Supramolecular metathesis: Co-former exchange in co-crystals of pyrazine with (R,R)-, (S,S)-, (R,S)- and (S,S/R,R)-tartaric acid

Article abstract of DOI:10.1039/c0ce00576b

Co-crystals of dextro-(R,R), levo-(S,S), meso-(R,S) and racemic (R,R-S,S)-tartaric acid with pyrazine were obtained by manual kneading and slurry experiments; subsequent reactions in the solid state between these co-crystals and the various forms of tartaric acid in the solid state and via slurry show that co-former exchange takes place according to the sequence of stability [(R,S)-ta]₂·py > (S,S/R,R)-ta·py > (R,R)-ta·py or (S,S)-ta·py.

Full text of DOI:10.1039/c0ce00576b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
CrystEngComm
Cite this: CrystEngComm, 2011, 13, 3122
www.rsc.org/crystengcomm
COMMUNICATION
Supramolecular metathesis: co-former exchange in co-crystals of pyrazine
with (R,R)-, (S,S)-, (R,S)- and (S,S/R,R)-tartaric acid†
*
Dario Braga, Fabrizia Grepioni and Giulio I. Lampronti
Received 15th June 2010, Accepted 16th September 2010
DOI: 10.1039/c0ce00576b
acid via N/H–O hydrogen bonds, but also because any unreacted
pyrazine can easily sublimate at room temperature as a pure
component,6 thus simplifying analysis and characterization of the
reaction product.
Co-crystals of dextro-(R,R), levo-(S,S), meso-(R,S) and racemic
(R,R–S,S)-tartaric acid with pyrazine were obtained by manual
kneading and slurry experiments; subsequent reactions in the solid
state between these co-crystals and the various forms of tartaric acid
in the solid state and via slurry show that co-former exchange takes
place according to the sequence of stability [(R,S)-ta]2$py > (S,S/
R,R)-ta$py > (R,R)-ta$py or (S,S)-ta$py.
First co-crystals of (R,R)-, (S,S)-, racemic and (R,S)-tartaric acid
and pyrazine, respectively called (R,R)-ta$py, (S,S)-ta$py, (S,S/R,R)-
ta$py and [(R,S)-ta]2$py, were obtained from three different methods
(see Scheme 1): (i) manual kneading of the solid pyrazine and tartaric
acid reactants, (ii) slurry in ethanol and (iii) crystallization from
ethanol solutions.‡ This latter method allowed the growth of single
crystals suitable for X-ray diffraction, thus providing not only
complete characterization of the co-crystals, but also a control of the
products purity that was performed via powder X-ray diffraction
measurements accompanied by Rietveld refinements (see ESI†).x
Thermal stability of all compounds was checked by DSC
measurements: above ca. 130 ꢀC for (S,S/R,R)-ta$py, [(S,S/R,R)-
ta]2$py (discussed below) and [(R,S)-ta]2$py, and ca. 120 ꢀC for
The paradigm of crystal engineering is the possibility of controlling
the preparation of crystal phases of a given molecule by a choice of
supramolecular bonding features. Molecules and ions are regarded as
building blocks in the construction of periodical supramolecular
frameworks.1 The introduction of a new component in a single
molecule crystalline material often allows the existence of materials
with different free energy states, hence to potentially more stable
crystal forms. Co-crystallization has been proved to be a route to the
preparation of entirely new multi-component crystalline systems.2
Clearly, the competition with kinetic factors, associated often to the
nucleation stage of the crystallization process, needs also to be taken
into account.3 A way to overcome the thermodynamic–kinetic
dualism is the ‘‘solvent-free’’ condition.
The number of papers reporting preparation of molecular co-
crystals and salts via solid state techniques is constantly increasing. If
not all, many processes can definitely be conducted in the absence of
solvent or with solvent present in minimal ‘‘catalytic’’ quantity
(kneading and liquid-assisted grinding).2i,4 It has also been shown that
in some systems chiral recognition can affect solid state reactions just
as it does in solution chemistry, and that products of grinding and
solution experiments can be different.5
In the present work we make use of mechanochemical methods to
first obtain, then convert a co-crystal into a different one that shares
a molecular component with the former. For this purpose dextro-
(R,R), levo-(S,S), meso-(R,S) and racemic (R,R–S,S)-tartaric acid,
C4O6H6, and pyrazine, C4N2H4, were chosen as starting materials.
