Struct Chem (2010) 21:185–190 DOI 10.1007/s11224-009-9560-7
ORIGINAL RESEARCH
Structural studies on lithocholyl-N-(2-aminoethyl)amide in the solid state Kari Ahonen • Babita Behera • Elina Sieva¨nen • Arto Valkonen • Manu Lahtinen • Minna Tolonen Reijo Kauppinen • Erkki Kolehmainen
•
Received: 23 October 2009 / Accepted: 18 November 2009 / Published online: 2 December 2009 Springer Science+Business Media, LLC 2009
Abstract The synthetic procedure of lithocholyl-N(2-aminoethyl)amide yielded a mixture of several forms detected by solid state 13C CP/MAS NMR although the solution state NMR unambiguously ascertained that the compound was pure. By recrystallization from various solvents one pure polymorph alongside with four solvates were isolated. The structures of the pure polymorph and the solvates were characterized by 13C and 15N CP/MAS NMR and powder X-ray diffraction (PXRD) methods. Variable contact time and dipolar dephasing experiments were employed to obtain optimized CP parameters and to distinguish various CHn (n = 0–3) resonances. CSA analyses of spinning side bands at different spinning rates showed small variations in the shielding tensor values of the carbonyl group between the pure polymorph (recrystallized from acetonitrile, tetrahydrofuran and 1,4-dioxane) and p-xylene solvate. Keywords 13C and 15N CP/MAS PXRD Polymorph Solvate Lithocholyl amide
Electronic supplementary material The online version of this article (doi:10.1007/s11224-009-9560-7) contains supplementary material, which is available to authorized users. K. Ahonen B. Behera E. Sieva¨nen A. Valkonen M. Lahtinen M. Tolonen R. Kauppinen E. Kolehmainen (&) Department of Chemistry, University of Jyva¨skyla¨, P.O. Box 35, 40014 Jyva¨skyla¨, Finland e-mail:
[email protected]
Introduction Polymorphism is an important phenomenon in solid state chemistry, which commonly arises from two reasons (i) different packing of molecules in the crystal lattice, called packing polymorphism, or (ii) packing of different conformers of the same molecule in an asymmetric unit, called conformational polymorphism [1]. In addition, different crystal structure types may be hydrates, solvates, and co-crystals [2–5]. The ability to detect the presence of different polymorphs in a sample, assay their relative amounts, and ultimately obtain the desired form in high purity, are key issues in solid state organic chemistry. Nowadays polymorph characterization is an active field of research in pharmaceutical and biomedical chemistry as physical and chemical properties of polymorphs are dependent on their structures [2, 3]. Bile acids and their derivatives have been extensively used in supramolecular chemistry, materials science, and nanotechnology [6]. Bile acids, being end products of cholesterol metabolism, are excellent building blocks in tailoring biologically interesting derivatives. Various pharmacological applications of bile acids and their derivatives have been reported in the literature [7–13]. For instance, N-substituted amides of lithocholic acid and derivatives of 3a-hydroxy-24-amino-5b-cholane have shown in vitro activity against Gram positive strains and mycetes [14]. Bile acids and steroids exhibit variable crystal structures, solvates, and co-crystals [15–22]. Although structural studies of systems involving bile acid derivatives have long been a challenge due to their complex and supramolecular nature, papers dealing with single crystal X-ray diffraction characterization of crystalline host–guest assemblies of steroidal and related molecules have been published. The review articles by Miyata et al.
123
186
[17, 18] for example, address the diverse structures of bile acid derivatives with the concept of hierarchy and supramolecular chirality. Single crystal X-ray diffraction is an established method in defining solid state structures of molecules [23], but preparation of single crystals suitable for single crystal X-ray crystallography of bile acid derivatives can be a rather difficult task. Structure solution based on theoretical calculations, powder X-ray diffraction (PXRD) methods, and/or solid state (SS) cross polarization (CP) magic angle spinning (MAS) NMR experiments has been demonstrated to be a very promising alternative [24–39], opening opportunities for investigation of solid state structures even in cases, where no suitable single crystals for single crystal X-ray crystallography are obtained. Especially in polymorph and solvate characterization, SS NMR techniques have shown that 13C NMR chemical shifts are influenced by crystal packing [25–31]. The isotropic chemical shift data as well as determination of the shielding tensors have been successfully used to distinguish various crystallographic forms [32, 33]. However, spectral broadening arising from different solid state interactions and overlapping of resonances due to similar electronic environments in the different polymorphs impose structural analysis. Thus, careful line shape analysis by the spectral deconvolution is essential in SS NMR spectroscopy for obtaining maximum information about the system under investigation.
