I. INTRODUCTION
Atazanavir, as the sulfate Reyataz®, is an HIV-1 protease inhibitor. Atazanavir is typically administered with other antiretrovirals, for example, ritonavir, resulting in a synergistic effect boosting the efficacy of each medication. As the free base, the IUPAC name for atazanavir (CAS Registry number 198904-31-3) is methyl N-[(2S)-1-[2-[(2S,3S)-2-hydroxy-3-[[(2S)-2-(methoxycarbonylamino)-3,3-dimethylbutanoyl]amino]-4-phenylbutyl]-2-[(4-pyridin-2-ylphenyl)methyl]hydrazinyl]-3,3-dimethyl-1-oxobutan-2-yl]carbamate. A two-dimensional molecular diagram for Atazanavir is shown in Figure 1.
Figure 1. The molecular structure of the atazanavir molecule.
A reduced cell search of the Cambridge Structural Database (Groom et al., Reference Groom, Bruno, Lightfoot and Ward2016), increasing the default tolerance from 1.5% to 2.0% of the longest cell dimension, yielded 10 hits, but no structure for atazanavir derivatives. A name search on “atazanavir” yielded a single-crystal structure of atazanavir at 150 K (Patel et al., Reference Patel, Luthra, Shamblin, Arora, Krzyzaniak and Taylor2018; CSD Refcode LISTEP; Pfizer; published during the course of this work) and two structures of atazanavir bisulfate (Kim et al., Reference Kim, Lotz, Malley, Gougoutas, Davidovich and Srivastava2011; CSD Refcodes LUQRAS and LUQREW; Bristol-Myers Squibb). The Kim et al. patent application reports the crystal structure and powder data for Form A of atazanavir bisulfate, a process for preparing Form C, and the structure of a triethanol solvate Form E3. Crystallographic data for these forms are also reported in Kim et al. (Reference Kim, Lotz, Malley, Gougoutas, Davidovich and Srivastava2005).
This work was carried out as part of a project (Kaduk et al., Reference Kaduk, Crowder, Zhong, Fawcett and Suchomel2014) to determine the crystal structures of large-volume commercial pharmaceuticals at room temperature for phase identification applications and include high-quality powder diffraction data for these pharmaceuticals in the Powder Diffraction File (Gates-Rector and Blanton, Reference Gates-Rector and Blanton2019).
II. EXPERIMENTAL
The sample labeled as “atazanavir sulfate” was a commercial reagent, purchased from Arking Pharma Scientific, Inc. (Lot #Arki-24306), 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 beamline 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 Argonne National Laboratory using a wavelength of 0.413891 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 a primitive monoclinic unit cell with a = 15.331, b = 5.902, c = 21.562 Å, β = 96.3°, V = 1939.3 Å3, and Z = 2 using Jade 9.8 (MDI, 2018). The suggested space group was P21, which was confirmed by the successful solution and refinement of the structure. After many attempts to solve the structure as a sulfate derivative, the S content was checked by X-ray Photoelectron Spectroscopy (XPS). A surface concentration of a few hundred ppm S was detected, consistent with the formulation as a free base. The unit cell volume of atazanavir bisulfate Form A is 2150(2) Å3. The unit cell volume observed here is 9.3% smaller, so the compound was unlikely to be a bisulfate.
Once the absence of S was realized, the atazanavir molecule was extracted from the crystal structure of Form E3 of the bisulfate using Materials Studio (Dassault, 2019) and converted to 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).
Initial Rietveld refinement of the structure led to a wR ~17%, an uncomfortably high value. The root-mean-square (rms) Cartesian displacement between the refined structure and a DFT-optimized structure (carried out using the methods described below for the correct structure) was 0.463 Å, outside the range for correct structures (van de Streek and Neumann, Reference van de Streek and Neumann2014). Fortunately, by this time, the LISTEP structure had appeared, and we could understand what was wrong with this initial structure from our study. The rms Cartesian displacement between the two structures is 0.942 Å (Figure 2). The differences are spread through the molecules but are largest in the center of the molecule and in the orientation of one of the methyl ester groups on the periphery. The differences are a good reminder that both Rietveld refinement (least squares) and a DFT optimization locate local minima and do not guarantee to find the global optimum structure.
Figure 2. (Color online) Comparison of the false-minimum/wrong (red) structure of atazanavir to the correct structure (orange) from the CSD entry LISTEP. The rms Cartesian displacement is 0.942 Å.
