Form I/Form II
Form III / Form I
462(1)
Results and discussion
Phase transitions
scans were collected between 5 and 35ꢁ in 2q with a step size of
0.02ꢁ and an acquisition time of 5 s. Samples were randomized by
gentle grinding prior to measurement to reduce the effects of
preferred orientation.
Three phases of 1, Fig. 1, exist: the orthorhombic Form I (space
group Pbcm), produced by crystallization from solution, is stable
between 263(1) and 480.7(5) K; the orthorhombic Form II (space
group Pbca), obtained by cooling Form I below 263(1) K, is
stable between 100(1) and 263(1) K; the monoclinic Form III
(space group P21/n), generated from the melt by cycling the seed
crystals between 466 and 467 K, is metastable between 100(1) and
466(1) K. The transformation between Forms I/II is an enan-
tiotropic first-order solid-state phase transition classified as type
k2.10 Form III, when in contact with seed crystals of Form I,
upon heating transforms into the higher melting Form I (crystals
of Form I grow with a visible growth front) in a first-order
monotropic phase transition of type t2.
Differential scanning calorimetry. The temperatures and
enthalpies of transformations of the polymorphs of 1 are
summarized in Table 2.
Investigation of the Form I/Form II transition. Three experi-
ments were conducted with ꢃ4 mg of Form I. The material was
loaded in an aluminium Tzero pan and heating curves for the
phase transition were collected from 250 K to 273 K followed by
cooling of the sample back down to 250 K and heating it to
523 K. The heating and cooling were performed at a rate of
10ꢁ per min.
The Form I/Form II transition was first detected by X-ray
diffractometry and later studied both by diffractometry and by
DSC. The DSC heating curve clearly indicated the onset of the
phase transition at 262.9(5) with DH ¼ 0.3(1) kJ molꢀ1. There is
no visible change in the crystals during this phase transition and
the IR spectra of Form I and Form II are nearly identical.
The relationship between the unit cells of Form I and Form II
is shown in Fig. 3. The a axis of the Form II is twice the length of
that in the Form I whereas the b and c axes are very similar
between the polymorphs. As a result of the difference of the
a axial length, the volume and Z of the Form II are two times
larger than those of the Form I. Fig. 3 shows one unit cell for the
Form II, two unit cells for the Form I, and the symmetry
elements to illustrate the differences between the two poly-
morphs. The most significant difference is that glide planes
perpendicular to the c axis in the Form II become mirror planes
perpendicular to the c axis in the Form I. This is the result of
molecules of 1 adopting a Cs-symmetrical conformation thus the
molecules of 1 reside on crystallographic mirror planes in the
Form I. The two-fold screw axes parallel to the a axis in the Form
II are ‘‘downgraded’’ to two-fold axes in the Form I.
Discovery of Form III. The first DSC heating curve for
a sample of 1, Form I, was recorded between 250 and 523 K and
showed the two expected transformations, the Form II/Form I
transition and the melting of Form I at 481 K. Reheating of the
sample produced evidence of a Tg (indicating that the sample did
not fully crystallize during cooling). Subsequent crystallization
was followed by melting and recrystallization at about 460 K,
and final melting at 481 K. All peaks were broad and temperature
depressed because of partial sample decomposition. These results
indicated the existence of a new polymorph that coexisted with
Form I below 460 K. The new polymorph was isolated as
described in the Experimental section by cycling of the seed
crystals of the new Form III polymorph between 466 and 467 K
in order to isolate large enough crystals for characterization.
