On the solid-state isomerization of terpinolene tetrabromide

Article abstract of DOI:10.1071/CH09606

The isomerization of terpinolene tetrabromide from the cis-2,4- (diaxial) dibromo isomer (A) to the more stable trans-2,4- (equatorialaxial) compound (B) has been investigated in the solid state. The progress of the solid-state reaction as a function of t

Full text of DOI:10.1071/CH09606

Full Paper
CSIRO PUBLISHING
Aust. J. Chem. 2010, 63, 458–462
www.publish.csiro.au/journals/ajc
On the Solid-State Isomerization ofTerpinoleneTetrabromide
Paul V. Bernhardt,A,B Raymond M. Carman,A and Tri T. LeA
ASchool of Chemistry and Molecular Biosciences, University of Queensland, Brisbane,
Qld 4072, Australia.
BCorresponding author. Email: p.bernhardt@uq.edu.au
The isomerization of terpinolene tetrabromide from the cis-2,4- (diaxial) dibromo isomer (A) to the more stable trans-2,4-
(equatorial–axial) compound (B) has been investigated in the solid state. The progress of the solid-state reaction as a
function of time was followed by solution NMR analysis, and, more effectively, by micropowder X-ray diffraction of the
bulk material in situ. Characterization of the solid-state reaction mixture is facilitated by the existence of X-ray crystal
structures of both compounds A (monoclinic) and B (reported previously) in addition to a new triclinic polymorph of
compound A reported herein, all three forms being observed by powder X-ray diffraction.
Manuscript received: 24 November 2009.
Manuscript accepted: 16 January 2010.
Br
Br
Introduction
Br
8
1
The rearrangement in solution of terpinolene tetrabromide (com-
pound A, Scheme 1) into the isomeric compound B has been
discussed previously.[1–4] Prior to these studies, several other
workershadbeenaware[5–9] ofthepresenceoftwodifferentcrys-
talline forms of ‘terpinolene tetrabromide’, but at the time they
were unable to draw on the physical characterization methods
now available, such as NMR and X-ray crystallography.
BrϪ
8
1
Me
Me
Me
2
Me
Me
Me
H
2
4
H
4
Brϩ
Br
Br
Br
Compound A
Br
Br
We were able to relate optical crystal measurements made
long ago[5–7] with crystals we now know to be monoclinic
compound A, which we characterized by single-crystal X-ray
crystallography.[4] Briefly, the cell dimensions we determined
for monoclinic compound A, when transformed to its equivalent
space group P21/a (a 15.423, b 9.915, c 9.477 Å, β 113.93◦),
match the optical measurements made on a crystalline sample
of ‘Compound A’ reported by Henry and Paget in 1931[7] and
Hintze earlier still[6] in 1885, predating X-ray crystallography
(or even X-rays themselves).
1
Ring inversion
Br
Me
Me
Me
H
4
Br
4
2
H
2
1
8
8
Me
Me
Br
Br
Br
Br
Me
Compound B
Scheme 1.
Since its original isolation, compoundA has caused consider-
able confusion in several laboratories. In 1971, M. D. Sutherland
supplied one of the present authors (R.M.C.) with a sample (pos-
sibly the Owen[8] 1955 sample) believed to be compound A, but
thatprovedtobecompound B.# B. N.Venzke, whoworkedexten-
sively with compound A,[2,3] experienced similar difficulties
later in the 1970s with crystals of compound A he had pre-
pared himself. It is not now possible to know how these samples
had been stored, or the temperature of their final recrystalliza-
tion, but both Sutherland and Venzke were meticulous in the
labelling and storage of their compounds, always in the dark and
normally in the refrigerator. Earlier, Sutherland reported[9] that
‘unstable’ (compound A) and ‘stable’ (compound B) tetrabro-
mides had already caused confusion in his laboratory, and he
advised recrystallization from hot solvent to force conversion
to the more stable isomer and simplify characterization of ter-
pinolene. Independently, Briasco and Murray reported[10] that a
freshly made tetrabromide (presumably compound A) was dif-
ferent from a sample of terpinolene tetrabromide they believed
to be ‘authentic’ (but presumably compound B).
We have previously shown,[2,3] from both labelling and chi-
rality experiments, that the rearrangement in solution involves
substitution reactions at C4 and C8, and that bromide ions par-
ticipate in the mechanism, but do not catalyze the process. The
isomerization of compoundA into compound B relieves the 2,4-
diaxial dibromo repulsion present in compound A (Scheme 1),
and there is no doubt that compound B is more stable and there is
no detectable presence of compound A (<0.5%) at equilibrium.