The LISTEP structure has a 68/32% disorder in one of the side chains (N6–C37–O6–O7–C38). Since an ordered model is necessary for a DFT calculation, only the major orientation of this chain was included in the Rietveld refinement. Refinement was carried out using GSAS-II (Toby and Von Dreele, Reference Toby and Von Dreele2013). Only the 1.5–22.0° portion of the pattern was included in the refinement (d min = 1.084 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 results were exported to a csv file. The Mogul average and standard deviation for each quantity were used as the restraint parameters and were incorporated using the new feature Restraints/Edit Restraints/Add MOGUL Restraints, which reads the bond distance and angle restraints from the csv file. The restraints contributed 2.4% to the final χ 2. The hydrogen atoms were included in calculated positions, which were recalculated during the refinement using Materials Studio (Dassault, 2019). Positions of the active hydrogens were derived by analysis of potential hydrogen bonding patterns. The U iso of the heavy atoms were fixed at 0.06 Å2, and those of the H atoms at 0.08 Å2. The background was modeled using a 3-term shifted Chebyshev polynomial and a 6-term diffuse scattering function to describe the Kapton capillary and any amorphous component.
The final refinement of 180 variables using 20 501 observations and 127 restraints yielded the residuals Rwp = 0.09408 and GOF = 1.15. The largest peak (0.23 Å from C5) and hole (1.12 Å from C9) in the difference Fourier map were 0.49 and −0.33(8) eÅ−3, respectively. The Rietveld plot is included as Figure 3. The largest errors in the fit are in the positions of some of the low-angle peaks and may indicate subtle changes in the beam during the measurement.
Figure 3. (Color online) The Rietveld plot for the refinement of atazanavir. 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 > 11.5°.
A density functional geometry optimization (fixed experimental unit cell) was carried out using VASP (Kresse and Furthmüller, Reference Kresse and Furthmüller1996) through the MedeA graphical interface (Materials Design, 2016). 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. 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 1 × 3 × 1 mesh, and took ~11 h. A single-point calculation on the VASP-optimized structure was 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). 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 ~82 min.
III. RESULTS AND DISCUSSION
Despite being labeled as “atazanavir sulfate”, this sample consists of atazanavir free base. The refined atom coordinates of atazanavir and the coordinates from the DFT optimization are reported in the CIFs attached in Supplementary Material. As would be expected from a Rietveld-refined structure derived from a single-crystal determination, the rms Cartesian displacement of the non-hydrogen atoms in the Rietveld-refined and DFT-optimized structures is 0.077 Å (Figure 4), well within the range expected for correct structures (van de Streek and Neumann, Reference van de Streek and Neumann2014). There is no sign of the disorder noted in the single-crystal structure. Perhaps the samples actually differ, or powder diffraction is not sensitive enough to detect the disorder. This discussion concentrates on the VASP-optimized structure, with the CRYSTAL14 fixed-point calculation. The asymmetric unit (with atom numbering; the LISTEP numbering has been retained) is illustrated in Figure 5, and the crystal structure is presented in Figure 6.
Figure 4. (Color online) Comparison of the Rietveld-refined (red) and VASP-optimized (blue) structures of atazanavir. The rms Cartesian displacement is 0.077 Å.
Figure 5. (Color online) The asymmetric unit of atazanavir, with the atom numbering. The atoms are represented by 50% probability spheroids.
Figure 6. (Color online) The crystal structure of atazanavir, viewed down the b-axis.
Both the profile residuals (0.09408 vs. 0.17246) and the energies (−1.8 kcal mol−1) clearly show that the LISTEP structure is preferred over the initially obtained false-minimum structure (Figure 2). Since the structure solution involves 28 degrees of freedom, we should not be surprised that Monte Carlo simulated annealing might fall into a false minimum. The structure consists of an array of extended-conformation molecules parallel to the ac-plane (Figure 6). The unit cell expansion between 150 and 295 K is not especially anisotropic (Table I). Visual inspection suggests that aromatic–aromatic ring interactions and t-butyl–t-butyl interactions are significant, but this is not supported by the population analysis. All of the bond distances, bond angles, and torsion angles in atazanavir 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).
Table I. Lattice parameters of atazanavir free base. Space group P21.
