II. EXPERIMENTAL

Cefprozil monohydrate was a commercial reagent, purchased from United States Pharmacopeia (Lot # F0L131) and was used as-received. The white powder was packed into a 1.5 mm diameter Kapton capillary and rotated during the measurement at ~50 Hz. The powder pattern was measured at 295 K at beam line 11-BM (Lee et al., Reference Lee, Shu, Ramanathan, Preissner, Wang, Beno, Von Dreele, Ribaud, Kurtz, Antao, Jiao and Toby2008; Wang et al., Reference Wang, Toby, Lee, Ribaud, Antao, Kurtz, Ramanathan, Von Dreele and Beno2008) of the Advanced Photon Source at the Argonne National Laboratory using a wavelength of 0.413342 Å from 0.5 to 50° 2θ with a step size of 0.001° and a counting time of 0.1 s step⁻¹.

The pattern was indexed on an orthorhombic unit cell with a = 11.4453, b = 11.3116, c = 14.8125 Å, V = 1917.7 Å³, and Z = 4 using N-TREOR (Altomare et al., Reference Altomare, Cuocci, Giacovazzo, Moliterni, Rizzi, Corriero and Falcicchio2013). The suggested space group was P2₁2₁2₁. A reduced cell search in the Cambridge Structural Database (Groom et al., Reference Groom, Bruno, Lightfoot and Ward2016) combined with the chemistry C, H, N, O, and S yielded four hits. Among them were two hits for cefadroxil monohydrate (Shin and Cho, Reference Shin and Cho1992; JOSWAP; Seetharaman et al., Reference Seetharaman, Rajan and Srinivasan1993; JOSWAP01), which has a methyl group instead of a propenyl side chain. We saved the JOSWAP structure as a .cif file, read the .cif into Mercury (Macrae et al., Reference Macrae, Bruno, Chisholm, Edington, McCabe, Pidcock, Rodriguez-Monge, Taylor, van de Streek and Wood2008), saved it as a .mol2 file, and read the .mol2 file into Spartan ’18 (Wavefunction, 2018). We converted the methyl group at C3 into a Z-propenyl group and added the hydrogen atoms to build a cefprozil molecule. The cefprozil molecule was saved as a .mol2 file and converted into a Fenske-Hall Z-matrix using OpenBabel (O'Boyle et al., Reference O'Boyle, Banck, James, Morley, Vandermeersch and Hutchison2011). The structure was solved using FOX (Favre-Nicolin and Černý, Reference Favre-Nicolin and Černý2002) with cefprozil and an oxygen atom (water molecule) as fragments.

Rietveld refinement was carried out using GSAS II (Toby and Von Dreele, Reference Toby and Von Dreele2013). Only the 2.0–25.0° portion of the pattern was included in the refinement (d _min = 0.955 Å). All non-H-bond distances and angles were subjected to restraints, based on a Mercury/Mogul Geometry Check (Bruno et al., Reference Bruno, Cole, Kessler, Luo, Motherwell, Purkis, Smith, Taylor, Cooper, Harris and Orpen2004; Sykes et al., Reference Sykes, McCabe, Allen, Battle, Bruno and Wood2011) of the molecule. The Mogul average and standard deviation for each quantity were used as the restraint parameters. The hydrogen atoms were included in calculated positions, which were recalculated during the refinement using Materials Studio (Dassault, 2018). Positions of the active hydrogens were derived by analysis of potential hydrogen bonding patterns. The U _iso were grouped by chemical similarity. The background was modeled using a 1-term shifted Chebyshev polynomial, with a 4-term diffuse scattering function to model the Kapton capillary and any amorphous component.

The final refinement of 105 variables using 23 074 observations (23 003 data points and 71 restraints) yielded the residuals R _wp = 0.1348 and GOF = 2.79. Although the fit (Figure 2) looked reasonable (except for some profile shape errors), closer examination showed that many peaks were split and/or had shoulders. The peaks were manually located by interactive profile fitting using Jade 9.8 (MDI, 2018). The pattern was re-indexed using DICVOL14 (Louër and Boultif, Reference Louër and Boultif2014) on a primitive monoclinic unit cell with a = 11.2685, b = 11.3338, c = 14.7071 Å, β = 90.141°, V = 1878.31 Å³, and Z = 4. The space group was assumed to be P2₁, a subgroup of P2₁2₁2₁ (Figure 3).