Samples of the monoclinic Form III were prepared for DSC
analysis by melting 3–5 mg of material in a Tzero aluminium pan
at 491 K for 30 seconds to ensure complete melting, verified by
visual inspection. The sample was rapidly quenched to room
temperature, and then heated at 443 K for ꢃ30 seconds to
complete crystallization of the monoclinic Form III. Even a very
small amount of orthorhombic polymorph (Form I) present in
contact with Form III is sufficient to induce phase trans-
formation during heating in the DSC cycle, hence preparation of
a phase-pure sample is essential for measurement of the thermal
properties. DSC melting curves were collected between 298 and
503 K at a heating rate of 10ꢁ per min.
Phase transition Form II / Form I was monitored by single
crystal X-ray diffractometry first by cooling a crystal of 1 to
100 K and then slowly warming it to 300 K. The unit cell was
redetermined stepwise every 1–30 K in a procedure analogous to
the one previously reported by us.11 The temperature dependence
of the axial lengths is plotted in Fig. 4. Upon heating from 100 K,
the a axial length slowly contracts until the onset of the phase
transition at 263(1) K, at which point its length is halved. Further
heating to 300 K is characterized by a small positive thermal
expansion coefficient in the [100] direction. The b axis remains
virtually unchanged throughout the temperature range whereas
the c axis has a small positive thermal expansion coefficient. The
phase transition was also monitored by following the intensities
of two reflections with hkl, h ¼ even (662) and hkl, h ¼ odd (762),
Fig. 5. Upon heating the crystal from 100 K, the intensities of
both reflections decrease at a similar rate, an effect associated
Investigation of the Form III / Form I transition. The new
polymorph, metastable Form III, showed no signs of conversion
to Form I after several days at RT and could be examined in the
entire range between 100 and 462 K. At 462(1) K it melts in the
absence of seed crystals of Form I or undergoes an exothermic
phase transition into Form I in the presence of crystals of Form I.
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 3444–3450 | 3447

View Article Online
in P21/n, and it may also crystallize in Pbcm or possibly Pca21—
the crystal and data quality did not allow precise structural
determination in the orthorhombic space group.13 rac-cis-a-
(Triethylenetetramine)-bis(nitrito-N)-rhodium chloride upon
heating undergoes two phase transitions, Pbcm / C2/c / P21/n,
thus no Pbcm / P21/n transformation has been investigated.14
The CSD contains no compounds that form triads of poly-
morphs in space group numbers 14/57/61 (P21/n, Pbcm, Pbca)
observed for Forms I–III.
Molecular geometry
The molecular geometries of the three polymorphs of 1 are
similar overall due to the few rotational degrees of freedom in the
molecule, yet show noticeable differences. The molecule of 1 in
Form I resided in a crystallographic symmetry plane and thus
possesses Cs symmetry. The molecules in Forms II and III are
devoid of symmetry. Table 3 lists selected structural parameters
for the three polymorphs of 1, while Fig. 6 illustrates the
differences among the molecules. The most striking feature is the
orientation of ring C (see designation in Fig. 1): in Form I it is at
the right angle to the plane of ring B; in Form II ring C is tilted by
ꢃ20ꢁ relative to ring B; in Form III ring C forms only a ꢃ10ꢁ
dihedral angle with ring B, but now ring C is laterally displaced
to one side rendering atom N1 pyramidal. The displacement of
Fig. 3 The relationship between the two orthorhombic polymorphic
Forms I and II of 1. Two unit cells are shown for the Form I and one for
Form II. Note that the molecules of 1, represented by black and open
circles in the Form II, are mirror images of each other.
with the increase in atomic thermal motion. At 263(1) K reflec-
tion (762) disappears, corresponding to the doubling of the a axis
and providing further evidence of a phase transformation.
The phase transition between the space groups Pbcm and Pbca
has been heretofore reported only once, for acetone.12 Thus,
literature comparisons and generalizations in this particular case
are not possible. Nonetheless, our data mining of the 15 438
polymorphs (the ones with known space groups) of 6243
compounds as reported to the CSD uncovered 26 polymorphs
with space group Pbcm and 597 polymorphs with Pbca. For
Pbcm, there are 21 types of polymorphic pairs with different
second space groups, thus no meaningful trend can be identified.
In the case of Pbca, there are 39 types of polymorphic pairs and
the second space groups observed with high frequency are listed
ꢀ
atom C10 from plane B is a substantial ꢀ0.2960(16) A.
Another way to compare and contrast the geometries in Forms
I–III is to examine the rms deviations for superimposed struc-
tures computed (a) for all 21 non-hydrogen atoms and (b) for all
non-hydrogen atom except for the six atoms of ring C (a total of
2
ꢀ
15 atoms). These values comprised 0.195 and 0.095 A for Forms
2
I/II, 0.274 and 0.067 A for Forms I/III, and 0.328 and 0.104 A
2
ꢀ
ꢀ
for Forms II/III. Consequently, whereas the atoms of rings A
and B are in similar positions, ring C is the part of the molecules
exhibiting the conformational variations. It is also noteworthy
that the N1–C10 bond length connecting rings B and C in Form
III is probably statistically significantly shorter than those in
Forms I and II. This is a counterintuitive result because the bond
distances should increase with the loss of delocalization of
p-electron density.