Quantum chemical geometry optimization of the atazanavir molecule (DFT/B3LYP/6-31G*/water) using Spartan ‘18 (Wavefunction, 2018) indicated that the observed solid-state conformation is 9.9 kcal mol−1 higher in energy than the local minimum. The geometry differences are spread throughout the molecule (Figure 7) and lie mainly in the orientations of the arms on the outside of the molecule. Molecular mechanics conformational analysis indicated that the observed conformation is 8.1 kcal mol−1 higher in energy than the global minimum energy conformation (Figure 8). The two conformations differ considerably (rms Δ = 4.091 Å), suggesting that intermolecular interactions play an important role in determining the solid-state conformation.
Figure 7. (Color online) Comparison of the observed solid-state conformation of atazanavir (red) to the local minimum energy conformation of an isolated molecule (green). The rms Cartesian displacement is 1.097 Å.
Figure 8. (Color online) Comparison of the observed solid-state conformation of atazanavir (red) to the global minimum energy conformation of an isolated molecule (purple).
Analysis of the contributions to the total crystal energy using the Forcite module of Materials Studio (Dassault, 2019) suggests that the intramolecular deformation energy is small, as might be expected for a flexible molecule. The intermolecular energy is dominated by electrostatic attractions, which in this force field-based analysis include cation coordination and hydrogen bonds. The hydrogen bonds are better analyzed using the results of the DFT calculation.
Although the atazanavir molecule contains only four classical hydrogen bond donors, hydrogen bonding is, surprisingly, important to the crystal energy (Table II). Both intra- and intermolecular hydrogen bonds are significant. The hydroxyl group O2–H48 forms bifurcated intramolecular hydrogen bonds to the ketone oxygen atom O3 and the amide nitrogen N3. The energy of the O–H⋯O hydrogen bond was calculated from the correlation of Rammohan and Kaduk (Reference Rammohan and Kaduk2018). The amide nitrogen N4 acts as a donor to the hydroxyl group O2, and the amide nitrogen N1 acts as a donor to the carbonyl oxygen O1. The principal classical hydrogen bonds are illustrated in Figure 9. The energies of the N–H⋯O hydrogen bonds were calculated using the correlation of Wheatley and Kaduk (Reference Wheatley and Kaduk2019). The amide nitrogen N2 acts as a donor to the amide nitrogen N1. Several methyl, methylene, methyne, and phenyl hydrogens participate in hydrogen bonds to carbonyl oxygens, the amide nitrogen N2, and the pyridine nitrogen N5.
Figure 9. (Color online) The principal classical hydrogen bonds in atazanavir.
Table II. Hydrogen bonds (CRYSTAL14) in atazanavir.
a Intramolecular.
The volume enclosed by the Hirshfeld surface (Figure 10; Hirshfeld, Reference Hirshfeld1977; Turner, et al., Reference Turner, McKinnon, Wolff, Grimwood, Spackman, Jayatilaka and Spackman2017) is 958.39 Å3, 98.75% of one-fourth the unit cell volume. The molecules are, thus, not tightly packed. All of the significant close contacts (red in Figure 10) involve the hydrogen bonds. The volume/non-hydrogen atom is 19.0 Å3.
Figure 10. (Color online) The Hirshfeld surface of atazanavir. 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 needle-like morphology for atazanavir, with 
$\langle 010 \rangle$ as the needle axis, as might be expected from the anisotropy of the lattice parameters. A second-order spherical harmonic model for preferred orientation was incorporated into the refinement. The texture index was 1.004, indicating that preferred orientation was not significant in this rotated capillary specimen. The powder pattern of atazanavir from this synchrotron data set is included in the Powder Diffraction File™ as entry 00-065-1426.
IV. DEPOSITED DATA
The Crystallographic Information Framework (CIF) files containing the results of the Rietveld refinement (including the raw data) and the DFT geometry optimization were deposited with the ICDD. The data can be requested at [email protected].
ACKNOWLEDGEMENTS
The use of the Advanced Photon Source at Argonne National Laboratory was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-06CH11357. This work was partially supported by the International Centre for Diffraction Data. We thank Lynn Ribaud and Saul Lapidus for their assistance in the data collection, Andrey Rogachev for the use of computing resources at IIT, and Gerry W. Zajac of Ineos Technologies for the XPS analysis.
CONFLICTS OF INTEREST
The authors have no conflicts of interest to declare.
