pharmaceuticals and natural products Acta Crystallographica Section C
A new, more precise, refinement of the 6-APA structure and a detailed structural analysis are presented herein.
Crystal Structure Communications ISSN 0108-2701
Crystal structure and packing energy calculations of (+)-6-aminopenicillanic acid Sofiane Saouane,a Gernot Buthb and Francesca P. A. Fabbiania*
Keywords: crystal structure; (+)-6-aminopenicillanic acid; zwitterion structures; semi-synthetic penicillin precursor; PIXEL calculations; computational methods.
According to the literature, the crystal structure of 6-APA was first described in 1963 by R. D. Diamand (Diamand, 1963). This is the surname that appears in the original DPhil. thesis, although in all subsequent references to this work the surname is spelled as Diamond. In the thesis, the author was unsure whether the molecule in the crystal is zwitterionic or neutral. Gałdecki & Werfel (1978) subsequently reported a preliminary refinement of 6-APA; this corresponds to the refcode AMPENA in the Cambridge Structural Database (CSD; (Version 5.34, including updates to February 2013; Allen, 2002) and the entry shows a neutral molecule. In the original paper, which corresponds to a conference abstract, no details on H atoms are given but a reference to a ‘carboxylate group’ is made; from this it can be inferred that the authors identified a zwitterion, despite the reported bond lengths of 1.329 and ˚ . No atomic coordinates for any crystal form that 1.175 A includes 6-APA have been deposited in the CSD. In their original paper, Gałdecki & Werfel report a final R value of 0.123 and list all bond lengths but do not report fractional coordinates for their structure. The reference for Gałdecki & Werfel (1978) is sometimes wrongly cited (Kirschbaum, 1986; Clayden et al., 1986). 6-APA has also been studied by computational methods. Mwangi & Garside (1996) stated that misplaced H atoms of 6-APA in the structure reported by Diamand (1963) rendered it difficult to use the data in any molecular mechanics calculations. Stroganov et al. (2003) reported that the molecular geometry obtained for their single-molecule gas-phase quantum mechanical calculations by the RHF/6-31G* method is in good agreement with the experimental data of Gałdecki & Werfel (1978), although the chemical formula in their study appears to be incorrect. In view of the above, we have redetermined the crystal structure of (I) and extended the study with lattice energy calculations.
1. Introduction
2. Experimental
(+)-6-Aminopenicillanic acid (6-APA), (I), is a molecule with antibacterial properties (Rolinson & Stevens, 1961) produced by hydrolysis of the natural benzylpenicillin known as penicillin G (Sakaguchi & Murao, 1950). Furthermore, 6-APA is the main intermediate for the production of various semisynthetic penicillins on an industrial scale (Ferreira et al., 2006). Despite the fact that 6-APA is the core of all penicillins, there is a lack of fully defined structural data in the literature.
2.1. Synthesis and crystallization
a
Abteilung Kristallographie, GZG, Georg-August-Universita¨t Go¨ttingen, Goldschmidtstrasse 1, D-37077 Go¨ttingen, Germany, and bInstitut fu¨r Synchrotronstrahlung (ISS/ANKA), Karlsruhe Institute of Technology (KIT), Hermann-vonHelmholtz-Platz 1, D-76344 Eggenstein-Leopoldshafen, Germany Correspondence e-mail: [email protected] Received 31 May 2013 Accepted 19 September 2013
The X-ray single-crystal structure of (2S,5R,6R)-6-amino-3,3dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid, commonly known as (+)-6-aminopenicillanic acid (C8H12N2O3S) and a precursor of a variety of semi-synthetic penicillins, has been determined from synchrotron data at 150 K. The structure represents an ordered zwitterion and the crystals are nonmerohedrally twinned. The crystal structure is composed of a three-dimensional network built by three charge-assisted hydrogen bonds between the ammonium and carboxylate groups. The complementary analysis of the crystal packing by the PIXEL method brings to light the nature and ranking of the energetically most stabilizing intermolecular interaction energies. In accordance with the zwitterionic nature of the structure, PIXEL lattice energy calculations confirm the predominance of the Coulombic term (379.1 kJ mol1) ahead of the polarization (141.4 kJ mol1), dispersion (133.7 kJ mol1) and repulsion (266.3 kJ mol1) contributions.
1238
# 2013 International Union of Crystallography
6-APA (Sigma) was used as supplied (purity 96%). Single crystals of (I) were obtained by slow evaporation at ambient temperature of an aqueous solution of the compound. Crystals were small in size and exhibited nonmerohedral twinning, as confirmed by screening several crystals by diffraction. Data were collected on beamline SCD at the ANKA synchrotron source at the Karlsruhe Institute of Technology using a Bruker
doi:10.1107/S0108270113025924
Acta Cryst. (2013). C69, 1238–1242
pharmaceuticals and natural products Table 1 Experimental details. Crystal data Chemical formula Mr Crystal system, space group Temperature (K) ˚) a, b, c (A ˚ 3) V (A Z Radiation type (mm1) Crystal size (mm) Data collection Diffractometer Absorption correction Tmin, Tmax No. of measured, independent and observed [I > 2(I)] reflections Rint ˚ 1) (sin /)max (A Refinement R[F 2 > 2(F 2)], wR(F 2), S No. of reflections No. of parameters H-atom treatment ˚ 3) max, min (e A Absolute structure
Absolute structure parameter
C8H12N2O3S 216.26 Orthorhombic, P212121 150 6.1954 (4), 10.4543 (7), 14.7654 (9) 956.33 (11) 4 ˚ Synchrotron, = 0.71075 A 0.32 0.08 0.06 0.04
Bruker SMART APEXII CCD areadetector diffractometer Multi-scan (TWINABS; Sheldrick, 2008b) 0.643, 0.745 2130, 2130, 1966 0.043 0.609
0.029, 0.073, 1.05 2130 142 H atoms treated by a mixture of independent and constrained refinement 0.21, 0.20 Flack x parameter determined using 678 quotients [(I+) (I)]/ [(I+) + (I)] (Parsons & Flack, 2004) 0.06 (6)
Computer programs: APEX2 (Bruker, 2012), SAINT (Bruker, 2012), SHELXS97 (Sheldrick, 2008a), SHELXL2012 (Sheldrick, 2008a), Mercury (Macrae et al., 2008), publCIF (Westrip, 2010) and SHELXLE (Hu¨bschle et al., 2011).
SMART APEX CCD diffractometer with silicon-mono˚. chromated radiation of = 0.71075 A 2.2. Refinement
The crystal structure presented in this work was refined using nonmerohedrally twinned data presenting two major domains. A third domain was not taken into account on the basis of its very small contribution and negligible overlap with the other domains. The nonmerohedral twin law corresponds to a twofold axis about the [001] direct-lattice direction. The fractional contribution of the second component was refined to 0.507 (4). The H atoms of the final refined model were normalized using the RETCIF and RETCOR modules of the CLP program (Gavezzotti, 2011), before calculating the molecular electron density using the program GAUSSIAN09W (Frisch et al., 2009) at the MP2/6-31G** level of theory. The electron-density model of the molecule was then used in the PIXELC (Gavezzotti, 2011) module for energy calculations, using a distance cut-off from the central molecule ˚. of 30 A Crystal data, data collection and structure refinement details are summarized in Table 1. The program TWINABS (Sheldrick, 2008b) was used to generate a merged reflection file in the SHELXL HKLF5 format (Sheldrick, 2008a). Data Acta Cryst. (2013). C69, 1238–1242
statistics were as follows: 6566 data (1807 unique) for domain 1 only with a mean I/(I) value of 12.9, 6634 data (1806 unique) for domain 2 only with a mean I/(I) value of 13.4, and 590 data (349 unique) for both domains with a mean I/(I) value of 21.2. The quoted Rint value (0.043) comes from scaling all single and composite reflections involving both domains. For the generation of an HKLF5 file, all observations containing domain 2 were chosen. Single reflections that also occur in composites were included for merging. H atoms on C atoms were placed geometrically and constrained to ride on their parent atoms, with Uiso(H) values assigned in the range 1.2–1.5Ueq of the parent atom. H atoms on the N atom were clearly visible in a difference Fourier map and their positions were freely refined. The chirality of the compound was known from the specifications of the manufacturer, and refinement of the Flack x parameter to a value of 0.06 (6) using Parsons’ quotients method confirmed it (Parsons & Flack, 2004). The four omitted low-angle reflections have partial contribution from domain 1 and were not properly integrated. Closer inspection of the diffraction frames revealed that these reflections were compromised by a distinctively higher background or were partially shaded by the beam stop.