Pyrazine was chosen not only because it would easily bind to tartaric
Scheme 1 Reactivity of the various forms of tartaric acid towards pyr-
azine, both in the solid state and in solution.
Dipartimento di Chimica G. Ciamician, Universitaꢀdi Bologna, Via F. Selmi
2, 40126 Bologna, Italy. E-mail: dario.braga@unibo.it; Fax: +39 051
2099456
† Electronic supplementary information (ESI) available: Experimental,
calculated and difference X-ray powder patterns. CCDC reference
numbers 781833–781836. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c0ce00576b
Fig. 1 O(H)COOH/N hydrogen bonding interactions in (R,R)-ta$py (a)
and in (S,S/R,R)-ta$py (b).
3122 | CrystEngComm, 2011, 13, 3122–3124
This journal is ª The Royal Society of Chemistry 2011

View Article Online
(R,R)-ta$py the co-crystals are no longer stable, and separate out into
the pure reagents; this is followed by sublimation of pyrazine.
X-Ray single crystal determinations were essential for the identi-
fication of the main hydrogen bonding features, responsible for the
behaviour of these systems with respect to co-former exchange in
kneading or slurry experiments.
different forms of tartaric acid or by cross-reacting the preformed co-
crystals. Kneading experiments were all conducted with an excess of
pyrazine, and always led to quantitative co-crystal formation, since
the excess of pyrazine sublimes. A list of experiments and stoichio-
metric ratios are reported in Table 1.
[(R,S)-ta]2$py was obtained from (R,S)-tartaric acid and (R,R)-
ta$py or (S,S/R,R)-ta$py by kneading and slurry experiments.
Interestingly, while (R,R)-ta$py and (S,S)-ta$py react in a slurry
experiment to form (S,S/R,R)-ta$py, i.e. maintaining the stoichio-
metric ratio 1 : 1 between the acid and the base, manual kneading
resulted in the formation of the 2 : 1 stoichiometry product [(S,S/
R,R)-ta]2$py, with sublimation of the unreacted pyrazine. Recrys-
tallization via seedingk of this co-crystal from an ethanol solution
allowed complete structural characterization of the product by single-
crystal X-ray diffraction. As observed in the case of [(R,S)-ta]2$py,
the 2 : 1 stoichiometry allows the formation of carboxylic dimers in
addition to the ubiquitous (CO)OH/N hydrogen bond (see Fig. 3).
This result provides further evidence to the fact that solid state
reactions can rapidly produce pure phases in a solvent-free condi-
tion.9 In this particular case the formation of the 2 : 1 product by
manual kneading is probably due to partial sublimation of pyrazine
during the kneading process; the same reaction conducted with
In (R,R)-ta$py and (S,S)-ta$py the tartaric acid molecules link one
to the other via OOH(H)/OOH/OCO hydrogen bonds, while the
carboxylic OH group is employed only in the OH/N bonds with
pyrazine (see Fig. 1a).
A similar arrangement, as far as the carboxylic groups are con-
cerned, is observed in (S,S/R,R)-ta$py (see Fig. 1b).
The pattern is completely changed, though, in [(R,S)-ta]2$py
(Fig. 2), where the 2 : 1 acid : pyrazine stoichiometry, in a centro-
symmetric space group, allows the formation of carboxylic dimeric
rings (evidenced in yellow) in addition to the favourite (CO)OH/N
hydrogen bonds. Structural effects of chiral versus racemic tartaric
acid in the synthesis of adducts with diamines have also been inves-
tigated by Aakeroy et al.7 and by Glidewell et al.8
Our co-crystals were then used as starting materials for slurry and
kneading experiments,{ in the attempt of converting one phase into
another by further reaction of pre-formed co-crystals with the
Fig. 2 O(H)COOH/OCOOH carboxylic dimeric rings are present in addition to O(H)COOH/N hydrogen bonding interactions in [(R,S)-ta]2$py.