Experimental Lithocholyl-N-(2-aminoethyl)amide 1 (Scheme 1) was prepared by the procedure given in the literature [40–42]. Its purity was checked by TLC, liquid state 1H, and 13C NMR spectroscopy (Supplementary Material: Fig. S1), and ESI–TOF–MS. In contrast, the 13C CP/MAS NMR spectrum of the synthesis product 1 showed several resonances for each carbon when deconvoluted using DMFIT program [43]. Polymorph 2 was separated from 1 by slowly dissolving 1 in boiling solvent (acetonitrile, tetrahydrofuran, or 1,4-dioxane) and by stirring for 6 h. The resulting solution was filtered and kept at room temperature. Polymorph 2 crystallized out, was separated by filtration, and dried. Solvates 3, 4, and 5 were prepared by the same procedure with recrystallization of 1 from p-xylene, iso-propyl alcohol, and diethyl ether, respectively. Chlorobenzene solvate 6 was gel-like material and it was separated in RT by solvent evaporation and dried in vacuum. All solvents used were of analytical grade and used without further purification. The 13C CP/MAS NMR spectra were recorded on a Bruker Avance 400 NMR spectrometer equipped with a 4 mm standard bore CP/MAS probehead, whose X channel
123
Struct Chem (2010) 21:185–190
Scheme 1 Structure and numbering of lithocholyl-N-(2-aminoethyl)amide
was tuned to 100.62 MHz for 13C and 40.55 MHz for 15N. The other channel was tuned to 400.13 MHz for broad band 1H decoupling. Approximately 100 mg of thoroughly dried and finely powdered sample was packed into the ZrO2 rotor, which was closed with a Kel-F cap. The spinning rate was 5 or 10 kHz unless otherwise stated. The 13C CP/MAS NMR experiment for all samples was carried out under Hartmann-Hahn conditions with TPPM decoupling. The p/2 pulses for proton and carbon were found to be 4.2 and 5 ls at power levels of -4.6 and -0.8 dB, respectively. The optimized contact time of 2 ms was used to obtain efficient polarization transfer. A total of 6000 scans (unless otherwise stated) were recorded with a 4 s recycle delay for each sample. All FIDs were processed by exponential apodization function with line broadening of 20–40 Hz prior to FT. The spectra were deconvoluted using DMFIT software [43]. Originally, the 13C CP/MAS NMR spectra were referenced to glycine, after which the chemical shifts were recalculated to tetramethylsilane d(13C) = 0.0 ppm. In order to distinguish various CHn (n = 0–3) carbons in polymorph 2, three experiments viz. using a short contact time CP (SCP), short cross-polarization polarization inversion (SCPPI), and long cross-polarization depolarization (LCPD) were performed [44] The contact time for LCPD experiments was 2 ms, and that for SCP and SCPPI 40 ls. The depolarization time in the LCPD experiment was 100 ls and the polarization inversion time in the SCPPI experiment 35 ls. The number of scans for each experiment with a recycle delay of 4 s was 2000 (these 13C CP/MAS NMR spectra are depicted in Supplementary Material: Fig. S2). The 15N CP/MAS NMR experiments were carried out for all samples at a 10 kHz spinning rate under HartmannHahn condition. The p/2 pulses for proton and nitrogen were found to be 4.2 and 5 ls at power levels of -4.6 and -0.8 dB, respectively. The optimized contact time of 4 ms was used for efficient polarization transfer with a 5 s recycle delay to acquire the CP/MAS spectra. A total of 16000 scans were acquired overnight to obtain the 15N
Struct Chem (2010) 21:185–190
CP/MAS NMR spectra for each sample. All FIDs were processed by exponential apodization function with line broadening of 20–40 Hz prior to FT. Originally, the 15N CP/MAS NMR spectra were referenced to glycine, after which the chemical shifts were recalculated to nitromethane d(15N) = 0.0 ppm. The X-ray powder diffraction data was measured with PANalytical X0 Pert PRO diffractometer in Bragg–Brentano geometry using Johansson monochromator (a1 setup) to ˚ ; 40 kV, 30 mA) produce pure CuKa1 radiation (1.5406 A and x/2h step-scan technique in 2h range of 3–70. The data collection was made in continuous scanning mode using step size of 0.0167 with a counting time of 60 s at each step. Programmable divergence slit (PDS) was used to set irradiated length on sample to 15 mm together with a 15 mm incident beam mask. Soller slits of 0.02 rad were used on both incident and diffracted beam sides together with anti-scatter slits of 4 and 13 mm, respectively. The diffraction data was collected using X0 Celerator detector. The powder diffraction data was examined by PANalytical Highscore Plus v. 2.2c software package.