3. Results and discussion As previously reported, 6-APA crystallizes in the chiral orthorhombic P212121 space group with one molecule in the asymmetric unit (Fig. 1). Difference Fourier maps clearly indicate the zwitterionic nature of the molecule. The compound was first described by Batchelor et al. (1959) as possessing definite antibacterial properties of a lower order than those of benzylpenicillin but a longer metabolization cycle. Subsequently and by biological assays, Rolinson & Stevens (1961) stated that 6-APA activity is similar to that of benzylpenicillin depending on the bacterial species. In comparison, the theoretical structure–activity relationship described by the Cohen (c) (1983) and Woodward (h) (1980)
Figure 1 The molecular structure of (I), showing the atom-numbering scheme. Displacement ellipsoids are drawn at the 50% probability level. Saouane et al.
C8H12N2O3S
1239
pharmaceuticals and natural products Table 2 ˚ , ). Hydrogen-bond geometry (A D—H A i
N13—H13A O10 N13—H13B O9ii N13—H13C O9iii N13—H13C O14iv C6—H6 O9iii
D—H
H A
D A
D—H A
0.98 (4) 0.96 (5) 0.90 (4) 0.90 (4) 1.00
1.68 (4) 1.79 (5) 2.04 (4) 2.66 (4) 2.53
2.647 (3) 2.738 (4) 2.825 (3) 3.242 (3) 3.018 (3)
170 (4) 169 (4) 145 (4) 124 (3) 110
Symmetry codes: (i) x þ 2; y þ 12; z þ 32; (ii) x þ 1; y þ 12; z þ 32; (iii) x þ 32, y þ 1; z 12; (iv) x þ 12; y þ 32; z þ 1.
Figure 2 A projection of the structure of 6-APA, (I), along the a axis. Hydrogenbond patterns showing two distinct graph-set assignments C(8)b and C(8)c are highlighted (cf. Table 2). H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
parameters is in good agreement with the biological assays. The Woodward parameter (h) defines the height of the pyramid with atoms C2, C5 and C7 as the base and N1 as the apex. The Cohen parameter (c) defines the distance between the lactam carbonyl O atom (O14) and the carboxylate C atom (C8). The values calculated using PLATON (Spek, 2009) of c =
˚ and h = 0.333 (2) A ˚ are within the reported ranges 4.475 (4) A ˚ of 3.0–4.5 and 0.25–0.5 A, respectively, for active -lactam structures (Nangia et al., 1996). In addition, the third parameter introduced by Nangia et al. (1996), concerning the torsion angle between the carboxylic acid and amide functions (O14—N1—C2—C8 herein), of 139.19 (18) also lies within the defined range (30–160 ) for biologically active -lactam structures. The crystal packing of 6-APA is mainly stabilized by hydrogen bonds. Each H atom of the ammonium group is involved in hydrogen bonds as a donor to the carboxylate groups of three neighbouring molecules (Table 2); carboxylate atom O9 is a double hydrogen-bond acceptor. Hence, the firstlevel graph-set motifs (Etter, 1990; Etter et al., 1990; Bernstein et al., 1995) are described as three C(8) chains. Two of these [C(8)a and C(8)b] run along the b axis and the third one [C(8)c] runs along the c axis (cf. Figs. 2–4). These chains combine to form a series of further chains at a secondary level, as well as large rings containing up to 38 atoms involving six molecules. Whilst there are three strong hydrogen bonds per molecule, carbonyl atom O14 forms an additional weak hydrogen bond with the ammonium N13—H13C group of a neighbouring molecule (see Table 2 for details). In addition to these classical hydrogen bonds, there is a short C—H O interaction involving the carboxylate group (Table 2). Overall, the hydrogen-bonded pattern is best described as threedimensional. For a better understanding of the intermolecular interactions in 6-APA and quantification of the interaction energies, the PIXEL method, as incorporated in the CLP program (Gavezzotti, 2011), was used for molecule–molecule and lattice energy calculations. PIXEL calculations allow the analysis of lattice and intermolecular interaction energies between pairs of molecules in terms of Coulombic, polarization, dispersion and repulsion contributions. Full details of the present PIXEL calculations are available in the Supplementary materials. The total PIXEL energy, which is the sum of these four energy contributions, gives an indication of the overall interaction energy for a particular dimer and for crystal packing, with the latter being traceable back to the sublimation energy. However, it is the breakdown of these energies into the four different terms that makes the PIXEL method a powerful tool for crystal structure analysis. The
Figure 3 A projection of the structure of 6-APA along the b axis. The C(8)c graph-set forming extended hydrogen-bonded chains of edge-to-edge molecules is highlighted. H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
1240
Saouane et al.
C8H12N2O3S
Acta Cryst. (2013). C69, 1238–1242
pharmaceuticals and natural products
Figure 5 Molecular pairs identified by PIXEL calculations as having the three most significant interaction energies; see Table 3 for details. H atoms have been omitted for clarity.
Table 4 PIXEL output of the lattice energy (kJ mol1).
Figure 4 A projection of the structure of 6-APA along the c axis. C(8)a and C(8)b graph sets form parallel motifs which run along the b axis. H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
symmetry-independent dimers with the three most significant intermolecular energies are detailed in Table 3 and depicted in Fig. 5; the predominant interaction in each dimer corresponds to the three hydrogen bonds described in Table 2. At 28.5 kJ mol1, the next energetically significant interaction [corresponding to symmetry code (x + 52, y + 1, z 12)] is substantially less stabilizing. From Table 3, a directly proportional correlation between the Coulombic and the total PIXEL energies is apparent. The predominance of the Coulombic term can be explained by the presence of the zwitterionic form of 6-APA, which causes a sharp distribution of charges. As expected, the lattice energy is also dominated by the Coulombic term (Table 4). The largest interaction energy involves C(8)a hydrogen-bonded chains: this intermolecular interaction not only comprises the shortest hydrogen bond among all other dimers but there is also a contribution from a short S Ocarboxylate contact, which is not ˚ , is shorter than seen in other dimers and which, at 3.256 (2) A the sum of the van der Waals radii (Bondi, 1964). The second
Coulombic
Polarization
Dispersion
Repulsion
Total PIXEL energy
379.1
141.4
133.7
266.3
387.9
largest contribution comes from the repulsion term. This is significantly large for the C(8)a and C(8)b graph-set motifs and can be ascribed to symmetry effects: in the C(8)a and C(8)b motifs, pairs of molecules are juxtaposed, whilst in the C(8)c motif the molecules are more linearly distributed, as evidenced by the large distance between the centres of mass (cf. Fig. 5 and Table 3). On the basis of crystal growth and habit considerations, Mwangi & Garside (1996) reported that the strongest of the intermolecular interactions in 6-APA are along the h100i and h001i directions. This interpretation can be partially reconciled with the analysis of hydrogen bonds in the structure as follows: the intermix of hydrogen bonds C(8)a and C(8)b results in a C22 (6) motif at the secondary-level graph set, and it is clear that this motif builds columns of molecules along the a axis of the unit cell. It is interesting to note that the interaction with the highest intermolecular energy according to the PIXEL calculations and which runs along the b axis was also identified by Mwangi & Garside (1996) as a strong interaction, but bonding in the h010i directions was described as ‘involving only weak bonding’.
Table 3 Output from PIXEL calculations of the three most significant intermolecular interaction energies. D denotes the donor, A the acceptor and Cm the centre of mass. Symmetry code
D A distance ˚) (A
Graph set
Cm—Cm distance ˚) (A
Coulombic (kJ mol1)
Polarization (kJ mol1)
Dispersion (kJ mol1)
Repulsion (kJ mol1)
Total (kJ mol1)
Direction
(x + 2, y + 12, z + 32) (x + 32, y + 1, z 12) (x + 1, y + 12, z + 32)
2.647 (3) 2.825 (3) 2.738 (4)
C(8)a C(8)c C(8)b
6.495 7.709 5.746
183.6 135.1 119.5
73 38.7 61.2
24 21.1 40.4
102.9 43.5 92.2
177.7 151.4 128.9
b axis c axis b axis
Acta Cryst. (2013). C69, 1238–1242
Saouane et al.
C8H12N2O3S
1241
pharmaceuticals and natural products Funding from the Deutsche Forschungsgemeinschaft DFG (project No. FA 964/1-1 for FPAF and SS) and from the Georg-August-Universita¨t Go¨ttingen (Dorothea Schlo¨zer Forschungsstipendium for FPAF) is gratefully acknowledged. The award of synchrotron beamtime at KIT–ISS–Anka is also gratefully acknowledged. The authors thank the Editorial Office of the IUCr for retrieving a copy of the original paper of Gałdecki & Werfel (1978) and Ms Karen Langdon of the Bodleian Library of the University of Oxford for providing a copy of the DPhil. thesis by R. D. Diamand. The authors also thank Professor Angelo Gavezzotti (Milan) for his advice on running PIXELC calculations. Supplementary data for this paper are available from the IUCr electronic archives (Reference: FM3006). Services for accessing these data are described at the back of the journal.