Figure 2.

The Rietveld plot for the refinement of orthorhombic cefprozil monohydrate. The blue crosses represent the observed data points, and the green line is the calculated pattern. The cyan curve is the normalized error plot. The vertical scale has been multiplied by a factor of 20× for 2θ > 12.1°.

Figure 3.

The Rietveld plot for the refinement of monoclinic cefprozil monohydrate. The blue crosses represent the observed data points, and the green line is the calculated pattern. The cyan curve is the normalized error plot.

The structure was solved using DASH 3.4.0 (David et al., Reference David, Shankland, van de Streek, Pidcock, Motherwell and Cole2006) using two cefprozil molecules as fragments. Six of the 100 trials yielded residuals much lower than the others, and the best of these was selected to begin refinement. Refinement was carried out using GSAS II (Toby and Von Dreele, Reference Toby and Von Dreele2013). Voids in the structure were identified using Mercury 4.0.0 using a probe radius of 1.0 Å. Coordinates of the centers of the voids were obtained by manually measuring the computer screen; O93 and O94 were added at these positions. Refining the occupancies of these two water molecules yielded values of 1.18 and 0.92. O94 moved too close to other atoms, was removed from the model, and a difference Fourier map was calculated. The strongest peak in the map occurred at a reasonable position for a water molecule, so a new O94 was added at that position. The refinement used the same strategy as for the orthorhombic model. The peak profiles were described using the generalized microstrain model. The restraints contributed 6.5% to the final χ ². The hydrogen atoms were included in calculated positions, which were recalculated during the refinement using Materials Studio (Dassault, 2018). Positions of the active hydrogens were derived by analysis of potential hydrogen bonding patterns. The two molecules were constrained to have the same displacement coefficients. A common U _iso was refined for the C, N, and O atoms of the fused-ring system, another U _iso for the carboxylic acid groups, another for the central part of the molecule, another for the phenyl substituent, another for the propenyl side chain, another for the water molecules, and another for the S atom.

The final refinement of 204 variables using 23 145 observations (23 003 data points and 142 restraints) yielded the residuals R _wp = 0.0958 and GOF = 1.99. The largest peak (0.59 Å from S1) and a hole in the difference Fourier map were 0.29 and −0.23(6) eÅ⁻³, respectively. The refinement was significantly better and the monoclinic cell explained all of the split peaks and shoulders, which were not accounted for by the orthorhombic cell. The largest errors in the fit are in the shapes of some of the strong low angle peaks and may represent aspects of the microstructure not accounted for by our profile function.

A density functional geometry optimization (fixed experimental unit cell) was carried out using VASP (Kresse and Furthmüller, Reference Kresse and Furthmüller1996). The calculation used the GGA-PBE functional, a plane wave cutoff energy of 400.0 eV, and a k-point spacing of 0.5 Å⁻¹ leading to a 2 × 2 × 1 mesh. The calculation was carried out on 16 2.4 GHz processors (each with 4 Gb RAM) of a 64-processor HP Proliant DL580 Generation 7 Linux cluster at North Central College and took ~60 h. A density functional geometry optimization was also carried out using CRYSTAL14 (Dovesi et al., Reference Dovesi, Orlando, Erba, Zicovich-Wilson, Civalleri, Casassa, Maschio, Ferrabone, De La Pierre, D-Arco, Noël, Causà and Kirtman2014). The basis sets for the H, C, N, and O atoms were those of Gatti et al. (Reference Gatti, Saunders and Roetti1994), and the basis set for S was that of Peintinger et al. (Reference Peintinger, Vilela Oliveira and Bredow2013). The calculation was run on eight 2.1 GHz Xeon cores (each with 6 Gb RAM) of a 304-core Dell Linux cluster at IIT, using 8 k-points and the B3LYP functional, and took ~119 h.

The experimental infrared spectrum of cefprozil was measured using a ThermoFisher Nicolet iS5 FTIR spectrometer equipped with an attenuated total reflectance (ATR) attachment.