ꢁ
here in the descending order: P21/c, P1, C2/c, Pca21, Cmca,
P212121, and Cmcm.
There are only two compounds to have polymorphic pairs with
space groups Pbcm–P21/n reported to the CSD, however no
reports of direct transformation between these space groups have
been published. Pyrazinium tetrafluoroborate was characterized
The three polymorphs have also been characterized by X-ray
powder diffraction. The experimental patterns for Forms I and
III and simulated pattern for Form II are shown in Fig. 7. The
patterns for Form I and II have several features in common with
Fig. 4 The axes length of 1 between 100 K and 300 K. The a axis is the
middle red line, the b axis is the top green line, and the c axis is the bottom
purple line.
Fig. 5 Temperature dependency of two representative hkl reflections.
The blue top line is hkl ¼ 662, h ¼ even and the red bottom line is hkl ¼
762, h ¼ odd.
3448 | CrystEngComm, 2011, 13, 3444–3450
This journal is ª The Royal Society of Chemistry 2011

View Article Online
Table 3 Selected geometrical parameters of 1 in its three polymorphs. The ring labels are from Fig. 1
Torsion
angle/deg
Vector/plane
angle/deg
Bond
length/A
Dihedral
angles/deg
ꢀ
Displacement of
C10 from plane B/A
Sum of bond angles
about N1/deg
ꢀ
Rings A/B
Rings B/C
N2–C1–S1–C16
N1–C10/ring B
N1–C10
Form I
Form II
Form III
0
90
69.56(4)
80.16(4)
0
0
1.435(3)
1.4325(14)
1.4279(12)
0
360
359.98(10)
357.22(8)
7.21(2)
4.92(7)
3.99(11)
5.97(10)
1.57(8)
12.66(7)
ꢀ0.0445(18)
ꢀ0.2960(16)
index value for Forms I/II of 0.959 is close to unity, attesting to
the resemblance of spatial molecular arrangements in these
lattices. Mercury attempted to match 15 molecules from each
lattice and successfully matched all 15. The similarity index
values for Forms I/III and II/III were 0.895 and 0.917 and in both
cases the algorithm matched only 2 out of 15 molecules. The
molecular arrangement in the lattice of Form III is very different
from that in the other polymorphs. This is illustrated in Fig. 8.
To gain insight into the thermodynamic aspects of the phase
transition several conformations of 1 were optimized with
density functional theory methods at the B3LYP/6-31+G(d) level
of theory.17 The molecular geometry with the Cs symmetry
(observed in the Form I) was 0.50 kJ molꢀ1 less stable than the
geometry with the C3–N1–C10–C11 torsion angle constrained to
69.56(4)ꢁ observed in the Form II. The difference is very small
and may be statistically insignificant given the precision of the
method. The Cs-symmetrized geometry is not an energy
minimum but rather a maximum characterizing a first-order
transition point corresponding to the rotation of the C10 phenyl
ring about the N1–C10 bond. An attempt to optimize the
experimentally observed low temperature geometry without
constraints did not converge and revealed a very shallow
minimum on the potential energy surface near the C3–N1–C10–
C15 torsion angle value of about 111ꢁ. A subsequent relaxed
Fig. 6 The three polymorphs of 1 viewed along the pyrazine ring. Left to
right: Form I, II, and III. Color codes: grey for C, blue for N, red for O,
and yellow for S.
the largest differences observed in the peak positions at ꢃ18ꢁ and
between 24 and 28ꢁ in 2q. The pattern of Form III is distinctly
different.
The similarity of the molecular arrangements in the lattices of
Forms I–III was assessed with the crystal structure similarity
index based on the algorithm of De Gelder, Wehrens and
Hageman15 incorporated into program Mercury.16 The similarity
Fig. 8 The overlay of the lattices of the orthorhombic Form I (pink) and
monoclinic Form III (blue). The ‘‘anchor’’ molecules of the two poly-
morphs with the closest fit are displayed as ball-n-stick models.
Fig. 7 Powder patterns of the three polymorphs of 1. From bottom to
top: Form I, II (simulated), and III.
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 3444–3450 | 3449

View Article Online
potential energy surface scan for the C3–N1–C10–C11 torsion
angle varied between 0 and 90ꢁ with 3ꢁ steps was characterized by
an energy change of less than 0.67 kJ molꢀ1 for the angle span
between 51 and 90ꢁ. Thus, whereas the molecular conformation
of Form I is not as stable as that of Form II, its packing must be
more energetically favorable.