# According to his MSc dissertation, Owen heated one of his preparations of terpinolene tetrabromide on a water bath to remove excess bromine, but then
obtained compound B (‘β’). He also reports that an old comparison sample available in the Sutherland laboratory, although still crystalline, was ‘much
decomposed’, and afforded compound B (‘β’) on recrystallization. It is not clear in his work how Owen distinguished his α compound, mp 116–117.5◦C
(now known to be compound A) from his β isomer (compound B), mp 117–118◦C.
© CSIRO 2010
10.1071/CH09606
0004-9425/10/030458

Solid-State Isomerization of Terpinolene Tetrabromide
459
(a)
The reaction has a half-life, at room temperature, of a few days
in solution in the absence of any other species.[2–4]
Br4
Br2
C1
C6
We too have experienced similar unexpected results with this
system through assuming that preservation of the crystalline
appearance of compounds A and B over time was indicative
of retention of structure, only to be surprised by the appear-
ance of compound B from a sample thought to be compound A.
This ongoing controversy has prompted us to look more closely
at the compositions of solid samples of the isomers A and B
of terpinolene tetrabromide as a function of time, and we have
identified an unusual solid-state reaction whereby a crystalline
sample of compound A converts to a microcrystalline sample
of compound B, which parallels the known interconversion in
solution.[4]
C7
C10
C9
C4
C8
C5
C2
C3
Br1
Br8
(b)
Br4
Br2
C6
Results and Discussion
C10
C7
C4
Solid-State Isomerization
C5
Compound A was stored in the dark at room temperature and
examined at various intervals by NMR spectroscopy. When it
became clear that the rearrangement A to B was slow under
these conditions (∼1% per week at ambient temperature), a par-
allel set of experiments was conducted on material stored at 37◦
and 67◦C (both kept in the dark). Samples were dissolved in
CDCl3 immediately before NMR examination, as compound A
is known to isomerize in solution to give significant quantities of
compound B within a few minutes.[4] We also found that traces
of bromine accelerate this isomerization in solution and in the
solid state (see below).
C8
C1
C2
C3
C9
Br1
Br8
Fig. 1. ORTEP views of compound A in its monoclinic (a) and triclinic
(b) forms; 30% probability ellipsoids shown. The conformations of the two
molecules are indistinguishable.
NMR analysis was straightforward by integration of the
methine H2 resonance (δ 4.81 (dt) of compound A; δ 4.91 (dd)
of compound B); see Scheme 1 for atom numbering and Acces-
sory Publication Fig. S1 for an example.Analytical samples were
prepared from crystals lightly ground in a mortar and pestle
so as to ensure representative sampling. Control experiments
showed that brief grinding did not significantly affect the ratios
of compounds A and B.
and a view of the molecule is shown in Fig. 1 (left). No signifi-
cant differences are found between the 150 and 293 K structures
of compound A apart from the expected thermal expansion of
unit cell volume; both structures are isomorphous (monoclinic,
P21/n) and no phase transformation occurs between 150 and
293 K.
A New Polymorph: Triclinic Compound A
Visual examination of the crystals during the rearrangement
showed a diminution of the lustre and size of individual mono-
clinic crystals of A, with the appearance of a very slight
purple tinge, but the material remained crystalline. Individual
(unground) crystals of A were also examined by NMR spec-
troscopy at various times. The reaction is likely to proceed by
an intramolecular mechanism as shown in Scheme 1, although
the migrating bromide ‘ion’ need not necessarily be as free as is
depicted. Possible intermediates in the interconversion between
compoundsA and B may involve theloss of Br2 or HBr (although
no vinyl carbons were observed in NMR spectra), and at least
bromine might be expected to effectively catalyze the reaction
through the formation of Br−3 .
A single crystal for X-ray work was selected from a sample orig-
inally of compound A (that had been stored at 37◦C for several
weeks), which in the solid state had apparently rearranged sub-
stantially to compound B by NMR analysis. To our surprise,
this crystal proved to be compound A, although in a previously
unknown triclinic form. The molecular structure of the triclinic
form of compound A is shown in Fig. 1b, and clearly is indis-
tinguishable from that of monoclinic compound A (Fig. 1a). A
better appreciation of the structural differences between mono-
clinic and triclinic compound A is gained by inspection of
their lattices (Accessory Publication Fig. S2a). The presence of
a two-fold screw axis and n glide plane in monoclinic com-
pound A results in a very different arrangement of molecules
compared with the triclinic form, where centres of inversion are
the only symmetry elements present. The disparate lattices of
monoclinic and triclinic forms of compound A lead to an impor-
tant conclusion: the two polymorphs are not related by a simple
reversible phase transition but instead they exist independently.