Results and discussion To the best of our knowledge, only a few studies on polymorphism of bile acid derivatives by SS NMR have been published [19–21]. The distinct resonances are clearly visible in cases of methyl groups CH3-18 and CH3-21, which do not overlap with the other resonances. The liquid state 13C NMR spectra of 1, however, gave only single resonances for each methyl carbon, revealing that the synthesis products are pure compounds (see Supplementary Material: Fig. S1). The number of peaks observed in the solid state can be explained in two ways. Either each 13C isotropic signal represents an individual polymorph, or one pure polymorph has more than one molecule in the crystallographic asymmetric unit, which results in several resonance lines. Thus, it can be concluded that multiple signals arise from different polymorphic forms, i.e., different packing of molecules, or from several molecules in the crystallographic asymmetric unit. The 13C CP/MAS NMR spectra of pure polymorph 2 and solvates 3–6 are depicted in Fig. 1. The resonances in solid state spectra were assigned tentatively based on the assignments of the liquid state spectra [40–42] and various 13 C CP/MAS techniques (Supplementary Material: Fig. S2). The 13C isotropic shifts of polymorph 2 are listed in Table 1. The 13C resonances in solution spectra are somewhat deshielded compared with those recorded by CP/MAS, which can be understood by crystal packing in the solid state.
187
Solid state CHn (n = 0–3) carbon resonance assignments of 2 (Table 1) were confirmed by dipolar dephasing experiments in conjunction with CP from (i) short crosspolarization (SCP), (ii) short cross-polarization polarization inversion (SCPPI), and (iii) long cross-polarization depolarization (LCPD) dephasing experiments [44], (Supplementary Material: Fig. S2). Since the cross-polarization time constants for the non-protonated carbons and methyl carbons are longer (generally hundreds of ls) than those for CH and CH2 carbons (usually tens of ls), the CH and CH2 magnetizations are almost completely dephased during the depolarization period in the LCPD experiment with a well chosen depolarization time (100 ls). As a result, a single LCPD experiment with a suitable depolarization time (100 ls) filters the CH and CH2 resonances out leading to a spectrum with signals of quaternary and methyl carbons only. In the SCP experiment, on the other hand, the short contact time (40 ls) is not sufficient enough to remove the methyl and the non-protonated carbon signals resulting in weakening of the CH3 signals and almost complete disappearance of the signals originating from the quaternary carbons. When the SCP experiment is performed with 35 ls polarization inversion, all but CH2 resonances, which appear in the inverted phase, are eliminated. The quaternary carbons C-10, C-13, and C-24 resonate at the chemical shift values of d = 32.8, 39.2, and 169.0 ppm, respectively, as obtained from the