References Allen, F. H. (2002). Acta Cryst. B58, 380–388. Batchelor, F. R., Doyle, F. P., Nayler, J. H. C. & Rolinson, G. N. (1959). Nature (London), 183, 257–258. Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew. Chem. Int. Ed. Engl. 34, 1555–1573. Bondi, A. (1964). J. Phys. Chem. 68, 441–451. Bruker (2012). APEX2 and SAINT. Bruker AXS Inc., Madison, Wisconsin, USA. Clayden, N. J., Dobson, C. M., Lian, L. & Marktwyman, J. (1986). J. Chem. Soc. Perkin Trans. 2, pp. 1933–1940.
1242
Saouane et al.
C8H12N2O3S
Cohen, N. C. (1983). J. Med. Chem. 26, 259–264. Diamand, R. D. (1963). DPhil thesis, University of Oxford, England. Etter, M. C. (1990). Acc. Chem. Res. 23, 120–126. Etter, M. C., MacDonald, J. C. & Bernstein, J. (1990). Acta Cryst. B46, 256– 262. Ferreira, J. S., Straathof, A. J. J., Li, X., Ottens, M., Franco, T. T. & van der Wielen, L. A. M. (2006). Ind. Eng. Chem. Res. 45, 6740–6744. Frisch, M. J. et al. (2009). GAUSSIAN09W. Gaussian Inc., Wallingford, Connecticut, USA. Gałdecki, Z. & Werfel, M. (1978). Acta Cryst. A34, S90. Gavezzotti, A. (2011). New J. Chem. 35, 1360–1368. Hu¨bschle, C. B., Sheldrick, G. M. & Dittrich, B. (2011). J. Appl. Cryst. 44, 1281– 1284. Kirschbaum, J. (1986). Analytical Profiles of Drug Substances, Vol. 15, edited by K. Florey, ch. 11, pp. 427–507. New Jersey: Academic Press Inc. Macrae, C. F., Bruno, I. J., Chisholm, J. A., Edgington, P. R., McCabe, P., Pidcock, E., Rodriguez-Monge, L., Taylor, R., van de Streek, J. & Wood, P. A. (2008). J. Appl. Cryst. 41, 466–470. Mwangi, S. M. & Garside, J. (1996). J. Cryst. Growth, 166, 1078–1083. Nangia, A., Biradha, K. & Desiraju, G. R. (1996). J. Chem. Soc. Perkin Trans. 2, pp. 943–953. Parsons, S. & Flack, H. (2004). Acta Cryst. A60, s61. Rolinson, G. N. & Stevens, S. (1961). Proc. R. Soc. London Ser. B, 154, 509– 513. Sakaguchi, K. & Murao, S. (1950). J. Agric. Chem. Soc. Jpn, 23, 411. Sheldrick, G. M. (2008a). Acta Cryst. A64, 112–122. Sheldrick, G. M. (2008b). TWINABS. Bruker AXS Inc., Madison, Wisconsin, USA Spek, A. L. (2009). Acta Cryst. D65, 148–155. Stroganov, O. V., Chilov, G. G. & Sˇvedas, V. K. (2003). J. Mol. Struct. (THEOCHEM), 631, 117–125. Westrip, S. P. (2010). J. Appl. Cryst. 43, 920–925. Woodward, R. B. (1980). Philos. Trans. R. Soc. London Ser. B, 289, 239– 250.
Acta Cryst. (2013). C69, 1238–1242
supplementary materials
supplementary materials Acta Cryst. (2013). C69, 1238-1242
[doi:10.1107/S0108270113025924]
Crystal structure and packing energy calculations of (+)-6-aminopenicillanic acid Sofiane Saouane, Gernot Buth and Francesca P. A. Fabbiani Computing details Data collection: APEX2 (Bruker, 2012); cell refinement: SAINT (Bruker, 2012); data reduction: SAINT (Bruker, 2012); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008a); program(s) used to refine structure: SHELXL2012 (Sheldrick, 2008a); molecular graphics: Mercury (Macrae et al., 2008); software used to prepare material for publication: publCIF (Westrip, 2010). (2S,5R,6R)-6-Amino-3,3-dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid Crystal data C8H12N2O3S Mr = 216.26 Orthorhombic, P212121 a = 6.1954 (4) Å b = 10.4543 (7) Å c = 14.7654 (9) Å V = 956.33 (11) Å3 Z=4 F(000) = 456
Dx = 1.502 Mg m−3 Synchrotron radiation, λ = 0.71075 Å Cell parameters from 2730 reflections θ = 2.4–25.6° µ = 0.32 mm−1 T = 150 K Block, colourless 0.08 × 0.06 × 0.04 mm
Data collection Bruker SMART APEXII CCD area-detector diffractometer Radiation source: fine-focus sealed tube Silicon monochromator ω scans Absorption correction: multi-scan (TWINABS; Sheldrick, 2008b) Tmin = 0.643, Tmax = 0.745
2130 measured reflections 2130 independent reflections 1966 reflections with I > 2σ(I) Rint = 0.043 θmax = 25.6°, θmin = 2.4° h = −7→7 k = −12→12 l = −17→17
Refinement Refinement on F2 Least-squares matrix: full R[F2 > 2σ(F2)] = 0.029 wR(F2) = 0.073 S = 1.05 2130 reflections 142 parameters 0 restraints Primary atom site location: structure-invariant direct methods
Acta Cryst. (2013). C69, 1238-1242
Secondary atom site location: difference Fourier map Hydrogen site location: mixed H atoms treated by a mixture of independent and constrained refinement w = 1/[σ2(Fo2) + (0.034P)2 + 0.1358P] where P = (Fo2 + 2Fc2)/3 (Δ/σ)max < 0.001 Δρmax = 0.21 e Å−3 Δρmin = −0.20 e Å−3
sup-1
supplementary materials Absolute structure: Flack x parameter determined using 678 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons & Flack, 2004) Absolute structure parameter: −0.06 (6) Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Refinement of F2 against ALL reflections. From TWINABS: 6566 data (1807 unique) involve domain 1 only, mean I/σ 12.9; 6634 data (1806 unique) involve domain 2 only, mean I/σ 13.4; 590 data (349 unique) involve 2 domains, mean I/σ 21.2. The quoted Rint value (0.043) comes from scaling all single and composite reflections involving both domains. Rint = 0.039 from scaling all singles of domain 2. Rint = 0.0554 from scaling single and composite reflections of domain 1. For the generation of an HKLF5 file, all observations containing domain 2 were chosen. Data were merged in TWINABS according to point group 222; single reflections that also occur in composites were included for merging. Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)
N1 C2 H2 C3 S4 C5 H5 C6 H6 C7 C8 O9 O10 C11 H11A H11B H11C C12 H12A H12B H12C N13 H13A H13B H13C O14
x
y
z
Uiso*/Ueq
0.7032 (4) 0.6065 (5) 0.4597 0.5848 (5) 0.79266 (11) 0.8738 (5) 1.0223 0.8143 (5) 0.9168 0.6148 (5) 0.7478 (5) 0.6525 (4) 0.9474 (3) 0.6248 (6) 0.7699 0.5175 0.6126 0.3620 (5) 0.3524 0.2516 0.3387 0.7738 (5) 0.883 (7) 0.630 (8) 0.781 (7) 0.4475 (4)