Furthermore, the isolation of this triclinic form as a single crys-
tal (among a mixture of compounds A and B) is remarkable, as
genuine recrystallization has occurred in the solid state. As we
shall show, triclinic compound A is metastable and has never
been obtained in pure form on recrystallization of compound A,
which always appears in its monoclinic form owing evidently to
its lower solubility.
The Monoclinic Crystal Structure of Compound A
A crystalline sample of freshly recrystallized compound A was
examined by single-crystal X-ray crystallography. In our previ-
ous structural study employing a point detector diffractometer,[4]
the required ∼24 h-data acquisition period was too long for the
crystals of compound A to survive at room temperature, which
necessitated data collection at low temperature (150 K). Now
with the benefit of a charge coupled device (CCD) diffracto-
meter, we were able to collect a full dataset of compound A at
room temperature within an hour before any significant crys-
tal degradation was observed. This structure is reported herein

460
P. V. Bernhardt, R. M. Carman, and T. T. Le
Monoclinic compound A
t ϭ 0
Monoclinic
compound A
45 min
Triclinic compound A
Compound B
16 h
Triclinic
compound A
2 days
3 days
6 days
Compound B
5
10
15
20
25
30
35
40
45
5
10
15
20
25
30
35
40
45
2θ [°]
2θ [°]
Fig. 2. Calculated powder X-ray diffraction patterns (CuKα radiation,
λ = 1.5418 Å, T = 293 K) for the monoclinc (top) and triclinic (middle)
forms of compound A and compound B (bottom). Profiles calculated with
the program Mercury using the atomic coordinates of their known crystal
structures published here and previously.[4]
Fig. 3. Experimental X-ray diffraction patterns for solids that had under-
gone thermally induced rearrangement in the solid state at 67◦C. The
conversion from monoclinic compound A to triclinic compound A, then
to compound B is apparent.
100
Polymorphism such as that found here in compound A is not
at all unusual; ∼5% of compounds in the Cambridge Structural
Database exhibit polymorphism.[11] It has also been suggested
that the observation of polymorphism in any compound is cor-
related with the level of intensity with which it is studied,[12]
e.g. chemicals of commercial importance are commonly found
in polymorphic forms, so the actual proportion of compounds
exhibiting polymorphism is most likely much greater than 5%.
Inspection of the intermolecular Br4· · ·Br8 contacts in the
monoclinic and triclinic structures of compound A is revealing
(Accessory Publication Fig. S2b). In the monoclinic structure,
the Br4· · ·Br8 intermolecular contacts (3.600 Å) are 0.100 Å
shorter than the sum of their van der Waals radii, whereas in the
triclinic form, these non-bonded repulsions are relieved some-
what with Br4· · ·Br8 distances of 3.679 Å. Generally speaking,
the triclinic form of compound A is less tightly packed, which is
supported by its slightly lower density relative to the monoclinic
form.
75
% monoclinic compound A
% triclinic compound A
% compound B
50
25
0
0
1
2
3
4
5
6
7
Time [days]
Fig. 4. Speciation plot as a function of time of the three species present
during the solid-state reaction at 67◦C (from Fig. 3).
Time-Dependent Powder XRD Analysis
Micropowder X-ray diffraction (XRD) patterns of all solid sam-
ples were acquired with CuKα radiation (employing the same
diffractometer used for single-crystal work) on solid samples
that had been finely ground in a mortar and pestle and packed
into glass capillaries (∼0.3 mm internal diameter). This tech-
nique is useful in that miniscule amounts of compound (<1 mg)
are required. The calculated XRD patterns were generated with
the program Mercury (version 2.2, Cambridge Crystallographic
Data Centre) using the known 293 K crystal structures of mono-
clinic A and triclinic A (reported herein) and compound B
published earlier, and these are shown in Fig. 2, all normalized
to their most intense peak.[4]
The percentage of the three species was calculated at each
point in time (Fig. 4) from the relative intensities of peaks in
the region 7 < 2θ < 13◦, where monoclinic and triclinic com-
pound A and compound B show well-separated XRD peaks (see
Fig. 2). Monoclinic compound A exhibits a unique XRD peak at
2θ = 10.8◦ (corresponding to the 0 1 1 reflection), triclinic com-
pound A shows a peak at 11.9◦ (0 −1 1), whereas compound B
gives a low-angle XRD peak at 7.9◦ (0 0 2).