dipolar dephasing spectrum. The signals at the chemical shift values of d = 8.0, 13.7, and 19.2 ppm correspond to methyl carbons C-18, C-21, and C-19, and the ones at d = 54.3, 55.3, and 67.6 ppm to carbons C-17, C-14, and C-3, respectively. The 13C CP/MAS NMR spectrum of form 2 differs clearly from those of 3, 4, 5, and 6, which show solvent peaks in their spectra (Fig. 1b–e) revealing that these solvents tend to include in the crystal lattice of lithocholyl-N-(2-aminoethyl)amide. Although it is considered in the literature that lithocholylamides rarely form inclusion compounds [17, 18], our results indicate that 1 forms inclusion crystals/solvates at least with p-xylene, isopropyl alcohol, diethyl ether, and chlorobenzene. In solvate 3, the resonances at d = 136.9 and 130.3 ppm correspond to the aromatic carbons of p-xylene (Fig. 1b), in solvate 4 the resonance at d = 70.1 ppm corresponds to the methine carbon of iso-propyl alcohol (Fig. 1c), in solvate 5 the resonance at d = 69.4 ppm corresponds to the methylene carbon of diethyl ether (Fig. 1d), and in solvate 6 the resonances from aromatic carbons of chlorobenzene are from d = 125.7 to 133.6 ppm (Fig. 1e). In all solvates it is observed that the carbonyl resonances are somewhat deshielded compared with the pure polymorph 2. In order to obtain more information, we compared the anisotropic shielding tensor values of carbonyl carbons between the pure polymorph (2) and the p-xylene solvate (3). It is well
123
188
Fig. 1
Struct Chem (2010) 21:185–190
13
C CP/MAS NMR spectra of 2 (a), 3 (b), 4 (c), 5 (d), and 6 (e)
Table 1 13C CP/MAS NMR chemical shifts (ppm from TMS) of polymorph 2 Nucleus
d(13C)
Nucleus
d(13C)
Nucleus
d(13C)
1
30.9
10
32.8
19
2
28.4
11
19.2
20
3
67.6
12
37.9
4
36.3
13
39.2
5
37.9
14
55.3
23
28.4
6
26.2
15
20.4
24
169.0
7
26.2
16
24.6
25
39.2
8
33.4
17
54.3
26
39.9
9
39.9
18
8.0
Parameter
2
3
19.2
d11
221.0
219.4
33.4
d22
170.4
175.3
21
13.7
d33
115.3
119.8
22
26.2
g
0.90
0.85
d
-80.4
-77.5
diso
169.0
171.5
known that detailed information about the electronic surroundings of each nucleus, which reflects internuclear interactions, can be obtained from the inspection of the three-dimensional tensor of the chemical shift. For rotating solids this information can be obtained from spectral spinning side band (SSB) analysis. The 13C dij tensor parameters were obtained for the pure polymorph 2 recrystallized from acetonitrile and the p-xylene solvate 3 from the DMFIT simulation of the intensities of the spinning side bands at different spinning rates (1000, 2000, and 3000 Hz). The average values of dij parameters are listed in Table 2. The orientation of the d22 tensor component is close to the carbonyl bond vector and it is therefore most sensitive for hydrogen bonding [32]. Deshielding of this tensor component in the solvate 3 compared to the pure polymorph may be caused by an interaction with the solvent molecule or different packing pattern of the host molecule.