III. RESULTS AND DISCUSSION

Both E and Z (indicating the configuration around the double bond in the propenyl side chain) cefprozil monohydrate are commercially available and are comparable in price. Some of the patent literature discusses producing one or the other. The molecule studied here is (Z)-cefprozil monohydrate. The powder pattern for the cefprozil monohydrate sample studied here differs significantly from the powder data for the “cefprozil compound” reported by (Qing, Reference Qing2014; Zhuhai Jinhong Pharmaceutical Co. Ltd.) (Figure 4). The refined atom coordinates of cefprozil monohydrate and the coordinates from the DFT optimization are reported in the CIFs attached as Supplementary Material.

Figure 4.

Comparison of the synchrotron pattern of cefprozil monohydrate with the laboratory diffraction pattern from Qing (Reference Qing2014; Zhuhai Jinhong Pharmaceutical Co. Ltd.). The literature pattern (measured using CuKα radiation) was digitized using UN-SCAN-IT (Silk Scientific, 2013) and scaled to the synchrotron wavelength of 0.413342 Å Using Jade 9.8 (MDI, 2018).

The root-mean-square Cartesian displacement of the non-hydrogen atoms in cefprozil molecule 1 (the lower atom numbers) of the Rietveld refined, and the DFT-optimized structures are 0.301 Å (Figure 5(a)). The maximum displacement is 0.607 Å, at C22, reflects a difference in the orientation of the phenyl ring. Similar differences also occur in the propenyl side chain. For molecule 2, the RMS displacement is 0.159 Å (Figure 5(b)). The maximum difference of 0.289 Å is at C71 in the propenyl side chain; other similar differences occur in the phenyl ring. The reasonable agreement between the refined and optimized structures is evidence that the experimental structure is correct (van de Streek and Neumann, Reference van de Streek and Neumann2014). The remaining discussion concentrates on the CRYSTAL-optimized structure. The asymmetric unit (with atom numbering) is illustrated in Figure 6, and the crystal structure is presented in Figure 7.

Figure 5.

(a) Comparison of the Rietveld refined (red) and CRYSTAL14-optimized (blue) structures of cefprozil molecule 1. (b) Comparison of the Rietveld refined (red) and CRYSTAL14-optimized (blue) structures of cefprozil molecule 2.

Figure 6.

The asymmetric unit of cefprozil monohydrate, with the atom numbering. The atoms are represented by 50% probability spheroids.

Figure 7.

The crystal structure of cefprozil monohydrate, viewed down the b-axis.

The RMS difference between molecule 1 and molecule 2 (Figure 8) is 0.339 Å. The largest difference (1.313 Å) is between N17 and N63, the protonated/neutral amino groups. Removing this pair from the calculation decreases the RMS difference to 0.230 Å. The geometries of the two molecules are not very different. The RMS displacement between molecule 1 and the orthorhombic structure is 0.418 Å (Figure 9(a)). The maximum displacement is 1.256 Å, in the propenyl side chain. The orientation of the phenyl rings also differs. The RMS displacement between molecule 2 and the orthorhombic structure is 0.712 Å (Figure 9(b)). The maximum displacements are in the propenyl side chain and the phenyl ring. As expected, the general arrangement of the molecules is similar in the monoclinic and orthorhombic structures, but the higher symmetry obscures differences between the two independent molecules in the monoclinic model. In particular, in the monoclinic model, the protonated N17 is 2.698 Å from the ionized carboxylate oxygen atoms O57 and O58, and the unionized N63 is 3.031 and 3.184 Å from the oxygen atoms O11 and O12 in the intact carboxylic acid group. In the orthorhombic model, the N18⋯O11 distance is 2.886 Å, the average of the four distances in the monoclinic model.

Figure 8.

Comparison between molecule 1 and molecule 2 in cefprozil monohydrate. Molecule 1 is in green, and molecule 2 is purple.

Figure 9.

(a) Comparison between molecule 1 in the monoclinic cefprozil structure (green) and the molecule in the orthorhombic structure (orange). (b) Comparison between molecule 2 in the monoclinic cefprozil structure (purple) and the molecule in the orthorhombic structure (orange).