0.5375 (2) 0.5155 (2) 0.4774 0.6508 (3) 0.75459 (7) 0.6344 (2) 0.5997 0.6525 (3) 0.6046 0.5693 (3) 0.4232 (2) 0.34077 (19) 0.4377 (2) 0.6501 (3) 0.6170 0.5954 0.7374 0.7053 (3) 0.7927 0.6520 0.7063 0.7812 (2) 0.841 (4) 0.810 (4) 0.776 (4) 0.5422 (2)
0.71283 (16) 0.80083 (18) 0.7927 0.84665 (19) 0.79121 (5) 0.70872 (19) 0.7188 0.60750 (18) 0.5677 0.6307 (2) 0.85603 (18) 0.90362 (13) 0.84992 (15) 0.94842 (19) 0.9607 0.9781 0.9720 0.8262 (2) 0.8498 0.8551 0.7606 0.57153 (17) 0.594 (3) 0.586 (3) 0.511 (3) 0.59357 (15)
0.0181 (5) 0.0172 (6) 0.021* 0.0174 (6) 0.01909 (19) 0.0180 (6) 0.022* 0.0196 (6) 0.023* 0.0211 (6) 0.0170 (6) 0.0211 (5) 0.0249 (5) 0.0241 (7) 0.036* 0.036* 0.036* 0.0237 (7) 0.036* 0.036* 0.036* 0.0185 (5) 0.048 (12)* 0.044 (11)* 0.049 (12)* 0.0317 (6)
Acta Cryst. (2013). C69, 1238-1242
sup-2
supplementary materials Atomic displacement parameters (Å2)
N1 C2 C3 S4 C5 C6 C7 C8 O9 O10 C11 C12 N13 O14
U11
U22
U33
U12
U13
U23
0.0218 (12) 0.0179 (15) 0.0174 (16) 0.0185 (3) 0.0182 (14) 0.0256 (17) 0.0278 (17) 0.0204 (17) 0.0225 (12) 0.0175 (12) 0.0306 (18) 0.0193 (16) 0.0201 (14) 0.0369 (14)
0.0164 (11) 0.0160 (13) 0.0162 (13) 0.0183 (3) 0.0170 (13) 0.0181 (13) 0.0171 (14) 0.0148 (12) 0.0224 (10) 0.0265 (11) 0.0232 (14) 0.0222 (14) 0.0200 (12) 0.0318 (12)
0.0160 (11) 0.0178 (15) 0.0186 (15) 0.0205 (3) 0.0188 (14) 0.0149 (14) 0.0185 (15) 0.0158 (13) 0.0185 (10) 0.0308 (12) 0.0186 (15) 0.0295 (16) 0.0153 (12) 0.0265 (12)
−0.0014 (9) −0.0017 (11) −0.0009 (11) −0.0030 (3) −0.0004 (11) −0.0005 (13) −0.0027 (12) 0.0001 (11) −0.0028 (8) 0.0036 (8) 0.0056 (14) 0.0023 (11) −0.0017 (10) −0.0142 (11)
−0.0010 (10) −0.0014 (12) −0.0004 (11) 0.0005 (3) −0.0003 (13) 0.0020 (12) −0.0014 (13) −0.0005 (11) −0.0016 (9) 0.0013 (9) −0.0005 (13) 0.0012 (14) 0.0002 (11) −0.0123 (10)
−0.0006 (10) 0.0026 (11) 0.0017 (12) −0.0029 (3) 0.0010 (12) −0.0028 (11) −0.0027 (11) −0.0027 (11) 0.0067 (8) 0.0069 (10) −0.0012 (12) 0.0049 (13) 0.0002 (9) 0.0039 (10)
Geometric parameters (Å, º) N1—C7 N1—C2 N1—C5 C2—C8 C2—C3 C2—H2 C3—C11 C3—C12 C3—S4 S4—C5 C5—C6 C5—H5 C6—N13 C6—C7
1.372 (4) 1.449 (4) 1.466 (4) 1.537 (4) 1.574 (4) 1.0000 1.523 (4) 1.524 (4) 1.872 (3) 1.820 (3) 1.551 (4) 1.0000 1.469 (4) 1.550 (4)
C6—H6 C7—O14 C8—O10 C8—O9 C11—H11A C11—H11B C11—H11C C12—H12A C12—H12B C12—H12C N13—H13A N13—H13B N13—H13C
1.0000 1.206 (4) 1.249 (4) 1.259 (3) 0.9800 0.9800 0.9800 0.9800 0.9800 0.9800 0.98 (5) 0.96 (5) 0.90 (4)
C7—N1—C2 C7—N1—C5 C2—N1—C5 N1—C2—C8 N1—C2—C3 C8—C2—C3 N1—C2—H2 C8—C2—H2 C3—C2—H2 C11—C3—C12 C11—C3—C2 C12—C3—C2 C11—C3—S4 C12—C3—S4 C2—C3—S4
131.8 (3) 94.8 (2) 116.4 (2) 109.8 (2) 106.2 (2) 112.6 (2) 109.4 109.4 109.4 110.2 (2) 113.9 (2) 109.2 (2) 108.8 (2) 108.6 (2) 105.91 (18)
C7—C6—H6 C5—C6—H6 O14—C7—N1 O14—C7—C6 N1—C7—C6 O10—C8—O9 O10—C8—C2 O9—C8—C2 C3—C11—H11A C3—C11—H11B H11A—C11—H11B C3—C11—H11C H11A—C11—H11C H11B—C11—H11C C3—C12—H12A
110.8 110.8 133.5 (3) 135.8 (3) 90.7 (2) 126.0 (3) 116.7 (2) 117.3 (2) 109.5 109.5 109.5 109.5 109.5 109.5 109.5
Acta Cryst. (2013). C69, 1238-1242
sup-3
supplementary materials C5—S4—C3 N1—C5—C6 N1—C5—S4 C6—C5—S4 N1—C5—H5 C6—C5—H5 S4—C5—H5 N13—C6—C7 N13—C6—C5 C7—C6—C5 N13—C6—H6
94.75 (13) 87.3 (2) 104.49 (19) 119.68 (19) 113.9 113.9 113.9 117.3 (3) 120.0 (2) 84.8 (2) 110.8
C3—C12—H12B H12A—C12—H12B C3—C12—H12C H12A—C12—H12C H12B—C12—H12C C6—N13—H13A C6—N13—H13B H13A—N13—H13B C6—N13—H13C H13A—N13—H13C H13B—N13—H13C
109.5 109.5 109.5 109.5 109.5 110 (3) 111 (2) 111 (3) 107 (3) 110 (4) 107 (3)
C7—N1—C2—C8 C5—N1—C2—C8 C7—N1—C2—C3 C5—N1—C2—C3 N1—C2—C3—C11 C8—C2—C3—C11 N1—C2—C3—C12 C8—C2—C3—C12 N1—C2—C3—S4 C8—C2—C3—S4 C11—C3—S4—C5 C12—C3—S4—C5 C2—C3—S4—C5 C7—N1—C5—C6 C2—N1—C5—C6 C7—N1—C5—S4 C2—N1—C5—S4 C3—S4—C5—N1
152.5 (3) −82.8 (3) −85.4 (3) 39.3 (3) −143.5 (3) −23.2 (3) 92.9 (3) −146.8 (2) −23.9 (2) 96.4 (2) 128.3 (2) −111.7 (2) 5.46 (19) −12.3 (2) −154.3 (2) 107.7 (2) −34.3 (3) 14.16 (19)
C3—S4—C5—C6 N1—C5—C6—N13 S4—C5—C6—N13 N1—C5—C6—C7 S4—C5—C6—C7 C2—N1—C7—O14 C5—N1—C7—O14 C2—N1—C7—C6 C5—N1—C7—C6 N13—C6—C7—O14 C5—C6—C7—O14 N13—C6—C7—N1 C5—C6—C7—N1 N1—C2—C8—O10 C3—C2—C8—O10 N1—C2—C8—O9 C3—C2—C8—O9
109.4 (2) 129.5 (3) 24.3 (4) 10.88 (19) −94.3 (2) −35.0 (5) −167.3 (4) 144.6 (3) 12.3 (2) 46.8 (5) 168.0 (4) −132.8 (2) −11.6 (2) 39.8 (3) −78.4 (3) −140.7 (2) 101.2 (3)
Hydrogen-bond geometry (Å, º) D—H···A i
N13—H13A···O10 N13—H13B···O9ii N13—H13C···O9iii N13—H13C···O14iv C6—H6···O9iii
D—H
H···A
D···A
D—H···A
0.98 (4) 0.96 (5) 0.90 (4) 0.90 (4) 1.00
1.68 (4) 1.79 (5) 2.04 (4) 2.66 (4) 2.53
2.647 (3) 2.738 (4) 2.825 (3) 3.242 (3) 3.018 (3)
170 (4) 169 (4) 145 (4) 124 (3) 110
Symmetry codes: (i) −x+2, y+1/2, −z+3/2; (ii) −x+1, y+1/2, −z+3/2; (iii) −x+3/2, −y+1, z−1/2; (iv) x+1/2, −y+3/2, −z+1.
PIXEL calculations output of the three most significant intermolecular interaction energies Symmetry D···A code distancea (-x+2, y+1/2, 2.647 (3) z+3/2)
Graph set
Cmb—Cm distance
Coulombic Polarisation Dispersion Repulsion Total (kJ (kJ mol-1) (kJ mol-1) (kJ mol-1) (kJ mol-1) mol-1)
Direction
C(8)a
6.495
-183.6
b axis
Acta Cryst. (2013). C69, 1238-1242
-73
-24
102.9
-177.7
sup-4
supplementary materials (-x+3/2, 2.825 (3) y+1, z-1/2) (-x+1, y+1/2, 2.738 (4) z+3/2)
C(8)c
7.709
-135.1
-38.7
-21.1
43.5
-151.4
c axis
C(8)b
5.746
-119.5
-61.2
-40.4
92.2
-128.9
b axis
Notes: (a) donor–acceptor distance; (b) centre of mass.
PIXEL output of the lattice energy (kJ mol-1) Coulombic -379.1
Polarization -141.4
Acta Cryst. (2013). C69, 1238-1242
Dispersion -133.7
Repulsion 266.3
Total PIXEL -387.9
sup-5
A new, more precise, refinement of the 6-APA structure and a detailed structural analysis are presented herein.