Table 1 Reagents stoichiometric ratio and products of kneading and slurry experiments
Reagents
Products (kneading)
Products (slurry)
Co-crystala + tartaric acid
Molar ratio
Co-crystal + tartaric acid
Co-crystal + tartaric acid
(R,R)-ta$py + (R,S)-ta
1 : 2
1 : 2
2 : 1
2 : 1
2 : 1
[(R,S)-ta]2$py + (R,R)-ta
[(R,S)-ta]2$py + (S,S/R,R)-ta
No reaction
No reaction
No reaction
[(R,S)-ta]2$py + (R,R)-ta
[(R,S)-ta]2$py + (S,S/R,R)-ta
No reaction
No reaction
No reaction
(S,S/R,R)-ta$py + (R,S)-ta
[(R,S)-ta]2$py + (R,R)-ta
[(R,S)-ta]2$py + (S,S/R,R)-ta
(S,S/R,R)-ta$py + (R,R)-ta
Co-crystala + co-crystal
Molar ratio
Co-crystal
Co-crystal
(R,R)-ta$py + (S,S)-ta$py
(R,R)-ta$py + (S,S)-ta$py + py
1 : 1
1 : 1 : 10
[(S,S/R,R)-ta]2$pyb
(S,S/R,R)-ta$pyb
(S,S/R,R)-ta$pyb
(S,S/R,R)-ta$pyb
a
b
See the products in Scheme 1. The 2 : 1 co-crystal could only be obtained via kneading, probably due to loss of volatile pyrazine during the
experiment; in order to obtain the 1 : 1 co-crystal, an excess of pyrazine was used. Slurry experiments invariably yielded the 1 : 1 co-crystal.
Fig. 3 The acid : pyrazine 2 : 1 stoichiometry in [(S,S/R,R)-ta]2$py allows the presence of O(H)COOH/OCOOH carboxylic dimeric rings in addition to
O(H)COOH/N hydrogen bonding interactions, contrary to what observed for (S,S/R,R)-ta$py (Fig. 1b).
This journal is ª The Royal Society of Chemistry 2011
CrystEngComm, 2011, 13, 3122–3124 | 3123

View Article Online
conducted starting from the crystal structures refined from single crystal
data and treating the single molecules as rigid bodies. Shifted Chebyshev
function with 6 parameters and Pseudo-Voigt function were used to fit
respectively background and peak shape. An overall thermal parameter
for each molecule was adopted. For (S,S)-ta$py data crystal structure
was derived from (R,R)-ta$py co-crystal. Refinements of kneading
experiment products for phases (R,R)-ta$py, (S,S)-ta$py, (S,S/R,R)-
ta$py, [(S,S/R,R)-ta]2$py and [(R,S)-ta]2$py converged with Rwp respec-
tively 14.80%, 11.68%, 11.58%, 14.62% and 9.91%, and RF respectively
11.69%, 12.17%, 11.03%, 13.12%, and 10.99%. DSC measurements were
performed with a Perkin-Elmer Diamond. Samples (3–5 mg) were placed
in open aluminium pans. Heating was carried out at 5 ꢀC minꢁ1 for all co-
crystals and 1 ꢀC minꢁ1 for (S,S/R,R)-ta$py co-crystal, in the temperature
range 25 to 160 ꢀC.
a large excess of pyrazine (10 : 1, see Table 1) results in fact in the 1 : 1
co-crystal, as observed by slurry.
We assume that the products of slurry experiments represent the
most stable forms, since the experimental time is much longer than
manual grinding experiments (days vs. minutes). In terms of
molecular recognition we may say that each pyrazine molecule has
a chance to eventually pass from one crystal structure to a more
stable one.
Conclusions
We have shown that solid–solid reactions with co-crystals or between
co-crystals can be used not only to produce new crystal forms with
respect to conventional reactions in solution, but also to interconvert
crystal forms, in a sort of supramolecular metathesis. The combined
experiments suggest a scale of solid state stability [(R,S)-ta]2$py >
(S,S/R,R)-ta$py > [(S,S/R,R)-ta]2$py > (R,R)-ta$py or (S,S)-ta$py.
The fact that pyrazine is volatile is instrumental to the preparation of
pure phases because of the sublimation of the excess reactant.
{ Metathesis experiments. Kneading experiments. 1 mmol of tested co-
crystal and tested isomeric tartaric acid in stoichiometric quantity (see
Table 1) were manually ground for 30 minutes after adding three to five
drops of ethanol corresponding to a few ml. 1 mmol of (R,R)-ta$py and 1
mmol of (S,S)-ta$py co-crystals with or without 10 mmol of pyrazine (see
Table 1) were manually ground for 30 minutes after adding three to five
drops of ethanol corresponding to a few ml. The powders were left at
room temperature for at least 24 hours before XRD characterization.