The crystal structure (Figure 7) exhibits some parallel stacking of phenyl rings and 3D hydrogen bond network, but no other noteworthy features. The Mulliken atomic charges calculated by CRYSTAL14 are shown in Figure 10. As expected, the most positive carbon atoms are doubly bonded to oxygen atoms. The most positive carbon atoms are those involved in hydrogen bonds. The most negative carbon atoms (C2, C27, C48, and C73) participate in C–H⋯O hydrogen bonds.

Figure 10.

Mulliken atomic charges in cefprozil, calculated by CRYSTAL14.

Although checking the structure at intermediate stages for extra symmetry using checkCIF (Strickland et al., Reference Strickland, Hoyland, McMahon, Hall and McMahon2006), Platon (Spek, Reference Spek2009), and Materials Studio (Dassault, 2018) indicated that the structure was orthorhombic, similar checks on the final DFT optimized structure confirmed that it was monoclinic. CRYSTAL14 indicates that the orthorhombic structure is 0.4 kcal mol⁻¹ molecule⁻¹ lower in energy than the monoclinic structure. VASP indicates that the monoclinic structure is more stable by 1.3 kcal mol⁻¹ molecule⁻¹. The structures are similar in energy; these differences are comparable to the accuracy of such calculations. The displacement coefficients of the atoms in the propenyl side chain are significantly larger than those in the rest of the molecule, indicating greater thermal motion and/or disorder.

All of the bond distances, almost all of the bond angles, and all of the torsion angles fall within the normal ranges indicated by a Mercury Mogul Geometry check (Macrae et al., Reference Macrae, Bruno, Chisholm, Edington, McCabe, Pidcock, Rodriguez-Monge, Taylor, van de Streek and Wood2008). The C3–C2–S1 and C49–C48–S47 angles of 118.8° and 119.3° [average = 111.4(21), Z-scores = 3.4 and 3.6] are flagged as unusual. This angle is a part of a fused ring system, so it is not unreasonable even though it is unusual. The C27–C26–C25 and C73–C72–C71 angles of 134.4° and 132.2° [average = 125.3(19), Z-scores = 4.7 and 3.5] are flagged as unusual. These are in the propenyl side chains, whose atoms have the largest displacement coefficients and thus may be located less accurately. The C71–C49–C50 angle of 119.3° is [average = 122.2(7), Z-score = 4.0] also flagged as unusual; the Z-score is the result of the exceptionally low standard uncertainty on the average.

Quantum chemical geometry optimizations (DFT/B3LYP/6-31G*/water) using Spartan ‘18 (Wavefunction, 2018) indicated that the observed conformations of molecules 1 and 2 are 9.9 and 7.2 kcal/mol higher in energy than their respective local minima. The global minimum energy conformation (found by molecular mechanics conformational search) was more compact than the observed conformations and contained intramolecular hydrogen bonds. Thus these intermolecular interactions are important in determining the solid-state conformation.

Analysis of the contributions to the total crystal energy using the Forcite module of Materials Studio (Dassault, 2018) suggests that angle distortion terms are significant in the intramolecular deformation energy, as might be expected from a fused ring system. The intermolecular energy contains significant contributions from van der Waals attraction, which in this force-field-based analysis include hydrogen bonds and electrostatic repulsion. The hydrogen bonds are better analyzed using the results of the DFT calculation.

The hydrogen bonds are prominent in the crystal structure (Table I). The water molecule O93 forms a strong O–H96⋯N hydrogen bond to the amine N63. The other hydrogen (H95) of this water molecule acts as a hydrogen donor to S1. The other water molecule O94 forms an O–H97⋯O to the ketone oxygen O15. The other hydrogen (H98) of this water molecule forms bifurcated hydrogen bonds to O57 of the ionized carboxylate and the sulfur S47.

TABLE I.

Hydrogen bonds (CRYSTAL14) in cefprozil monohydrate.

^a Intramolecular.

