New polymorphic hydrogen bonding donor–acceptor system with two temperature coincident solid–solid transitions

Abstract

The polymorphism study of dibenzylsquaramide reveals the appearance of two forms sharing a solid–solid transition at the same temperature towards the highest melting polymorph.

---

Cyclobutenediones exhibit interesting properties due to their high ring tension and have attracted the attention of different areas such as medicinal chemistry, enantioselective catalysis and supramolecular chemistry. In particular, the diamidoderivatives (also known as squaramides) have been studied as pharmaceutically active compounds,¹ chiral auxiliaries,² chemosensors³ and in molecular recognition systems.⁴ Previous studies of secondary squaramides in solution demonstrate that these compounds can exist in several conformations due to the partially restricted rotation of the C–N bond (anti/anti and anti/syn conformers) (Fig. 1).⁵ However, to the best of our knowledge nothing is known about this conformational preference in the solid state and only a few crystal structures of disecondary squaramides have been described.⁶

Fig. 1 Different conformations of dibenzylsquaramide (DBZSQ), i.e., 3,4-bis-benzylamino-cyclobut-3-ene-1,2-dione.

It is well known that these molecules exhibit a dual donor–acceptor hydrogen bonding ability. This has motivated us to study the solid state behaviour of secondary squaramides in order to establish relations between the aforementioned properties and their possible polymorphism.

Polymorphism is defined as the ability of the same compound to crystallize in different crystal forms. The differences in crystal packing often lead to considerable differences in solubility, hygroscopicity, bioavailability and physical stability. Therefore, crystal engineering, namely the design of crystal structures starting from molecular building blocks, is an extremely useful approach to both academic and pharmaceutical research as it permits a rational approach to the investigation of crystal forms (solvates, salts and co-crystals) and their polymorphs.⁷

In this study we have chosen dibenzylsquaramide (DBZSQ) as the model compound to explore its possible polymorphic behaviour. DBZSQ was obtained according to the literature from diethylsquarate and benzylamine in ethanol as a chemically pure white solid.⁸ In order to obtain as many crystal forms as possible a polymorphic screening was carried out. DBZSQ is only soluble in polar media such as DMF and DMSO. Therefore different combinations of those solvents with polar and non polar antisolvents were tested at several concentrations and temperatures, with variable cooling rates, in both thermodynamic and kinetic conditions, revealing a polymorphic system consisting of three polymorphs according to their different X-ray patterns (Fig. 2). XRPD were collected in a capillary spinner in order to reduce if not to eliminate problems of preferential orientation. Form A was obtained by slow cooling of a DMF saturated DBZSQ solution at 70 °C. Forms B and C could be obtained by slow diffusion of diethyl ether or dioxane, respectively, into a DMSO saturated DBZSQ solution at room temperature. For A and C forms, single crystals suitable for X-ray structure determination were grown in the former conditions. Each different form obtained was characterized by means of DSC, XRPD, IR and TGA. IR spectroscopy did not allow us to distinguish among the three forms. Thermogravimetric analysis confirmed that neither a solvate nor a hydrate was obtained.

Fig. 2 XRPD patterns of Forms A, B and C of DBZSQ.

The DSC analysis⁹ of the three forms show the same sharp endothermic phenomenon at 311 °C but two of them also show a low intensity broad endothermic phenomenon at 257 °C (Fig. 3). Variable heating rates confirm that the first phenomenon is a solid–solid transition, with the second one being a melting process. This can be interpreted in the following manner: two forms (B and C) transform during the DSC analysis in the same polymorph (A) which melts at 311 °C.

Fig. 3 DSC curves of Forms A, B and C of DBZSQ.

Curiously, the transition temperature is observed around the same value in all the cases.¹⁰ Although several polymorphs can present the same melting value (e.g. conformational polymorphism),¹¹ to the best of our knowledge this is the first time that two forms share the same transition temperature. Therefore, a misleading interpretation of this system could be concluded if based only on the DSC analysis.

It is of great relevance to know whether polymorphic modifications can transform reversibly (enantiotropy) or irreversibly (monotropy) at atmospheric pressure. Should an organic compound exhibit polymorphism of an enantiotropic type, the knowledge of the different domains of thermodynamic stability for each form is essential in order to obtain the desired form through a reliable crystallization procedure and to define the optimum storage conditions.¹² According to the heat of transition rule¹³ forms B and C are enantiotropically related to form A. This means that form A is a metastable form at room temperature.