Crystal Structure Communications ISSN 0108-2701
Crystal structure and packing energy calculations of (+)-6-aminopenicillanic acid Sofiane Saouane,a Gernot Buthb and Francesca P. A. Fabbiania*
Keywords: crystal structure; (+)-6-aminopenicillanic acid; zwitterion structures; semi-synthetic penicillin precursor; PIXEL calculations; computational methods.
According to the literature, the crystal structure of 6-APA was first described in 1963 by R. D. Diamand (Diamand, 1963). This is the surname that appears in the original DPhil. thesis, although in all subsequent references to this work the surname is spelled as Diamond. In the thesis, the author was unsure whether the molecule in the crystal is zwitterionic or neutral. Gałdecki & Werfel (1978) subsequently reported a preliminary refinement of 6-APA; this corresponds to the refcode AMPENA in the Cambridge Structural Database (CSD; (Version 5.34, including updates to February 2013; Allen, 2002) and the entry shows a neutral molecule. In the original paper, which corresponds to a conference abstract, no details on H atoms are given but a reference to a ‘carboxylate group’ is made; from this it can be inferred that the authors identified a zwitterion, despite the reported bond lengths of 1.329 and ˚ . No atomic coordinates for any crystal form that 1.175 A includes 6-APA have been deposited in the CSD. In their original paper, Gałdecki & Werfel report a final R value of 0.123 and list all bond lengths but do not report fractional coordinates for their structure. The reference for Gałdecki & Werfel (1978) is sometimes wrongly cited (Kirschbaum, 1986; Clayden et al., 1986). 6-APA has also been studied by computational methods. Mwangi & Garside (1996) stated that misplaced H atoms of 6-APA in the structure reported by Diamand (1963) rendered it difficult to use the data in any molecular mechanics calculations. Stroganov et al. (2003) reported that the molecular geometry obtained for their single-molecule gas-phase quantum mechanical calculations by the RHF/6-31G* method is in good agreement with the experimental data of Gałdecki & Werfel (1978), although the chemical formula in their study appears to be incorrect. In view of the above, we have redetermined the crystal structure of (I) and extended the study with lattice energy calculations.
1. Introduction
2. Experimental
(+)-6-Aminopenicillanic acid (6-APA), (I), is a molecule with antibacterial properties (Rolinson & Stevens, 1961) produced by hydrolysis of the natural benzylpenicillin known as penicillin G (Sakaguchi & Murao, 1950). Furthermore, 6-APA is the main intermediate for the production of various semisynthetic penicillins on an industrial scale (Ferreira et al., 2006). Despite the fact that 6-APA is the core of all penicillins, there is a lack of fully defined structural data in the literature.
2.1. Synthesis and crystallization
a
Abteilung Kristallographie, GZG, Georg-August-Universita¨t Go¨ttingen, Goldschmidtstrasse 1, D-37077 Go¨ttingen, Germany, and bInstitut fu¨r Synchrotronstrahlung (ISS/ANKA), Karlsruhe Institute of Technology (KIT), Hermann-vonHelmholtz-Platz 1, D-76344 Eggenstein-Leopoldshafen, Germany Correspondence e-mail: [email protected] Received 31 May 2013 Accepted 19 September 2013
The X-ray single-crystal structure of (2S,5R,6R)-6-amino-3,3dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid, commonly known as (+)-6-aminopenicillanic acid (C8H12N2O3S) and a precursor of a variety of semi-synthetic penicillins, has been determined from synchrotron data at 150 K. The structure represents an ordered zwitterion and the crystals are nonmerohedrally twinned. The crystal structure is composed of a three-dimensional network built by three charge-assisted hydrogen bonds between the ammonium and carboxylate groups. The complementary analysis of the crystal packing by the PIXEL method brings to light the nature and ranking of the energetically most stabilizing intermolecular interaction energies. In accordance with the zwitterionic nature of the structure, PIXEL lattice energy calculations confirm the predominance of the Coulombic term (379.1 kJ mol1) ahead of the polarization (141.4 kJ mol1), dispersion (133.7 kJ mol1) and repulsion (266.3 kJ mol1) contributions.
1238
# 2013 International Union of Crystallography
6-APA (Sigma) was used as supplied (purity 96%). Single crystals of (I) were obtained by slow evaporation at ambient temperature of an aqueous solution of the compound. Crystals were small in size and exhibited nonmerohedral twinning, as confirmed by screening several crystals by diffraction. Data were collected on beamline SCD at the ANKA synchrotron source at the Karlsruhe Institute of Technology using a Bruker
doi:10.1107/S0108270113025924
Acta Cryst. (2013). C69, 1238–1242
pharmaceuticals and natural products Table 1 Experimental details. Crystal data Chemical formula Mr Crystal system, space group Temperature (K) ˚) a, b, c (A ˚ 3) V (A Z Radiation type (mm1) Crystal size (mm) Data collection Diffractometer Absorption correction Tmin, Tmax No. of measured, independent and observed [I > 2(I)] reflections Rint ˚ 1) (sin /)max (A Refinement R[F 2 > 2(F 2)], wR(F 2), S No. of reflections No. of parameters H-atom treatment ˚ 3) max, min (e A Absolute structure
Absolute structure parameter
C8H12N2O3S 216.26 Orthorhombic, P212121 150 6.1954 (4), 10.4543 (7), 14.7654 (9) 956.33 (11) 4 ˚ Synchrotron, = 0.71075 A 0.32 0.08 0.06 0.04
Bruker SMART APEXII CCD areadetector diffractometer Multi-scan (TWINABS; Sheldrick, 2008b) 0.643, 0.745 2130, 2130, 1966 0.043 0.609
0.029, 0.073, 1.05 2130 142 H atoms treated by a mixture of independent and constrained refinement 0.21, 0.20 Flack x parameter determined using 678 quotients [(I+) (I)]/ [(I+) + (I)] (Parsons & Flack, 2004) 0.06 (6)
Computer programs: APEX2 (Bruker, 2012), SAINT (Bruker, 2012), SHELXS97 (Sheldrick, 2008a), SHELXL2012 (Sheldrick, 2008a), Mercury (Macrae et al., 2008), publCIF (Westrip, 2010) and SHELXLE (Hu¨bschle et al., 2011).
SMART APEX CCD diffractometer with silicon-mono˚. chromated radiation of = 0.71075 A 2.2. Refinement
The crystal structure presented in this work was refined using nonmerohedrally twinned data presenting two major domains. A third domain was not taken into account on the basis of its very small contribution and negligible overlap with the other domains. The nonmerohedral twin law corresponds to a twofold axis about the [001] direct-lattice direction. The fractional contribution of the second component was refined to 0.507 (4). The H atoms of the final refined model were normalized using the RETCIF and RETCOR modules of the CLP program (Gavezzotti, 2011), before calculating the molecular electron density using the program GAUSSIAN09W (Frisch et al., 2009) at the MP2/6-31G** level of theory. The electron-density model of the molecule was then used in the PIXELC (Gavezzotti, 2011) module for energy calculations, using a distance cut-off from the central molecule ˚. of 30 A Crystal data, data collection and structure refinement details are summarized in Table 1. The program TWINABS (Sheldrick, 2008b) was used to generate a merged reflection file in the SHELXL HKLF5 format (Sheldrick, 2008a). Data Acta Cryst. (2013). C69, 1238–1242
statistics were as follows: 6566 data (1807 unique) for domain 1 only with a mean I/(I) value of 12.9, 6634 data (1806 unique) for domain 2 only with a mean I/(I) value of 13.4, and 590 data (349 unique) for both domains with a mean I/(I) value of 21.2. The quoted Rint value (0.043) comes from scaling all single and composite reflections involving both domains. For the generation of an HKLF5 file, all observations containing domain 2 were chosen. Single reflections that also occur in composites were included for merging. H atoms on C atoms were placed geometrically and constrained to ride on their parent atoms, with Uiso(H) values assigned in the range 1.2–1.5Ueq of the parent atom. H atoms on the N atom were clearly visible in a difference Fourier map and their positions were freely refined. The chirality of the compound was known from the specifications of the manufacturer, and refinement of the Flack x parameter to a value of 0.06 (6) using Parsons’ quotients method confirmed it (Parsons & Flack, 2004). The four omitted low-angle reflections have partial contribution from domain 1 and were not properly integrated. Closer inspection of the diffraction frames revealed that these reflections were compromised by a distinctively higher background or were partially shaded by the beam stop.
3. Results and discussion As previously reported, 6-APA crystallizes in the chiral orthorhombic P212121 space group with one molecule in the asymmetric unit (Fig. 1). Difference Fourier maps clearly indicate the zwitterionic nature of the molecule. The compound was first described by Batchelor et al. (1959) as possessing definite antibacterial properties of a lower order than those of benzylpenicillin but a longer metabolization cycle. Subsequently and by biological assays, Rolinson & Stevens (1961) stated that 6-APA activity is similar to that of benzylpenicillin depending on the bacterial species. In comparison, the theoretical structure–activity relationship described by the Cohen (c) (1983) and Woodward (h) (1980)
Figure 1 The molecular structure of (I), showing the atom-numbering scheme. Displacement ellipsoids are drawn at the 50% probability level. Saouane et al.