Acknowledgements
We are grateful to Prof. Lian Yu (UW-Madison) for the access to
his DSC and microscope equipment and to Dr. Evgueni G.
Mednikov (UW-Madison) and Dr. Bruce Noll (Bruker AXS) for
help with several experiments. EMG acknowledges the support
of the PhRMA foundation (postdoctoral research fellowship).
The molecular geometry of Form III was optimized with the
bond angles about the N1 atom and the ring B/ring C torsion
angle constrained to the experimentally observed values. This
conformation is 2.81 kJ molꢀ1 less stable than that of Form I. An
inspection of the packing of Form III in the lattice does not
immediately uncover the reason for the existence and stability of
this phase. It appears that there is not enough room around ring
C to allow an alignment of the N1–C10 vector with ring B,
however in principle the alignment can occur by the movement of
the rest of the molecule. There are several weak non-classical
interactions among the molecules in crystals of Forms I–III but
drawing robust conclusions from them in order to rationalize the
nature of the polymorphism of 1 is not feasible due to the
unknown interaction energies associated with individual
contacts.
References
1 E. K. Beloglazkina, A. G. Majouga, A. A. Moiseeva, N. V. Zyk and
N. S. Zefirov, Mendeleev Commun., 2009, 19, 69.
2 T.R.N. Project, Indiana University, 2002.
3 F. H. Allen, Acta Crystallogr., Sect. B: Struct. Sci., 2002, 58, 380.
4 A. G. Majouga, E. K. Beloglazkina, S. Z. Vatsadze, N. A. Frolova
and N. V. Zyk, Russ. Chem. Bull., 2004, 53, 2850.
5 D. M. Wolfe and P. R. Schreiner, Synthesis, 2007, 2002.
6 A. G. Majouga, E. K. Beloglazkina, S. Z. Vatsadze, A. A. Moiseeva,
K. P. Butin and N. V. Zyk, Mendeleev Commun., 2004, 3, 115.
7 APEXII, Sadabs, and SAINT, Bruker AXS Inc., Madison, WI, USA,
2000–2007.
8 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
64, 112.
9 I. A. Guzei, Internal Laboratory Computer Programs FCF_Filter and
Modicifer, 2006–2008.
10 S. Ivantchev, E. Kroumova, G. Madariaga, J. M. Perez-Mato and
M. I. Aroyo, J. Appl. Crystallogr., 2000, 33, 1190.
11 I. A. Guzei, A. Mitra and L. C. Spencer, Cryst. Growth Des., 2009, 9,
2287.
12 D. R. Allan, S. J. Clark, R. M. Ibberson, S. Parsons, C. R. Pulham
and L. Sawyer, Chem. Commun., 1999, 751.
ꢂ
13 A. Katrusiak and M. Szafranski, J. Am. Chem. Soc., 2006, 128,
15775–15785.
14 I. Bernal and J. Cetrullo, Inorg. Chim. Acta, 1988, 144, 227–232.
15 R. De Gelder, R. Wehrens and J. A. Hageman, J. Comput. Chem.,
2001, 22, 273–289.
16 I. J. Bruno, J. C. Cole, M. Kessler, J. Luo, W. D. S. Motherwell,
L. H. Purkis, B. R. Smith, R. Taylor, R. I. Cooper, S. E. Harris
and A. G. Orpen, J. Chem. Inf. Comput. Sci., 2004, 44, 2133.
17 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi,
V. Barone, B. Mennucci, et al., Gaussian 03, Revision C.02,
Gaussian, Inc., Wallingford, CT, USA, 2004.
Conclusions
Three polymorphs of 1 have been prepared and structurally
characterized. The most common orthorhombic Form I can be
obtained by crystallization from solution and is stable above
263(1) K. Upon cooling below 263(1) K it undergoes a first-order
enantiotropic k2 phase transition into the orthorhombic Form II
which contains 1 in a slightly more stable molecular conforma-
tion. When 1 is molten and cooled the metastable monoclinic
Form III can be isolated. Form III is indefinitely stable at RT
and can be examined in the entire range between 100 and its
melting point of 466 K. The molecular geometries of 1 differ
most noticeably in the position of the phenyl ring relative to the
rest of the molecule. The packing arrangements in Forms I and II
are similar but differ substantially from that in Form III.
3450 | CrystEngComm, 2011, 13, 3444–3450
This journal is ª The Royal Society of Chemistry 2011