In addition, the reversibility of the solid transition could be observed in both cases by a heating–cooling DSC experiment (Fig. 4). Both forms B and C return to form B after cooling (confirmed by XRPD).

Fig. 4 Heating-cooling DSC experiment, where the reversibility of the solid transition of Form B of DBZSQ can be observed.

The crystal structure of metastable form A was determined by single crystal X-ray diffraction, revealing a well defined head-to-tail H-bonding motif among the squaramide units (Fig. 5) (N–H⋯O 2.834(6) Å). Form A crystallizes in monoclinic non-centrosymmetric space groupC2.¹⁴

Fig. 5 Single crystal X-ray structure of Form A.

The crystal structure of form C was determined by single crystal X-ray diffraction, revealing again the same well defined head-to-tail H-bonding motif (Fig. 6) (N–H⋯O 2.779(3) Å). Form C crystallizes in monoclinic space groupC2/c.¹⁴

Fig. 6 Single crystal X-ray structure of Form C.

It is important to notice the different packing direction of the squaramide columns when comparing both structures. Whereas form A shows a packing in antiparallel chains but parallel planes, form C exhibits parallel chains and antiparallel planes arrangements. Secondary aromatic interactions such as CH–π can also be observed.

These data confirm the favourable anti/anti conformation, previously seen in solution.⁴ Its self-aggregation explains the low solubility of this compound in non hydrogen bonding competitive media and it is probably responsible for the non-existence of the anti/syn conformation in the solid state.

Unfortunately, it has not been possible to grow single crystals of form B suitable for X-ray structure determination; however an indexing from the powder pattern of form B was attempted using DICVOL,¹⁵ that returned an orthorhombic unit cell with lattice parameters: a = 29.58(6) b = 17.22(2) c = 5.67(9) Å, V = 2887 ų. The volume suggests the presence of eight molecules within the unit cell. The Pawley fit¹⁶ in space groupPcm2₁ gave a reasonable fit to the data, with a reduced χ2 of 5.25. Moreover, the volume found for form B, is consistent with the one obtained for polymorph A: a = 11.403(2) b = 4.4101(10) c = 30.270(6) Å β = 100.807(5) °, V = 1495 ų, space groupC2 and polymorph C a = 30.4956(18), b = 6.0254(3), c = 8.1984(5) Å, β = 96.760(6)°, V = 1495.97(15) ų. In spite of all efforts using DASH program we were unable to determine the structure of this form.

In conclusion, we have reported the polymorphism of dibenzylsquaramide. In this preliminary study we have observed the appearance of two forms sharing a solid–solid transition towards the highest melting polymorph at the same temperature, which is a rare observation. Therefore, a misleading interpretation of a system like this could be done if based only on the DSC analysis. Although thermal analysis is essential for studying the polymorphic behaviour of crystal solids, our results can be a useful warning to scientists involved in pharmaceutical research about the limitation of these techniques when characterizing crystal solids. The two crystal structures obtained, both showing a head-to-tail pattern, suggest that the H-bonding donor–acceptor duality of secondary squaramides can be used as a synthon in crystal engineering in the future.

The authors thank Dr X. Alcobé (XRD Unit of the Scientific-Technical Services, University of Barcelona) for the XRPD measurements and for technical guidance. We also thank MiUR Italy (project POLYM2006) for financial support (DB, MP) and gratefully acknowledge the input received from Dr Lucia Maini (Università di Bologna). We are indebted to Prof. Antoni Costa (Universitat de les Illes Balears) for the generous gift of a sample of dibenzylsquaramide.

