Crystal engineering of exemestane to obtain a co-crystal with enhanced urease inhibition activity

Co-crystallization is a phenomenon widely employed to enhance the physio-chemical and biological properties of active pharmaceutical ingredients (APIs). Exemestane, or 6-methyl­ideneandrosta-1,4-diene-3,17-dione, is an anabolic steroid used as an irreversible steroidal aromatase inhibitor, which is in clinical use to treat breast cancer. The present study deals with the synthesis of co-crystals of exemestane with thio­urea by liquid-assisted grinding. The purity and homogeneity of the exemestane–thio­urea (1:1) co-crystal were confirmed by single-crystal X-ray diffraction followed by thermal stability analysis on the basis of differential scanning calorimetry and thermogravimetric analysis. Detailed geometric analysis of the co-crystal demonstrated that a 1:1 co-crystal stoichiometry is sustained by N—H⋯O hydrogen bonding between the amine (NH₂) groups of thio­urea and the carbonyl group of exemestane. The synthesized co-crystal exhibited potent urease inhibition activity in vitro (IC₅₀ = 3.86 ± 0.31 µg ml⁻¹) compared with the API (exemestane), which was found to be inactive, and the co-former (thio­urea) (IC₅₀ = 21.0 ± 1.25 µg ml⁻¹), which is also an established tested standard for urease inhibition assays in vitro. The promising results of the present study highlight the significance of co-crystallization as a crystal engineering tool to improve the efficacy of pharmaceutical ingredients. Furthermore, the role of various hydrogen bonds in the crystal stability is successfully analysed quantitatively using Hirshfeld surface analysis.

1. Introduction

As defined by the FDA regulatory classification of pharmaceutical co-crystal guidance for industry, `co-crystals are crystalline materials composed of two or more different molecules, typically an active pharmaceutical ingredient (API) and a co-crystal former in the same crystal structure'. The approval of co-crystals by the FDA as new drug applicants has opened up new vistas in both the academic and the industrial research sectors (Gadade & Pekamwar, 2016), resulting in a significant increase in research in the field of co-crystals. It has already been established that the physical and biological properties of an API can be altered by co-crystallization (Caira et al., 2012; Ghosh & Reddy, 2012; Issa et al., 2012). The literature has disclosed many examples of fine tuning the biological and physical properties of many bioactive compounds and key pharmaceutical ingredients with suitable co-formers by using a number of crystal engineering approaches (Chen et al., 2014; Charron et al., 2013; Castro et al., 2011). Examples include the co-crystallization of naturally occurring anti-leishmanial seselin (Hussain et al., 2018), anti-tumour drug temozolomide (Sanphui et al., 2013), antibiotic nalidixic acid (Gangavaram et al., 2012) and well known quercetin (Smith et al., 2011). Co-crystals are also known to have uses in thr agrochemical industry, as well as paint, electronic and optical materials (Blagden et al., 2008; Papaefstathiou et al., 2004; Sokolov et al., 2006). The stabilization of co-crystals is known to be influenced by classical hydrogen bonding and π–π stacking. Non-covalent interactions (mainly hydrogen bonding) are involved in various biological systems due to their dynamic nature. The famous `lock-and-key' model proposed by Emil Fischer (Fischer, 1894) for enzyme and substrate interactions has two main features: supramolecular chemistry and molecular recognition (Desiraju, 2001, 2000; Lehn, 1995). Therefore, it is worth studying both qualitatively and quantitatively the various interactions which contribute to crystal stability. Hirshfeld surface analysis (Spackman & Jayatilaka, 2009) is an effective approach for the qualitative and quantitative analysis of non-covalent interactions in crystal packing.