Initially (Fig. 3, top), the XRD pattern is essentially that of
pure monoclinic compoundA (cf. Fig. 2, top), with a very minor
(<5%) amount of its triclinic polymorph. After 16 h at 67◦C, no
monoclinic compound A remains and the XRD pattern corre-
sponds to 80% triclinic compound A (cf. Fig. 2, centre) and
20% compound B. By Day 6, the solid is 100% compound B
(cf. Fig. 2, bottom). Similar behaviour was seen at 37◦C but
the rate of conversion was slower by a factor of ∼7, i.e. after
Solid samples that had been subjected to heating (at 37◦
or 67◦C) for various periods gave XRD patterns that could be
accounted for as mixtures of monoclinic compound A, triclinic
compound A and compound B. An example is shown in Fig. 3
for the solid-state reaction starting with monoclinic compoundA
at 67◦C.

Solid-State Isomerization of Terpinolene Tetrabromide
461
2 weeks, the reaction at 37◦C had reached the same point as
after 2 days at 67◦C (Accessory Publication Fig. S3). At 25◦C,
the reaction was very slow; after 3 weeks, the mixture was ∼65%
monoclinic and 35% triclinic compoundA with no apparent iso-
merization to compound B (Accessory Publication Fig. S4).This
result was confirmed by NMR analysis of the solid (see Experi-
mental section). By comparison, the solid-state reaction at 67◦C
had reached this same stage after only 45 min.
even in the solid state, an observation noted in several differ-
ent laboratories but until now not understood. In the course of
our investigations, the serendipitous observation of a triclinic
polymorph of compound A has enabled us to identify it as an
intermediate in the solid-state conversion of monoclinic com-
pound A to compound B. The relief of Br4· · ·Br8 contacts in the
triclinic form of compound A apparently aid the rearrangement
to compound B as required in Scheme 1, where both Br4 and
Br8 must migrate.
No quantitative analysis of the rates of these reactions has
been attempted. Over several experiments conducted at the same
temperature, we noted that the rates of conversion from mono-
clinic to triclinic compound A were dependent on degrees of
grinding and also the purity of the starting material (mono-
clinic compound A). Also, as the course of each reaction
proceeded, it was noted that the appearance of the sample
changed from a finely mechanically ground powder (initially)
to coarse needle-like crystals of the final product compound B
(Accessory Publication Fig. S5). As all time-dependent exper-
iments were carried out in situ with the same capillary, it was
not possible to re-grind the compound once mounted so as to
restore microcrystallinity. As a consequence, we observed that
heterogeneity in the morphology of the sample developed with
time. If the sample was centred in the X-ray beam in a region
dominated by coarse crystalline needles, the XRD pattern lost its
ring-like appearance and spots emerged emanating from diffrac-
tion of single crystals. The ensuing XRD trace was necessarily
poor owing to peak broadening and proper quantitative anal-
ysis was difficult. However, sampling different regions within
the same capillary did not reveal any variations in composition
with time.
Solid-state (solvent-free) transformations of organic com-
pounds are a novel approach to molecules that may be inac-
cessible by conventional synthetic routes and this area has been
comprehensively reviewed by Kaupp.[13] The mechanisms of
solid-state reactions also present challenges.[14] The present
example of the simultaneous migration of two Br atoms is to our
knowledge unique among the wide range of known solid-state
organic reactions[13] and the mechanism of this reaction is even
more intriguing given the emergence of triclinic compound A as
a likely intermediate.
The mechanism shown in Scheme 1 requires migration of
only Br4 and Br8 during the conversion of compound A to
compound B. We have shown that packing in monoclinic com-
poundA results in closer intermolecular Br4· · ·Br8 contacts that
are relieved significantly in the triclinic form. The relief of this
congestion in the triclinic form appears to trigger the progres-
sion of this reaction to the stable final product, compound B. In
other words, the monoclinic form of compound A is stabilized
(kinetically) in the solid state against isomerization to compound
B by congestion around atoms Br4 and Br8.
Experimental
NMR spectra were collected in CDCl3 solution at ambient tem-
perature using a Bruker 500 MHz spectrometer. All solutions
were made up immediately before examination. The compounds
decompose rapidly in the presence of light, so were always kept
in the dark when not in use.