References

1. B. Bang-Andersen, H. Ahmadian, S. M. Lenz, T. B. Stensbol, U. Madsen, K. P. Bogeso and P. Krogsgaard-Larsen, J. Med. Chem., 2000, 43, 4910–4918.
2. J. Zhang, H. B. Zhou, S. M. Lümmluo, R. G. Xie, M. C. K. Choi, Z. Y. Zhou, S. C. Chang and T. K. Yang, Tetrahedron: Asymmetry, 2001, 12, 1907–1912.
3. S. Tomas, R. Prohens, G. Deslongchamps, P. Ballester and A. Costa, Angew. Chem., Int. Ed., 1999, 38, 2208–2211; R. Prohens, G. Martorell, P. Ballester and A. Costa, Chem. Commun., 2001, 16, 1456–1457.
4. S. Tomas, R. Prohens, M. Vega, M. C. Rotger, P. M. Deyà, P. Ballester and A. Costa, J. Org. Chem., 1996, 61, 9394–9401.
5. M. C. Rotger, N. Piña, A. Frontera, G. Martorell, P. Ballester, P. M. Deyà and A. Costa, J. Org. Chem., 2004, 69, 2302–2308.
6. C. Rotger, B. Soberats, D. Quiñonero, A. Frontera, P. Ballester, J. Benet-Buchholz, P. M. Deyà and A. Costa, Eur. J. Org. Chem., 2008, 1864–1868; A. P. Davis, S. M. Draper, G. Dunne and P. Ashton, Chem. Commun., 1999, 2265–2266.
7. G. R. Desiraju, Angew. Chem., Int. Ed., 2007, 46, 8342–8356; D. Braga and F. Grepioni, Making crystals by design, Wiley-VCH: Weinheim, 2006; J. Bernstein, Polymorphism in Molecular Crystals, Clarendon Press, Oxford, 2002; R. D. Braga and F. Grepioni, Chem. Commun., 2005, 3635; J. Bernstein, Chem. Commun., 2005, 5007.
8. R. Prohens, S. Tomas, J. Morey, P. M. Deyà, P. Ballester and A. Costa, Tetrahedron Lett., 1998, 39, 1063–1066.
9. DSC analysis were carried out at a heating rate of 10 °C min⁻¹ and atmosphere of dry nitrogen.
10. Different samples of each polymorph analysed by DSC at a rate of 10 °C min⁻¹, show the solid–solid transition in the temperature range of 256–259 °C.
11. S. Roy, S. Aitipamula and A. Nangia, Cryst. Growth Des., 2005, 5, 2268–2276.
12. D. Giron, M. Mutz and S. Garnier, J. Therm. Anal. Calorim., 2004, 77, 709.
13. A. Burger and R. Ramberger, Mikrochim. Acta, Part I, 1979, 259 . This rule states that if an endothermic transition is observed at some temperature it may be assumed that there is a transition point below it and that the two forms are related enantiotropically.
14. Crystal-structure determination: data for form A and form C were collected at room temperature on Apex II Bruxer AXS diffractometer, Mo Kα radiation (λ = 0.71073 Å) C₁₈H₁₆N₂O₂M_r = 292.20, colourless crystals. Crystal data for form A: crystal dimensions: 0.22 × 0.19 × 0.11 mm³; monoclinic C2, a = 11.403(2), b = 4.4101(10), c = 30.270(6) Å, β = 100.807(5)°, V = 1495.2(5) ų, Z = 4, ρ_calcd = 1.298 g cm⁻³, R_int = 0.0574, R1 = 0.0623, wR2 = 0.1776 for all 1803 observed independent reflections; F(000) = 616: 2θ range = 1.36 – 53.46°. CCDC 676071. Crystal data for form C: only half molecule is present in the asymmetric unit, since the molecule has crystallographically imposed twofold symmetry; crystal dimensions: 0.20 × 0.18 × 0.12 mm³; monoclinic C2/c, a = 30.4956(18), b = 6.0254(3), c = 8.1984(5) Å, β = 96.760(6)°, V = 1495.97(15) ų, Z = 4, ρ_calcd = 1.298 g cm⁻³, R_int = 0.0536, R1 = 0.0760, wR2 = 0.1971 for all 1775 observed independent reflections; F(000) = 616; 2θ range = 5.38–58.28°. CCDC 686074. All non-hydrogen atoms were refined anisotropically. SHELXL97 was used for structure solution and refinement on F².
15. A. Boultif and D. Louër, J. Appl. Crystallogr., 1991, 24, 987.
16. G. S. Pawley, J. Appl. Crystallogr., 1981, 14, 357.

---

Footnote

† Electronic supplementary information (ESI) available: A table with the most important two theta positions (2θ/°) and relative intensities (I) of the powder X-ray diffraction patterns of DBZSQ crystal forms. Variable heating rates DSC experiments. CCDC reference numbers 676071 and 686074. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/b813086h

---