Enzyme inhibition is an important area of biomedical research contributing to the treatment of a wide range of disorders such as ulcers, cancer, inflammation, cardiovascular and central nervous system problems, and many infectious diseases. The binding of specific inhibitors to block the activity of an enzyme is the key to searching for potential drugs and to treating several physiological conditions associated with enzymes (Ramsay & Tipton, 2017). Urease is a pharmaceutically important and unique enzyme found in a wide variety of organisms and is known to catalyze the hydrolysis of urea into ammonia and carbonic acid (Amtul et al., 2002; Mobley et al., 1995). Urease also plays a key role in the formation of gastric ulcers (Mobley, 1996). The Helicobacter pylori infection (which can cause peptic ulcers and gastritis) is considered a worldwide problem with a high morbidity and mortality rate (Taha et al., 2015). It is estimated that about 50% of the world's population is infected with H. pylori (Rego et al., 2018). The urease-producing pathogen (H. pylori) is reported to survive in the highly acidic environment of the stomach because the urease-catalyzed hydrolysis of urea produces ammonia, which forms a protective shield around the pathogen to protect it from the acidic conditions, and is the actual damaging factor for the host tissues, resulting in gastritis and gastro-duodenal ulcers (Rego et al., 2018). Urea is a metabolic nitro­genous waste that is mainly produced in the liver, and the blood stream is responsible for carrying it towards the kidneys for excretion in urine. However, 20–25% of the urea remains in the body and is utilized as a substrate by the urolytic bacteria to produce high quantities of ammonia. During hepatic failure, the removal of toxic substances from the blood stream is affected and, as a result, the accumulation of toxic nitro­genous waste, including ammonia in the brain, is responsible for hepatic coma. The role of urease-producing pathogens is also reported in the progress of urinary catheter obstruction and in the production of kidney stones (urolithiasis) (Rego et al., 2018; Maroney & Ciurli, 2014). Therefore, the search for possible inhibitors of this key enzyme is an important therapeutic need, as the currently available inhibitor drugs are reported to have many side effects (Kosikowska & Berlicki, 2011).

Exemestane (EX), or 6-methyl­ideneandrosta-1,4-diene-3,17-dione, is an anabolic steroid, used as an irreversible steroidal aromatase inhibitor. It is used clinically to treat breast cancer (Shiraki et al., 2008). To the best of our knowledge, only one co-crystal of EX with maleic acid has been synthesized and reported by Shiraki and co-workers in 2008 in order to study the change in dissolution rate of poorly soluble APIs, i.e. EX.

By considering all of the above facts, the present study was conducted to observe changes in the biological activities of synthesized co-crystals of EX. The objective was to assimilate the co-former (thio­urea) in the molecular structure of the API (EX) to alter the crystal properties and therefore contribute to a change in the biological properties. Unfortunately, the synthesized exemestane:thio­urea (EX:TH) (1:1) co-crystal was found to be inactive for anti-cancer activity evaluated against the breast cancer cell line. The discouraging results regarding anti-cancer activity prompted us to evaluate the synthesized co-crystal against urease inhibition activity as our co-former (thio­urea) is already reported to be a tested standard in urease inhibition assays (Kanwal et al., 2018). Synthesized co-crystals were found to be several folds more active than tested standard thio­urea. Therefore, as per the approved guidelines of the FDA, the synthesized co-crystal is a potential candidate as a urease inhibitor for further studies. The detailed structural features of the synthesized co-crystal were established on the basis of single-crystal X-ray diffraction (SCXRD) studies followed by qualitative and quantitative analyses of intermolecular interactions contributing to the co-crystal stability by Hirshfeld surface analysis. Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) were also carried out to observe the thermal stability of the synthesized co-crystal.

2. Experimental

2.1. Data collection and refinement

Crystals of suitable sizes were mounted on a Bruker SMART APEX II X-ray diffractometer for data collection and structure determination. X-ray diffraction data were collected using Mo Kα radiation (λ = 0.71073 Å). The program SAINT (Bruker, 2016) was used for data integration and reduction. Direct methods followed by Fourier transformation by employing full-matrix least-squares calculations were carried out to solve the structure by using SHELXTL and SHELXL97 (Sheldrick, 1997) in the case of EX; the APEX3 Suite combined with SHELXL2014/7 and SHELXL2016/6 (Sheldrick, 2015) was used in the case of the co-crystal. Intermolecular interactions were calculated using PLATON (Spek, 2009). The program ORTEP3 (Farrugia, 1997) was utilized to generate a structural representation. Mercury 2.0 was used to create molecular graphics of the interactions (Macrae et al., 2008). Crystallographic data, experimental details and structure-refinement parameters are summarized in Table 1.