Slurry experiments. 1 mmol of tested co-crystal and tested isomeric tar-
taric acid in stoichiometric quantity (see Table 1) were suspended in
around 15 ml of ethanol in a closed vessel, and stirred at room temper-
ature for over two weeks. 1 mmol of (R,R)-ta$py and 1 mmol of (S,S)-
ta$py co-crystals with or without 10 mmol of pyrazine (see Table 1) were
suspended in around 15 ml of ethanol in a closed vessel, and stirred at
room temperature for over two weeks.
Notes and references
‡ Co-crystals preparation. Crystallization from solution. (R,R)-ta$py,
(S,S)-ta$py, (S,S/R,R)-ta$py and [(R,S)-ta]2$py co-crystals were grown
by crystallization from methanol solution. 1 mmol of tartaric acid (150
mg) and 3 mmol of pyrazine (240 mg) were separately dissolved in
methanol and then mixed. The solution was allowed to evaporate at room
temperature. Kneading experiments. (R,R)-ta$py, (S,S)-ta$py, (S,S/R,R)-
ta$py and [(R,S)-ta]2$py co-crystals were obtained by kneading experi-
ments using ethanol. 1 mmol of tartaric acid (150 mg) and 10 mmol of
pyrazine (800 mg) were manually ground for 30 minutes after adding
three to five drops of solvent. The powders were left at room temperature
for at least 24 hours before XRPD characterization. Slurry experiments.
(R,R)-ta$py, (S,S)-ta$py, (S,S/R,R)-ta$py and [(R,S)-ta]2$py co-crystals
were obtained by slurry experiments using ethanol. 1 mmol of tartaric
acid (150 mg) and 1 mmol of pyrazine (80 mg) were suspended in 15 ml of
ethanol in a closed vessel, and stirred at room temperature for two weeks.
x X-Ray diffraction. X-Ray data collected with an Oxford Diffraction
k Seeding experiment. 1 mmol of (S,S/R,R)-tartaric acid (150 mg) and 3
mmol of pyrazine (240 mg) were separately dissolved in methanol solu-
tion and then mixed. 100 mg of [(S,S/R,R)-ta]2$py were suspended in the
resulting solution. The solution was left to evaporate at room tempera-
ture.
1 D. Braga and F. Grepioni, Chem. Commun., 2005, 3635.
2 (a) D. Braga, S. L. Giaffreda, F. Grepioni, G. Palladino and
M. Polito, New J. Chem., 2008, 32, 820; (b) D. Braga,
S. L. Giaffreda, F. Grepioni, G. Palladino and M. Polito, New J.
Chem., 2008, 32, 820; (c) D. Braga, F. Grepioni, M. Polito,
M. R. Chierotti, S. Ellena and R. Gobetto, Organometallics, 2006,
25, 4627; (d) A. D. Bond, CrystEngComm, 2007, 9, 833; (e)
G. R. Desiraju, CrystEngComm, 2003, 5, 466; (f) C. B. Aakeroy and
D. J. Salmon, CrystEngComm, 2005, 7, 439; (g) O. Almarsson and
M. J. Zaworotko, Chem. Commun., 2004, 1889; (h) S. L. Childs and
M. J. Zaworotko, Cryst. Growth Des., 2009, 9, 4208; (i) S. Karki,
ꢁ
Xcalibur diffractometer; Mo Ka radiation (l ¼ 0.71073 A). Crystal data.
(R,R)-ta$py: chemical formula moiety C4H6O6$C4H4N2, M ¼ 230.19,
T ¼ 293 ꢀC, triclinic, space group P1, a ¼ 4.9179(5), b ¼ 5.4897(7), c ¼
ꢀ
ꢁ
9.5322(7)A, a ¼ 92.426(9), b ¼ 102.087(9), g ¼ 94.810(9) , V ¼ 250.28(5)