Both DFT calculations moved the carboxylic acids proton H78 to the amino group N17. The structure thus contains one ion pair and one pair of neutral molecules. The ammonium ion N17 forms N–H⋯O hydrogen bonds to two different oxygens of ionized carboxylate groups and one to the ketone O55. The cation/anion hydrogen bonds form a chain that runs roughly along the b-axis. The local environments of the two carboxyl groups are very different, making ionization of only one of them plausible. Using the PHONON module (Parlinski et al., Reference Parlinski, Li and Kawazoe1997) of MedeA (Materials Design, 2016), we calculated the infrared spectrum of cefprozil monohydrate (Figure 11). The high frequency region of the spectrum pattern has the expected broad hydrogen bond feature. The general pattern in the low frequency region shows a similar structure in the two spectra but there are differences, which we are not going to explore in detail. We will concentrate on the high frequency X-H stretching region. The highest frequency bands are higher in the observed spectrum than in the experimental.

Figure 11.

Observed (black) and calculated (red; PHONON) infrared spectra of cefprozil monohydrate.

Analyzing the phonon density of states, we could determine which of the hydrogen atoms contributes to the different high frequency bands. An unusual feature of the spectra are the two bands at 2584 and 2699 cm⁻¹. These involve H35 and H78, which make the hydrogen bonds to O57 and O58, the ionized carboxyl group. The third hydrogen on N17, which is H36, contributes to the 3281 cm⁻¹ band, which donates to the ketone oxygen O55. It is reasonable that these strong hydrogen bonds contribute to the lower frequency stretches. The observed intensities of these bands are much lower than observed, so perhaps the suggested hydrogen transfer is only partial. H81 and H82, the hydrogens of the neutral amine group, N63, contribute to the two highest frequency bands at 3405 and 3335 cm⁻¹ in the calculated spectrum; they are observed at higher frequencies. The water molecule O93–H95–H96 contributes to 2928, 3030, and 3337 cm⁻¹ bands, while water molecule O94–H97–H98 yields 3113, 3196, 3332, and 3412 cm⁻¹ bands. The secondary amines and hydroxyl groups contribute to several of the high frequency bands.

The neutral amine N63 acts as a donor to the ketone O12 and the hydroxyl group O11. The nonionized carboxylic acid O11 acts as a donor to the hydroxyl group O24. Both hydroxyl groups O24 and O70 act as hydrogen bond donors to the water molecules. Both the secondary amino groups N13 and N59 also act as donors to the water molecules. The energies of the O–H⋯O hydrogen bonds were calculated by the correlation of Rammohan and Kaduk (Reference Rammohan and Kaduk2018), and the energies of the N–H⋯O were calculated according to the correlation of Wheatley and Kaduk (Reference Wheatley and Kaduk2019). There are a surprisingly large number of C–H⋯O hydrogen bonds as well as a C–H⋯S hydrogen bond. Although individually these are probably weak, the number contributes to the crystalline energy.

The volume enclosed by the Hirshfeld surface (Figure 12; Hirshfeld, Reference Hirshfeld1977; Turner, et al., Reference Turner, McKinnon, Wolff, Grimwood, Spackman, Jayatilaka and Spackman2017) is 928.00 Å³, 98.66% of 1/2 the unit cell volume. The molecules are thus not tightly packed. All of the significant close contacts (red in Figure 12) involve the hydrogen bonds.

Figure 12.

The Hirshfeld surface of cefprozil monohydrate. Intermolecular contacts longer than the sums of the van der Waals radii are colored blue, and contacts shorter than the sums of the radii are colored red. Contacts equal to the sums of radii are white.

The Bravais–Friedel–Donnay–Harker (Bravais, Reference Bravais1866; Friedel, Reference Friedel1907; Donnay and Harker, Reference Donnay and Harker1937) morphology suggests that we might expect platy morphology for cefprozil monohydrate, with {001} as the major faces. A second-order spherical harmonic preferred orientation model was included in the refinement; the texture index was 1.032, indicating that preferred orientation was small in this rotated capillary specimen. The powder pattern of orthorhombic cefprozil monohydrate from this synchrotron data set is included in the Powder Diffraction File™ as entry 00-066-1615 and will be replaced by the new powder pattern based on the monoclinic structure determined in this study, and has been submitted to ICDD.