123
Table 2 CSA tensor components (ppm) of carbonyl carbon in 2 and 3
Assuming d11 [ d22 [ d33, diso = d11 ? d22 ? d33, g = (d22 - d11)/ d33 - diso, d = d33 - (d11 ? d22)/2
The 15N CP/MAS NMR spectra of forms 2–6 are depicted in Fig. 2, and the 15N chemical shifts are listed in Table 3. The resonances at the chemical shift values between d = -254.2 and -262.8 ppm (from ext. nitromethane) are assigned to originate from the amide (NH) and the ones between d = -356.3 and -363.1 ppm from the amino (NH2) nitrogens, respectively. Marked differences compared to the 15N CP/MAS NMR spectrum of the pure polymorph 2 (Fig. 2a) are observed for the rest of the samples most probably related to the inclusion of solvents. The nitrogen resonances of form 3 do not deviate much from those of the pure polymorph 2 (see Table 3). However, the signal of the NH2 nitrogen appears somewhat broader (Fig. 2b). The 15N CP/MAS NMR spectra of 4 and 5 show several NH and NH2 resonances (Fig. 2c, d), indicating mixtures of different forms. The nitrogen resonances of solvate 6 are narrow and deshielded compared to form 2, implying fixed position of the solvent molecule in the crystal lattice. The type and positioning of the solvent
Struct Chem (2010) 21:185–190
Fig. 2
189
15
Table 3 2–6
N CP/MAS NMR spectra of 2 (a), 3 (b), 4 (c), 5 (d), and 6 (e) 15
N NMR chemical shifts (ppm from ext. nitromethane) of
Polymorph/solvate
NH
NH2
2
-261.3
-361.0
3
-261.8
-363.1
4
-259.7 … -262.8
-357.6 … -362.0
5
-260.9
-359.7 … -362.0
6
-254.2
-356.3
Despite numerous recrystallization efforts of 1, we were unable to obtain single crystals suitable for single crystal X-ray structure determination. Nevertheless, we got powdered crystalline samples suitable for PXRD. Powder X-ray diffractograms of the pure polymorph recrystallized from acetonitrile (2) and the p-xylene solvate 3 are presented in Fig. 3. The polymorph and the solvate have clearly different diffraction patterns that corroborate our NMR interpretations. Hence lithocholyl-N-(2-aminoethyl)amide clearly exhibits different forms in the solid state.
Conclusions
Fig. 3 Powder X-ray diffractograms of solvate 3 (a) and pure polymorph 2 (b)
molecules seem to have a marked effect on the spectra. This is probably due to the interactions between the solvent and the sample molecule, which can affect some change in the side chain conformation.
We have shown that 13C and 15N CP/MAS NMR techniques assisted with PXRD form a useful strategy to investigate solid state structures of lithocholyl-N-(2-aminoethyl)amide. Based on 13C and 15N CP/MAS NMR spectra, the synthesis product was concluded to consist of a mixture of different forms. One pure polymorph was separated by recrystallization from three different solvents: acetonitrile, tetrahydrofuran, and 1,4-dioxane, whereas recrystallization from p-xylene, iso-propyl alcohol, diethyl ether, and chlorobenzene resulted in respective solvates. The polymorph/solvate formation characteristics of lithocholyl-N-(2-aminoethyl)amide may prove important in understanding the gel formation process as found in the case of chlorobenzene, or in developing novel drug delivery applications. Acknowledgments Academy of Finland is gratefully acknowledged for financial support (project no. 119616 for E.S., project no. 7127006 for K.A.).
123
190