C8H12N2O3S
1239
pharmaceuticals and natural products Table 2 ˚ , ). Hydrogen-bond geometry (A D—H A i
N13—H13A O10 N13—H13B O9ii N13—H13C O9iii N13—H13C O14iv C6—H6 O9iii
D—H
H A
D A
D—H A
0.98 (4) 0.96 (5) 0.90 (4) 0.90 (4) 1.00
1.68 (4) 1.79 (5) 2.04 (4) 2.66 (4) 2.53
2.647 (3) 2.738 (4) 2.825 (3) 3.242 (3) 3.018 (3)
170 (4) 169 (4) 145 (4) 124 (3) 110
Symmetry codes: (i) x þ 2; y þ 12; z þ 32; (ii) x þ 1; y þ 12; z þ 32; (iii) x þ 32, y þ 1; z 12; (iv) x þ 12; y þ 32; z þ 1.
Figure 2 A projection of the structure of 6-APA, (I), along the a axis. Hydrogenbond patterns showing two distinct graph-set assignments C(8)b and C(8)c are highlighted (cf. Table 2). H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
parameters is in good agreement with the biological assays. The Woodward parameter (h) defines the height of the pyramid with atoms C2, C5 and C7 as the base and N1 as the apex. The Cohen parameter (c) defines the distance between the lactam carbonyl O atom (O14) and the carboxylate C atom (C8). The values calculated using PLATON (Spek, 2009) of c =
˚ and h = 0.333 (2) A ˚ are within the reported ranges 4.475 (4) A ˚ of 3.0–4.5 and 0.25–0.5 A, respectively, for active -lactam structures (Nangia et al., 1996). In addition, the third parameter introduced by Nangia et al. (1996), concerning the torsion angle between the carboxylic acid and amide functions (O14—N1—C2—C8 herein), of 139.19 (18) also lies within the defined range (30–160 ) for biologically active -lactam structures. The crystal packing of 6-APA is mainly stabilized by hydrogen bonds. Each H atom of the ammonium group is involved in hydrogen bonds as a donor to the carboxylate groups of three neighbouring molecules (Table 2); carboxylate atom O9 is a double hydrogen-bond acceptor. Hence, the firstlevel graph-set motifs (Etter, 1990; Etter et al., 1990; Bernstein et al., 1995) are described as three C(8) chains. Two of these [C(8)a and C(8)b] run along the b axis and the third one [C(8)c] runs along the c axis (cf. Figs. 2–4). These chains combine to form a series of further chains at a secondary level, as well as large rings containing up to 38 atoms involving six molecules. Whilst there are three strong hydrogen bonds per molecule, carbonyl atom O14 forms an additional weak hydrogen bond with the ammonium N13—H13C group of a neighbouring molecule (see Table 2 for details). In addition to these classical hydrogen bonds, there is a short C—H O interaction involving the carboxylate group (Table 2). Overall, the hydrogen-bonded pattern is best described as threedimensional. For a better understanding of the intermolecular interactions in 6-APA and quantification of the interaction energies, the PIXEL method, as incorporated in the CLP program (Gavezzotti, 2011), was used for molecule–molecule and lattice energy calculations. PIXEL calculations allow the analysis of lattice and intermolecular interaction energies between pairs of molecules in terms of Coulombic, polarization, dispersion and repulsion contributions. Full details of the present PIXEL calculations are available in the Supplementary materials. The total PIXEL energy, which is the sum of these four energy contributions, gives an indication of the overall interaction energy for a particular dimer and for crystal packing, with the latter being traceable back to the sublimation energy. However, it is the breakdown of these energies into the four different terms that makes the PIXEL method a powerful tool for crystal structure analysis. The
Figure 3 A projection of the structure of 6-APA along the b axis. The C(8)c graph-set forming extended hydrogen-bonded chains of edge-to-edge molecules is highlighted. H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
1240
Saouane et al.
C8H12N2O3S
Acta Cryst. (2013). C69, 1238–1242
pharmaceuticals and natural products
Figure 5 Molecular pairs identified by PIXEL calculations as having the three most significant interaction energies; see Table 3 for details. H atoms have been omitted for clarity.
Table 4 PIXEL output of the lattice energy (kJ mol1).
Figure 4 A projection of the structure of 6-APA along the c axis. C(8)a and C(8)b graph sets form parallel motifs which run along the b axis. H atoms have been omitted for clarity. Hydrogen bonds are depicted as pale lines.
symmetry-independent dimers with the three most significant intermolecular energies are detailed in Table 3 and depicted in Fig. 5; the predominant interaction in each dimer corresponds to the three hydrogen bonds described in Table 2. At 28.5 kJ mol1, the next energetically significant interaction [corresponding to symmetry code (x + 52, y + 1, z 12)] is substantially less stabilizing. From Table 3, a directly proportional correlation between the Coulombic and the total PIXEL energies is apparent. The predominance of the Coulombic term can be explained by the presence of the zwitterionic form of 6-APA, which causes a sharp distribution of charges. As expected, the lattice energy is also dominated by the Coulombic term (Table 4). The largest interaction energy involves C(8)a hydrogen-bonded chains: this intermolecular interaction not only comprises the shortest hydrogen bond among all other dimers but there is also a contribution from a short S Ocarboxylate contact, which is not ˚ , is shorter than seen in other dimers and which, at 3.256 (2) A the sum of the van der Waals radii (Bondi, 1964). The second
Coulombic
Polarization
Dispersion
Repulsion
Total PIXEL energy
379.1
141.4
133.7
266.3
387.9
largest contribution comes from the repulsion term. This is significantly large for the C(8)a and C(8)b graph-set motifs and can be ascribed to symmetry effects: in the C(8)a and C(8)b motifs, pairs of molecules are juxtaposed, whilst in the C(8)c motif the molecules are more linearly distributed, as evidenced by the large distance between the centres of mass (cf. Fig. 5 and Table 3). On the basis of crystal growth and habit considerations, Mwangi & Garside (1996) reported that the strongest of the intermolecular interactions in 6-APA are along the h100i and h001i directions. This interpretation can be partially reconciled with the analysis of hydrogen bonds in the structure as follows: the intermix of hydrogen bonds C(8)a and C(8)b results in a C22 (6) motif at the secondary-level graph set, and it is clear that this motif builds columns of molecules along the a axis of the unit cell. It is interesting to note that the interaction with the highest intermolecular energy according to the PIXEL calculations and which runs along the b axis was also identified by Mwangi & Garside (1996) as a strong interaction, but bonding in the h010i directions was described as ‘involving only weak bonding’.
Table 3 Output from PIXEL calculations of the three most significant intermolecular interaction energies. D denotes the donor, A the acceptor and Cm the centre of mass. Symmetry code
D A distance ˚) (A
Graph set
Cm—Cm distance ˚) (A
Coulombic (kJ mol1)
Polarization (kJ mol1)
Dispersion (kJ mol1)
Repulsion (kJ mol1)
Total (kJ mol1)
Direction
(x + 2, y + 12, z + 32) (x + 32, y + 1, z 12) (x + 1, y + 12, z + 32)
2.647 (3) 2.825 (3) 2.738 (4)
C(8)a C(8)c C(8)b
6.495 7.709 5.746
183.6 135.1 119.5
73 38.7 61.2
24 21.1 40.4
102.9 43.5 92.2
177.7 151.4 128.9
b axis c axis b axis
Acta Cryst. (2013). C69, 1238–1242
Saouane et al.
C8H12N2O3S
1241
pharmaceuticals and natural products Funding from the Deutsche Forschungsgemeinschaft DFG (project No. FA 964/1-1 for FPAF and SS) and from the Georg-August-Universita¨t Go¨ttingen (Dorothea Schlo¨zer Forschungsstipendium for FPAF) is gratefully acknowledged. The award of synchrotron beamtime at KIT–ISS–Anka is also gratefully acknowledged. The authors thank the Editorial Office of the IUCr for retrieving a copy of the original paper of Gałdecki & Werfel (1978) and Ms Karen Langdon of the Bodleian Library of the University of Oxford for providing a copy of the DPhil. thesis by R. D. Diamand. The authors also thank Professor Angelo Gavezzotti (Milan) for his advice on running PIXELC calculations. Supplementary data for this paper are available from the IUCr electronic archives (Reference: FM3006). Services for accessing these data are described at the back of the journal.
References Allen, F. H. (2002). Acta Cryst. B58, 380–388. Batchelor, F. R., Doyle, F. P., Nayler, J. H. C. & Rolinson, G. N. (1959). Nature (London), 183, 257–258. Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew. Chem. Int. Ed. Engl. 34, 1555–1573. Bondi, A. (1964). J. Phys. Chem. 68, 441–451. Bruker (2012). APEX2 and SAINT. Bruker AXS Inc., Madison, Wisconsin, USA. Clayden, N. J., Dobson, C. M., Lian, L. & Marktwyman, J. (1986). J. Chem. Soc. Perkin Trans. 2, pp. 1933–1940.