ꢃꢃ ꢄ
T. Friscic and W. Jones, CrystEngComm, 2009, 11, 470.
3
ꢁ
A , Z ¼ 1, 1460 independent reflections (1729 measured), Rint ¼ 0.0168,
wR2 ¼ 0.0737, R1(obs) ¼ 0.0375. [(R,S)-ta]2$py: chemical formula moiety
ꢄ
3 B. Rodrıguez-Spong, C. P. Price, A. Jayasankar, A. J. Matzger and
ꢄ
N. Rodrıguez-Hornedo, Adv. Drug Delivery Rev., 2004, 56, 241.
ꢀ
ꢂ
C4H6O6$0.5(C4H4N2), M ¼ 380.29, T ¼ 293 C, triclinic, space group P1,
4 D. Braga, S. L. Gaffreda, K. Rubini, F. Grepioni, M. R. Chierotti and
R. Gobetto, CrystEngComm, 2007, 9, 39.
5 R. Kuroda, Y. Imai and T. Sato, Chirality, 2001, 13, 588; R. Kuroda,
J. Yoshida, A. Nakamura and S. Nishikiori, CrystEngComm, 2009,
11, 427; S. Chen, H. Xi, R. F. Henry, I. Marsden and
G. G. Z. Zhang, CrystEngComm, 2010, 12, 1485.
6 D. Lipkind and J. S. Chickos, Struct. Chem., 2009, 20, 49.
7 C. B. Aakeroy, P. B. Hitchcock and K. R. Seddon, J. Chem. Soc.,
Chem. Commun., 1992, 3635.
8 D. M. M. Farrell, G. Ferguson, J. Lough and C. Glidewell, Acta
Crystallogr., Sect. B: Struct. Sci., 2002, 58, 272.
ꢁ
a ¼ 4.9819(5), b ¼ 5.3114(7), c ¼ 14.748(2)A, a ¼ 89.543(12), b ¼
ꢀ
3
ꢁ
86.283(10), g ¼ 84.973(10) , V ¼ 387.91(9) A , Z ¼ 1, 1726 independent
reflections (2910 measured), Rint ¼ 0.0178, wR2 ¼ 0.1031, R1(obs) ¼
0.0419. (S,S/R,R)-ta$py: chemical formula moiety C4H6O6$C4H4N2,
M ¼ 230.19, T ¼ 293 ꢀC, monoclinic, space group P21/n, a ¼ 11.4966(4),
ꢀ
ꢁ
b ¼ 5.1521(2), c ¼ 17.0937(6) A, a ¼ 90, b ¼ 96.315(3), g ¼ 90 , V ¼
3
ꢁ
1006.34(6) A , Z ¼ 4, 2258 independent reflections (4566 measured),
Rint ¼ 0.0207, wR2 ¼ 0.0952, R1(obs) ¼ 0.0491. [(S,S/R,R)-ta]2$py:
chemical formula moiety C4H6O6$0.5(C4H4N2), M ¼ 380.29, T ¼ 293 ꢀC,
ꢂ
ꢁ
triclinic, space group P1, a ¼ 4.9168(5), b ¼ 5.4373(5), c ¼ 14.7869(14) A,
ꢀ
3
ꢁ
a ¼ 81.199(8), b ¼ 83.738(8), g ¼ 88.135 , V ¼ 388.28(6) A , Z ¼ 1, 1726
independent reflections (2909 measured), Rint ¼ 0.0222, wR2 ¼ 0.1651,
R1(obs) ¼ 0.0568. SHELX9710a and SCHAKAL9910b were used for
structure solution and graphical representation. Powder diffraction
patterns over 5ꢀ to 90ꢀ in 2q were collected on a PANalytical diffrac-
tometer with Bragg–Brentano geometry (Cu Ka radiation, detector:
X’celerator, step size D2q ¼ 0.0167ꢀ, and counting time per step ¼ 50 s).
The software GSAS10c was used for refinements. Rietveld analyses were
9 D. Braga, F. Grepioni and L. Maini, Chem. Commun., 2010, 46,
6232.
10 (a) G. M. Scheldrick, SHELXL97: Program for Crystal Structure
€
Determination, University of Gottingen, Germany, 1997; (b)
E. Keller, SCHAKAL99: Graphical Representation of Molecular
Models, University of Freiburg, Germany, 1999; (c) A. C. Larson
and R. B. Von Dreele, General Structure Analysis System (GSAS),
Los Alamos National Laboratory Report LAUR 86–748, 2000.
3124 | CrystEngComm, 2011, 13, 3122–3124
This journal is ª The Royal Society of Chemistry 2011