References 1. Desiraju GR (1997) Science 278:404 2. Hilfiker R (ed) (2006) Polymorphism: in the pharmaceutical industry. Wiley-VCH, Weinheim 3. Bernstein J (2002) Polymorphism in molecular crystals. Oxford University Press, Oxford 4. Liang J, Ma Y, Chen B, Munson EJ, Davis HT, Binder D, Chang H-T, Abbas S, Hsu F-L (2001) J Phys Chem B 105:9653 5. Kampschulte L, Lackinger M, Maier A-K, Kishore RSK, Griessel S, Schmittel M, Heckl WMJ (2006) Phys Chem B 110:10829 6. Virtanen E, Kolehmainen E (2004) Eur J Org Chem 3385 7. Marples BA, Stretton RJ (1990) WO 9013298 8. Hofmann AF (1995) Italian J Gastroenterol 27:106 9. Enhsen A, Kramer W, Wess G (1998) Drug Discov Today 3:409 10. Li C, Peters AS, Meredith EL, Allman GW, Savage PB (1998) J Am Chem Soc 120:2961 11. Li C, Budge LP, Driscoll CD, Willardson BM, Allman GW, Savage PB (1999) J Am Chem Soc 121:931 12. Criado JJ, Manzano JL, Rodriguez-Fernandez E (2003) J Inorg Biochem 96:311 13. Briz O, Macias RIR, Vallejo M, Silva A, Serrano MA, Marin JJG (2003) Mol Pharmacol 63:742 14. Bellini AM, Quaglio MP, Cavazzini G, Ceccherini R (1984) Farmaco Sci 39:305 15. Fantin G, Fogagnolo M, Bortolini O, Masciocchi N, Galli S, Sironi A (2003) New J Chem 27:1794 16. Epand RM, Epand RF, Hughes DW, Sayer BG, Borochov N, Bach D, Wachtel E (2005) Chem Phys Lipids 135:39 17. Miyata M, Tohnai N, Hisaki I (2007) Acc Chem Res 40:694 18. Miyata M, Tohnai N, Hisaki I (2007) Molecules 12:1973 19. Valkonen A, Kolehmainen E, Lahtinen M, Sieva¨nen E, Noponen V, Tolonen M, Kauppinen R (2007) Molecules 12:2161 20. Valkonen A, Lahtinen M, Tamminen J, Kolehmainen E (2008) J Mol Struct 886:197 21. Valkonen A, Lahtinen M, Kolehmainen E (2008) Steroids 73:1228 22. Yang J-H, Ho Y, Tzou D-LM (2008) Magn Reson Chem 46:718 23. Allen FH (2002) Acta Cryst B 58:380 24. Hu JZ, Wu X, Yang N, Li L, Ye C, Qin K (1996) Solid State Nucl Magn Reson 6:187 25. Smith EDL, Hammond RB, Jones MJ, Roberts KJ, Mitchell JBO, Price SL, Harris RK, Apperley DC, Cherryman JC, Docherty R (2001) J Phys Chem B 105:5818
123
Struct Chem (2010) 21:185–190 26. Portieri A, Harris RK, Fletton RA, Lancaster RW, Threlfall T (2004) Magn Reson Chem 42:313 27. Apperley DC, Forster AH, Fournier R, Harris RK, Hodgkinson P, Lancaster RW, Rades T (2005) Magn Reson Chem 43:881 28. Harris RK, Ghi PY, Puschmann H, Apperley DC, Griesser UJ, Hammond RB, Caiyun M, Roberts KJ, Pearce GJ, Yates JR, Pickard CJ (2005) Org Process Res Dev 9:902 29. Harris RK (2006) Analyst 13:351 30. Othman A, Evans JSO, Evans IR, Harris RK, Hodgkinson P (2007) J Pharm Sci 96:1380 31. Harris RK, Hodgkinson P, Larsson T, Muruganantham A, Yme´n I, Yufit DS, Zorin V (2008) Cryst Growth Des 8:80 32. Wei Y, Lee D-K, Ramamoorthy A (2001) J Am Chem Soc 123:6118 33. Chen YY, Luo SY, Hung SC, Chan SI, Tzou DLM (2005) Carbohydr Res 340:723 34. Offerdahl TJ, Salsbury JS, Dong Z, Grant DJW, Schroeder SA, Prakash I, Gorman EM, Barich DH, Munson EJ (2005) J Pharm Sci 94:2591 35. Wawer I, Wolniak M, Paradowska K (2006) Solid State Nucl Magn Reson 30:106 36. Maccaroni E, Alberti E, Malpezzi L, Masciocchi N, Vladiskovic C (2008) Int J Pharm 351:144 37. Fonseca I, Hayes SE, Blu¨mich B, Bertmer M (2008) Phys Chem Chem Phys 10:5898 38. Olejniczak S, Mikuła-Pacholczyk J, Hughes CE, Potrzebowski MJ (2008) J Phys Chem B 112:1586 39. Nieuwendaal RC, Bertmer M, Hayes SE (2008) J Phys Chem B 112:12920 40. Pandey PS, Singh RB (1997) Tetrahedron Lett 38:5045 41. Tamminen J, Kolehmainen E, Linnanto J, Vainiotalo P, Vuorikoski S, Kauppinen R (2000) J Inclusion Phenom Macrocycl Chem 37:121 42. Pandey PS, Rai R, Singh RB (2002) J Chem Soc Perkin Trans 1:918 43. Massiot D, Fayon F, Capron M, King I, Le Calve´ S, Alonso B, Durand J-O, Bujoli B, Gan Z, Hoatson G (2002) Magn Reson Chem 40:70 44. Wu X, Burns ST, Zilm KW (1994) J Magn Reson A 111:29