1242
Saouane et al.
C8H12N2O3S
Cohen, N. C. (1983). J. Med. Chem. 26, 259–264. Diamand, R. D. (1963). DPhil thesis, University of Oxford, England. Etter, M. C. (1990). Acc. Chem. Res. 23, 120–126. Etter, M. C., MacDonald, J. C. & Bernstein, J. (1990). Acta Cryst. B46, 256– 262. Ferreira, J. S., Straathof, A. J. J., Li, X., Ottens, M., Franco, T. T. & van der Wielen, L. A. M. (2006). Ind. Eng. Chem. Res. 45, 6740–6744. Frisch, M. J. et al. (2009). GAUSSIAN09W. Gaussian Inc., Wallingford, Connecticut, USA. Gałdecki, Z. & Werfel, M. (1978). Acta Cryst. A34, S90. Gavezzotti, A. (2011). New J. Chem. 35, 1360–1368. Hu¨bschle, C. B., Sheldrick, G. M. & Dittrich, B. (2011). J. Appl. Cryst. 44, 1281– 1284. Kirschbaum, J. (1986). Analytical Profiles of Drug Substances, Vol. 15, edited by K. Florey, ch. 11, pp. 427–507. New Jersey: Academic Press Inc. Macrae, C. F., Bruno, I. J., Chisholm, J. A., Edgington, P. R., McCabe, P., Pidcock, E., Rodriguez-Monge, L., Taylor, R., van de Streek, J. & Wood, P. A. (2008). J. Appl. Cryst. 41, 466–470. Mwangi, S. M. & Garside, J. (1996). J. Cryst. Growth, 166, 1078–1083. Nangia, A., Biradha, K. & Desiraju, G. R. (1996). J. Chem. Soc. Perkin Trans. 2, pp. 943–953. Parsons, S. & Flack, H. (2004). Acta Cryst. A60, s61. Rolinson, G. N. & Stevens, S. (1961). Proc. R. Soc. London Ser. B, 154, 509– 513. Sakaguchi, K. & Murao, S. (1950). J. Agric. Chem. Soc. Jpn, 23, 411. Sheldrick, G. M. (2008a). Acta Cryst. A64, 112–122. Sheldrick, G. M. (2008b). TWINABS. Bruker AXS Inc., Madison, Wisconsin, USA Spek, A. L. (2009). Acta Cryst. D65, 148–155. Stroganov, O. V., Chilov, G. G. & Sˇvedas, V. K. (2003). J. Mol. Struct. (THEOCHEM), 631, 117–125. Westrip, S. P. (2010). J. Appl. Cryst. 43, 920–925. Woodward, R. B. (1980). Philos. Trans. R. Soc. London Ser. B, 289, 239– 250.
Acta Cryst. (2013). C69, 1238–1242
supplementary materials
supplementary materials Acta Cryst. (2013). C69, 1238-1242
[doi:10.1107/S0108270113025924]
Crystal structure and packing energy calculations of (+)-6-aminopenicillanic acid Sofiane Saouane, Gernot Buth and Francesca P. A. Fabbiani Computing details Data collection: APEX2 (Bruker, 2012); cell refinement: SAINT (Bruker, 2012); data reduction: SAINT (Bruker, 2012); program(s) used to solve structure: SHELXS97 (Sheldrick, 2008a); program(s) used to refine structure: SHELXL2012 (Sheldrick, 2008a); molecular graphics: Mercury (Macrae et al., 2008); software used to prepare material for publication: publCIF (Westrip, 2010). (2S,5R,6R)-6-Amino-3,3-dimethyl-7-oxo-4-thia-1-azabicyclo[3.2.0]heptane-2-carboxylic acid Crystal data C8H12N2O3S Mr = 216.26 Orthorhombic, P212121 a = 6.1954 (4) Å b = 10.4543 (7) Å c = 14.7654 (9) Å V = 956.33 (11) Å3 Z=4 F(000) = 456
Dx = 1.502 Mg m−3 Synchrotron radiation, λ = 0.71075 Å Cell parameters from 2730 reflections θ = 2.4–25.6° µ = 0.32 mm−1 T = 150 K Block, colourless 0.08 × 0.06 × 0.04 mm
Data collection Bruker SMART APEXII CCD area-detector diffractometer Radiation source: fine-focus sealed tube Silicon monochromator ω scans Absorption correction: multi-scan (TWINABS; Sheldrick, 2008b) Tmin = 0.643, Tmax = 0.745
2130 measured reflections 2130 independent reflections 1966 reflections with I > 2σ(I) Rint = 0.043 θmax = 25.6°, θmin = 2.4° h = −7→7 k = −12→12 l = −17→17
Refinement Refinement on F2 Least-squares matrix: full R[F2 > 2σ(F2)] = 0.029 wR(F2) = 0.073 S = 1.05 2130 reflections 142 parameters 0 restraints Primary atom site location: structure-invariant direct methods
Acta Cryst. (2013). C69, 1238-1242
Secondary atom site location: difference Fourier map Hydrogen site location: mixed H atoms treated by a mixture of independent and constrained refinement w = 1/[σ2(Fo2) + (0.034P)2 + 0.1358P] where P = (Fo2 + 2Fc2)/3 (Δ/σ)max < 0.001 Δρmax = 0.21 e Å−3 Δρmin = −0.20 e Å−3
sup-1
supplementary materials Absolute structure: Flack x parameter determined using 678 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons & Flack, 2004) Absolute structure parameter: −0.06 (6) Special details Geometry. All e.s.d.'s (except the e.s.d. in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell e.s.d.'s are taken into account individually in the estimation of e.s.d.'s in distances, angles and torsion angles; correlations between e.s.d.'s in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell e.s.d.'s is used for estimating e.s.d.'s involving l.s. planes. Refinement. Refinement of F2 against ALL reflections. From TWINABS: 6566 data (1807 unique) involve domain 1 only, mean I/σ 12.9; 6634 data (1806 unique) involve domain 2 only, mean I/σ 13.4; 590 data (349 unique) involve 2 domains, mean I/σ 21.2. The quoted Rint value (0.043) comes from scaling all single and composite reflections involving both domains. Rint = 0.039 from scaling all singles of domain 2. Rint = 0.0554 from scaling single and composite reflections of domain 1. For the generation of an HKLF5 file, all observations containing domain 2 were chosen. Data were merged in TWINABS according to point group 222; single reflections that also occur in composites were included for merging. Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2)
N1 C2 H2 C3 S4 C5 H5 C6 H6 C7 C8 O9 O10 C11 H11A H11B H11C C12 H12A H12B H12C N13 H13A H13B H13C O14
x
y
z
Uiso*/Ueq
0.7032 (4) 0.6065 (5) 0.4597 0.5848 (5) 0.79266 (11) 0.8738 (5) 1.0223 0.8143 (5) 0.9168 0.6148 (5) 0.7478 (5) 0.6525 (4) 0.9474 (3) 0.6248 (6) 0.7699 0.5175 0.6126 0.3620 (5) 0.3524 0.2516 0.3387 0.7738 (5) 0.883 (7) 0.630 (8) 0.781 (7) 0.4475 (4)