Catalysis by Br2
A crystalline sample of compound A stored for 1 week at 18◦C
in a sealed vial with a trace of bromine vapour provided 100%
conversion into compound B as shown by powder XRD. A trace
of acetic acid provided no significant rate increase. In all exper-
iments with added bromine vapour, care needed to be taken to
ensure that no free bromine remained in the crystal before NMR
analysis, as free bromine greatly catalyzes the rearrangement in
solution, thus exaggerating the actual amount of compound B.
Thus a sample of compound A in CDCl3 containing a trace of
bromine provided 93% conversion to compound B as rapidly as
the spectrum could be obtained.
Tetrabromide Compound A
Tetrabromide compound A was prepared from terpinolene as
previously described,[4] and purified by recrystallization from
acetone at −18◦C to provide colourless crystals, with NMR data
as previously reported.[4] The compound was racemic, and the
crystals monoclinic.[4]
As compound A is prepared by the addition of bromine to
terpinolene, and as bromine catalyzes the rearrangement of com-
pound A to compound B, other workers may well find that their
samples rearrange at varying rates depending on their initial state
of purity. We recommend that compound A be stored as pure as
possible, at the lowest available temperature and in the absence
of light. Acetone provides a useful solvent for low-temperature
recrystallization as it scavenges any residual traces of bromine.
When solid samples that had undergone partial isomerization to
compound B were freed from any residual bromine by evapora-
tion, consistentNMRandXRDresultswereobtained(Accessory
Publication Fig. S6).
Tetrabromide Compound B
Tetrabromide compound A was refluxed in ethanol for 20 min,
when all the material was in solution. Cooling provided crystals
that were recrystallized from either acetone or ethanol to provide
authentic compound B with NMR data as previously reported.[4]
The compound was racemic, and the crystals monoclinic.[4]
Compound B stored in the dark at 37◦C for 20 weeks showed
little signs of decomposition.
Solid-State Isomerization of Compound A to B
Compound A stored in the dark in a sealed container at 25◦C
showed 0.3, 0.8, 1.5, 2.3, 26, 30, 32, 34% conversion to com-
pound Bafter0, 1, 2, 5, 10, 16, 17, 24weeksrespectivelybyNMR
analysis. The same material stored at 37◦C in a sealed container
showed 29, 38, 45, 58, 72, 82, and 98% conversion to com-
pound B after 1, 2, 3, 4, 5, 6, and 7 weeks respectively by NMR
analysis on solutions made up immediately before analysis.
Conclusions
The present study has provided answers to some longstanding
questions of how compound A, bearing a pair of 2,4-diaxial
bromine substituents on the cyclohexyl ring, is able to convert
to its more stable 2,4-equatorial–axial isomeric compound B

462
P. V. Bernhardt, R. M. Carman, and T. T. Le
Quantitative Validation of NMR (Solution)
and XRD (Solid-State) Analysis
Micropowder XRD profiles were also measured on an Oxford
Diffraction Gemini S Ultra CCD diffractometer using Cu-Kα
radiation (λ 1.5418 Å). Samples were ground finely and mounted
in a glass capillary (diameter ∼0.3 mm). A broad and asym-
metric amorphous scattering peak from the glass capillary was
observed centred at ∼2θ = 22◦. This was compensated for by
subtracting the XRD trace of an empty capillary that matched
that of amorphous vitreous silica.[18]
A 2:1 mixture of compounds A and B respectively was heated
in a sealed tube at 37◦C. Before analysis, the sample was lightly
ground in a mortar and pestle and left exposed at 37◦C for 2 h (in
the dark) to allow removal of any free molecular bromine. Sam-
ples then showed 40, 48, 53, 60% conversion to compound B
after 1, 2, 3, and 4 weeks respectively (by NMR analysis). Pow-
der XRD examination of the same solids (without dissolution)
provided similar results (Accessory Publication Fig. S6).
Accessory Publication
Representative NMR spectra of compounds A and B, crystal-
lographic packing diagrams for monoclinic and triclinic com-
pound A, and time dependent XRD speciation diagrams at 25
and 37 degrees are available on the Journal’s website.
Bromine Catalysis
Compound A dissolved in CDCl3 was treated with a trace
of Br2. Immediate NMR analysis showed 93% conversion to
compound B. Compound A was stored in the dark at room
temperature under bromine vapour (but not in direct contact)
for 1 week. Powder XRD analysis of the sample showed 100%
conversion to compound B.
Acknowledgements
R.M.C. thanks Dr James De Voss for providing office and laboratory space
and Dr Joanne Blanchfield and the De Voss research students for helpful
practical assistance.