Table 1
Experimental details

	Exemestane	Co-crystal
Empirical formula	C20H24O2	C21H28N2O2S
Formula weight	296.39	372.51
Temperature (K)	273 (2)	293 (2)
Wavelength (Å)	0.71073	0.71073
Crystal system	Orthorhombic	Monoclinic
Space group	P212121	P21
Unit-cell dimensions (Å, °)	a = 9.959 (4), α = 90	a = 10.3797 (14), α = 90
	b = 11.731 (4), β = 90	b = 8.3706 (11), β = 105.568 (3)
	c = 12.842 (5), γ = 90	c = 10.7275 (14), γ = 90
Volume (Å^3)	1500.2 (10)	897.9 (2)
Z	4	2
Density (calculated) (Mg m^−3)	1.312	1.378
Absorption coefficient (mm^−1)	0.083	0.199
F(000)	640	400
Crystal size	0.33 × 0.32 × 0.12	0.17 × 0.15 × 0.07
θ range (°)	2.352–28.365	1.971–28.390
Index ranges	−13 ≤ h ≤ 13, −15 ≤ k ≤ 14, −17 ≤ l ≤ 15	−13 ≤ h ≤ 13, −6 ≤ k ≤ 11, −14 ≤ l ≤ 14
Reflections collected	10477	6563
Independent reflections	3736 [R(int) = 0.0693]	3315
Completeness to theta	100.0%	25.242°
Refinement method	Full-matrix least-squares on F^2	Full-matrix least-squares on F^2
Data/restraints/parameters	3736/0/203	3315/1/244
Goodness-of-fit on F^2	0.968	1.034
Final R indices [I > 2σ(I)]	R1 = 0.0551, wR2 = 0.1005	R1 = 0.0472, wR2 = 0.0926
R indices (all data)	R1 = 0.1516, wR2 = 0.1329	R1 = 0.0733, wR2 = 0.1048
Absolute structure parameter	1 (3)	0.03 (14)
Extinction coefficient	0.012 (2)	NA
Largest difference, peak and hole (eÅ^−3)	0.121 and −0.116	0.173 and −0.199
Computer programs: SAINT, XPREP (Bruker, 2016), SADABS (Sheldrick, 2014), SHELXTL (Sheldrick, 2008), SHELXL2016/6 (Sheldrick, 2015).

2.2. Hirshfeld surface analysis

Qualitative and quantitative analyses of various hydrogen bonds contributing to the co-crystal stability were carried out with the aid of Hirshfeld surface analysis, utilizing the software package Crystal Explorer. The two-dimensional Hirshfeld surface for the API is generated using d_norm parameters in the range −0.580–1.325, with a shape index of −0.996–0.994 and a curvedness of −4.029–0.281. Whereas for the co-former, the d_norm surface range is −0.612–1.059, the shape index is −0.998–0.996 and the curvedness is −4.000–0.40. The electrostatic potential surface has also been studied using the TONTO package incorporated in Crystal Explorer (Wolff et al., 2012).

2.3. Synthesis and crystallization

Commercially available solvents and chemicals were utilized without any further purification. Thio­urea (TH) was purchased from Merck, Germany (index No. 612–082-0000). A small pestle and mortar was used for grinding. The commercially available drug Aromasin (1.411 g) was ground manually by using a pestle and mortar, and extracted in 100 ml di­chloro­methane (DCM) to obtain the API exemestane (0.350 g) after solvent evaporation under reduced pressure. The obtained API was further allowed to re-crystallize by dissolution in a 1:1 mixture of DCM and methanol after overnight evaporation. Manual grinding of crystallized EX (10 mg) and TH (20 mg) in a 1:1 stoichiometric ratio was carried out for 3 h, followed by methanol (1 ml) assisted grinding for 1 h. The resulting slurry was added to the DCM and methanol (1:1) and left overnight at room temperature (Fig. 1). The shiny co-crystals obtained (8 mg) were found to be suitable for SCXRD analysis.