0.5375 (2) 0.5155 (2) 0.4774 0.6508 (3) 0.75459 (7) 0.6344 (2) 0.5997 0.6525 (3) 0.6046 0.5693 (3) 0.4232 (2) 0.34077 (19) 0.4377 (2) 0.6501 (3) 0.6170 0.5954 0.7374 0.7053 (3) 0.7927 0.6520 0.7063 0.7812 (2) 0.841 (4) 0.810 (4) 0.776 (4) 0.5422 (2)
0.71283 (16) 0.80083 (18) 0.7927 0.84665 (19) 0.79121 (5) 0.70872 (19) 0.7188 0.60750 (18) 0.5677 0.6307 (2) 0.85603 (18) 0.90362 (13) 0.84992 (15) 0.94842 (19) 0.9607 0.9781 0.9720 0.8262 (2) 0.8498 0.8551 0.7606 0.57153 (17) 0.594 (3) 0.586 (3) 0.511 (3) 0.59357 (15)
0.0181 (5) 0.0172 (6) 0.021* 0.0174 (6) 0.01909 (19) 0.0180 (6) 0.022* 0.0196 (6) 0.023* 0.0211 (6) 0.0170 (6) 0.0211 (5) 0.0249 (5) 0.0241 (7) 0.036* 0.036* 0.036* 0.0237 (7) 0.036* 0.036* 0.036* 0.0185 (5) 0.048 (12)* 0.044 (11)* 0.049 (12)* 0.0317 (6)
Acta Cryst. (2013). C69, 1238-1242
sup-2
supplementary materials Atomic displacement parameters (Å2)
N1 C2 C3 S4 C5 C6 C7 C8 O9 O10 C11 C12 N13 O14
U11
U22
U33
U12
U13
U23
0.0218 (12) 0.0179 (15) 0.0174 (16) 0.0185 (3) 0.0182 (14) 0.0256 (17) 0.0278 (17) 0.0204 (17) 0.0225 (12) 0.0175 (12) 0.0306 (18) 0.0193 (16) 0.0201 (14) 0.0369 (14)
0.0164 (11) 0.0160 (13) 0.0162 (13) 0.0183 (3) 0.0170 (13) 0.0181 (13) 0.0171 (14) 0.0148 (12) 0.0224 (10) 0.0265 (11) 0.0232 (14) 0.0222 (14) 0.0200 (12) 0.0318 (12)
0.0160 (11) 0.0178 (15) 0.0186 (15) 0.0205 (3) 0.0188 (14) 0.0149 (14) 0.0185 (15) 0.0158 (13) 0.0185 (10) 0.0308 (12) 0.0186 (15) 0.0295 (16) 0.0153 (12) 0.0265 (12)
−0.0014 (9) −0.0017 (11) −0.0009 (11) −0.0030 (3) −0.0004 (11) −0.0005 (13) −0.0027 (12) 0.0001 (11) −0.0028 (8) 0.0036 (8) 0.0056 (14) 0.0023 (11) −0.0017 (10) −0.0142 (11)
−0.0010 (10) −0.0014 (12) −0.0004 (11) 0.0005 (3) −0.0003 (13) 0.0020 (12) −0.0014 (13) −0.0005 (11) −0.0016 (9) 0.0013 (9) −0.0005 (13) 0.0012 (14) 0.0002 (11) −0.0123 (10)
−0.0006 (10) 0.0026 (11) 0.0017 (12) −0.0029 (3) 0.0010 (12) −0.0028 (11) −0.0027 (11) −0.0027 (11) 0.0067 (8) 0.0069 (10) −0.0012 (12) 0.0049 (13) 0.0002 (9) 0.0039 (10)
Geometric parameters (Å, º) N1—C7 N1—C2 N1—C5 C2—C8 C2—C3 C2—H2 C3—C11 C3—C12 C3—S4 S4—C5 C5—C6 C5—H5 C6—N13 C6—C7
1.372 (4) 1.449 (4) 1.466 (4) 1.537 (4) 1.574 (4) 1.0000 1.523 (4) 1.524 (4) 1.872 (3) 1.820 (3) 1.551 (4) 1.0000 1.469 (4) 1.550 (4)
C6—H6 C7—O14 C8—O10 C8—O9 C11—H11A C11—H11B C11—H11C C12—H12A C12—H12B C12—H12C N13—H13A N13—H13B N13—H13C
1.0000 1.206 (4) 1.249 (4) 1.259 (3) 0.9800 0.9800 0.9800 0.9800 0.9800 0.9800 0.98 (5) 0.96 (5) 0.90 (4)
C7—N1—C2 C7—N1—C5 C2—N1—C5 N1—C2—C8 N1—C2—C3 C8—C2—C3 N1—C2—H2 C8—C2—H2 C3—C2—H2 C11—C3—C12 C11—C3—C2 C12—C3—C2 C11—C3—S4 C12—C3—S4 C2—C3—S4
131.8 (3) 94.8 (2) 116.4 (2) 109.8 (2) 106.2 (2) 112.6 (2) 109.4 109.4 109.4 110.2 (2) 113.9 (2) 109.2 (2) 108.8 (2) 108.6 (2) 105.91 (18)
C7—C6—H6 C5—C6—H6 O14—C7—N1 O14—C7—C6 N1—C7—C6 O10—C8—O9 O10—C8—C2 O9—C8—C2 C3—C11—H11A C3—C11—H11B H11A—C11—H11B C3—C11—H11C H11A—C11—H11C H11B—C11—H11C C3—C12—H12A
110.8 110.8 133.5 (3) 135.8 (3) 90.7 (2) 126.0 (3) 116.7 (2) 117.3 (2) 109.5 109.5 109.5 109.5 109.5 109.5 109.5
Acta Cryst. (2013). C69, 1238-1242
sup-3
supplementary materials C5—S4—C3 N1—C5—C6 N1—C5—S4 C6—C5—S4 N1—C5—H5 C6—C5—H5 S4—C5—H5 N13—C6—C7 N13—C6—C5 C7—C6—C5 N13—C6—H6
94.75 (13) 87.3 (2) 104.49 (19) 119.68 (19) 113.9 113.9 113.9 117.3 (3) 120.0 (2) 84.8 (2) 110.8
C3—C12—H12B H12A—C12—H12B C3—C12—H12C H12A—C12—H12C H12B—C12—H12C C6—N13—H13A C6—N13—H13B H13A—N13—H13B C6—N13—H13C H13A—N13—H13C H13B—N13—H13C
109.5 109.5 109.5 109.5 109.5 110 (3) 111 (2) 111 (3) 107 (3) 110 (4) 107 (3)
C7—N1—C2—C8 C5—N1—C2—C8 C7—N1—C2—C3 C5—N1—C2—C3 N1—C2—C3—C11 C8—C2—C3—C11 N1—C2—C3—C12 C8—C2—C3—C12 N1—C2—C3—S4 C8—C2—C3—S4 C11—C3—S4—C5 C12—C3—S4—C5 C2—C3—S4—C5 C7—N1—C5—C6 C2—N1—C5—C6 C7—N1—C5—S4 C2—N1—C5—S4 C3—S4—C5—N1
152.5 (3) −82.8 (3) −85.4 (3) 39.3 (3) −143.5 (3) −23.2 (3) 92.9 (3) −146.8 (2) −23.9 (2) 96.4 (2) 128.3 (2) −111.7 (2) 5.46 (19) −12.3 (2) −154.3 (2) 107.7 (2) −34.3 (3) 14.16 (19)
C3—S4—C5—C6 N1—C5—C6—N13 S4—C5—C6—N13 N1—C5—C6—C7 S4—C5—C6—C7 C2—N1—C7—O14 C5—N1—C7—O14 C2—N1—C7—C6 C5—N1—C7—C6 N13—C6—C7—O14 C5—C6—C7—O14 N13—C6—C7—N1 C5—C6—C7—N1 N1—C2—C8—O10 C3—C2—C8—O10 N1—C2—C8—O9 C3—C2—C8—O9
109.4 (2) 129.5 (3) 24.3 (4) 10.88 (19) −94.3 (2) −35.0 (5) −167.3 (4) 144.6 (3) 12.3 (2) 46.8 (5) 168.0 (4) −132.8 (2) −11.6 (2) 39.8 (3) −78.4 (3) −140.7 (2) 101.2 (3)
Hydrogen-bond geometry (Å, º) D—H···A i
N13—H13A···O10 N13—H13B···O9ii N13—H13C···O9iii N13—H13C···O14iv C6—H6···O9iii
D—H
H···A
D···A
D—H···A
0.98 (4) 0.96 (5) 0.90 (4) 0.90 (4) 1.00
1.68 (4) 1.79 (5) 2.04 (4) 2.66 (4) 2.53
2.647 (3) 2.738 (4) 2.825 (3) 3.242 (3) 3.018 (3)
170 (4) 169 (4) 145 (4) 124 (3) 110
Symmetry codes: (i) −x+2, y+1/2, −z+3/2; (ii) −x+1, y+1/2, −z+3/2; (iii) −x+3/2, −y+1, z−1/2; (iv) x+1/2, −y+3/2, −z+1.
PIXEL calculations output of the three most significant intermolecular interaction energies Symmetry D···A code distancea (-x+2, y+1/2, 2.647 (3) z+3/2)
Graph set
Cmb—Cm distance
Coulombic Polarisation Dispersion Repulsion Total (kJ (kJ mol-1) (kJ mol-1) (kJ mol-1) (kJ mol-1) mol-1)
Direction
C(8)a
6.495
-183.6
b axis
Acta Cryst. (2013). C69, 1238-1242
-73
-24
102.9
-177.7
sup-4
supplementary materials (-x+3/2, 2.825 (3) y+1, z-1/2) (-x+1, y+1/2, 2.738 (4) z+3/2)
C(8)c
7.709
-135.1
-38.7
-21.1
43.5
-151.4
c axis
C(8)b
5.746
-119.5
-61.2
-40.4
92.2
-128.9
b axis
Notes: (a) donor–acceptor distance; (b) centre of mass.
PIXEL output of the lattice energy (kJ mol-1) Coulombic -379.1
Polarization -141.4
Acta Cryst. (2013). C69, 1238-1242
Dispersion -133.7
Repulsion 266.3
Total PIXEL -387.9
sup-5