Crystallography
References
Compound A (monoclinic): C10H16Br4: M 455.87, temperature
293 K, monoclinic, space group P21/n, a 9.5390(7), b 9.9989(7),
c 14.577(1) Å, β 102.920(7)◦, V 1355.2(2) Å3, Z 4, F(000)
864, Dc 2.234 g cm−3, µ 11.843 mm−1, 2374 unique data (Rint
0.0552, 2θmax = 50◦), R1 0.0336 (for 1111 observed reflections
I > 2σ(I)), wR2 0.0497 (all data).
[1] M. J. Gallagher, M.Sc. Thesis: Studies in terpene chemistry 1960,
pp. A1–A4 (Brisbane: The University of Queensland).
[2] R. M. Carman, B. N. Venzke, Aust. J. Chem. 1971, 24, 1905.
[3] R. M. Carman, B. N. Venzke, Aust. J. Chem. 1974, 27, 383.
[4] P. V. Bernhardt, R. M. Carman, T. T. Le, Aust. J. Chem. 2005, 58, 535.
doi:10.1071/CH05043
Compound A (triclinic): C10H16Br4: M 455.87, temperature
293 K, triclinic, space group P1, a 7.7631(7), b 9.6651(8), c
10.207(1) Å, α 100.577(8)◦, β 92.661(9)◦, γ 112.706(8)◦, V
[5] O. Wallach, Liebigs Ann. 1885, 230, 262.
[6] C. Hintze, Z. Kristallogr. 1885, 10, 258.
[7] J. A. Henry, H. Paget, J. Chem. Soc. 1931, 25. doi:10.1039/
JR9310000025
[8] W. R. Owen, M.Sc. Thesis: Studies in the chemistry of alpha-
phellandrene 1955 (Brisbane: The University of Queensland).
[9] M. D. Sutherland, J. W. Wells, J. Org. Chem. 1956, 21, 1272.
doi:10.1021/JO01117A015
[10] J. D. Briasco, J. Murray, J. Prac. Chem. 1952, 2, 187.
[11] J. D. Dunitz, J. Bernstein,Acc. Chem. Res. 1995, 28, 193. doi:10.1021/
AR00052A005
[12] W. C. McCrone, Polymorphism in Physics and Chemistry of the
Organic Solid State (Eds D. Fox, M. M. Labes, A. Weissberger) 1965,
Vol. 11, pp. 726–767 (Interscience: New York, NY).
[13] G. Kaupp, CrystEngComm 2003, 5, 117. doi:10.1039/B303432A
[14] G. Kaupp, Top. Curr. Chem. 2005, 254, 95.
[15] L. J. Farrugia, J. Appl. Cryst. 1999, 32, 837. doi:10.1107/
S0021889899006020
[16] G. M. Sheldrick, SHELX97 – Programs for Crystal StructureAnalysis
(Release 97–2) 1998 (Universität Göttingen: Göttingen, Germany).
[17] L. J. Farrugia, J. Appl. Cryst. 1997, 30, 565. doi:10.1107/
S0021889897003117
688.7(1) Å3, Z 2, F(000) 432, Dc 2.198 g cm−3, µ 11.653 mm−1
,
2345 unique data (Rint 0.0357, 2θmax = 50◦), R 0.0527 (for 1270
observed reflections I > 2σ(I)), wR2 0.1658 (all data).
Single-crystal intensity data were collected on an Oxford
Diffraction Gemini S Ultra CCD diffractometer using graphite
monochromated Mo-Kα radiation (λ 0.71073 Å) operating in
the ω-scan mode. Data reduction and empirical absorption
corrections were performed with the CrysAlis program (Oxford
Diffraction, version 171.33.34d), while all other computations
were performed within the WinGX suite of programs.[15] Struc-
tures were solved by direct methods with SHELXS and refined by
full matrix least-squares analysis with SHELXL97.[16] All non-
H atoms were refined with anisotropic thermal parameters, and
H atoms were constrained at estimated positions using a riding
model. The atomic nomenclature is defined in Figs 1a and 1b,
drawn with ORTEP3.[17] Crystallographic data in CIF format
are available from the Cambridge Crystallographic Data Base
(CCDC) with CCDC Deposition numbers 755597 (monoclinic
compound A) and 755598 (triclinic compound A).
[18] B. E. Warren, J. Am. Ceram. Soc. 1934, 17, 249. doi:10.1111/J.1151-
2916.1934.TB19316.X