Figure 1 Schematic representation of the co-crystal synthesis.

2.4. Evaluation of biological activity

Urease enzyme inhibition activity in vitro against the urease enzyme from Jack bean (Canavalia ensiformis, EC 3.5.1.5) was evaluated by adopting the same procedure as that described by Kanwal and co-workers (Kanwal et al., 2018).

In vitro cytotoxicity and anti-cancer activity of the API and synthesized co-crystals were evaluated by using a standard MTT colorimetric assay as per the protocol adopted and explained by Park et al. (2008).

3. Result and discussions

SCXRD analysis revealed that EX crystallizes in the ortho­rhombic space group P2₁2₁2₁ (Görlitzer et al., 2006), whereas the co-crystal crystallizes in the monoclinic space group P2₁; hence EX and TH are independent asymmetric moieties in the unit cell with Z = 2, as depicted in Table 1. Structural analysis revealed that EX and its co-crystal are both composed of four transfused rings (A–D). As a result of the C1=C2 and C4=C5 olefinic bonds in conjugation with the C3 carbonyl, ring A is planar in nature, with a maximum deviation of 0.05 (1) and 0.03 (3) Å from the root-mean-square plane (r.m.s.) for the C3 atom in EX and the co-crystal, respectively. Transfused rings B and C were found to exist in chair conformations, with puckering parameters (Q = 0.506 Å, θ = 169.13°, φ = 306.26°) and (Q = 0.549 Å, θ = 4.88°, φ = 142.90°), and (Q = 0.510 Å, θ = 12.40°, φ = 253°) and (Q = 0.546 Å, θ = 5.19°, φ = 276.96°) for EX and the co-crystal (Boeyens, 1978), respectively, whereas the five-membered ring D is folded like an envelope, with a maximum deviation of 0.262 and 0.265 Å for C14 with puckering parameters of Q = 0.4116 Å, φ = 206.59° and Q = 0.4124 Å, φ = 210.11° for EX and the co-former (Cremer & Pople, 1975). The α,β-unsaturated carbonyl carbon (C3=O1) was found to be involved in the strong conjugation with the C1=C2 and C4=C5 olefinic bonds. Two methyl groups were found to be axially oriented (Fig. 2). The conformational features of the EX molecule in the co-crystal were found to be similar to that of the individual molecule. In comparison, the difference between the bond length of the C3=O1 olefinic bond was found to deviate between 1.194 and 1.189 Å, with the orientation of the angle C2—C3—O1 being 121.85 and 122.54°, and C3—C4—O1 being 120.97 and 121.09° for EX and the co-crystal, respectively (Fig. 3).

Figure 2 ORTEP drawing of the co-crystal of exemestane (6-methyl­ideneandrost-4-ene-3,17-dione) and thio­urea [(C20H24O2)(CH4N2S)] showing 30% probability ellipsoids.

Figure 3 Overlapped ORTEP drawing of EX and the co-crystal with 30% probability ellipsoids.

3.1. Supramolecular features

Sulfur and nitro­gen atoms of TH play an important role in providing stability to the co-crystal via N(1)—H(1A)⋯S(1), N(1)—H(1B)⋯O(2), N(2)—H(2A)⋯S(1), N(2)—H(2B)⋯O(1), C(4)—H(4)⋯O(2) and C(20)—H(20A)⋯O(2) intermolecular interactions to form a three-dimensional network [Fig. 4(c)]. Among these interactions, C(2)—H(2B)⋯O(1) was found to be the strongest, with a bond length of 1.96 Å. N(1)—H(1A)⋯S(1) and N(2)—H(2A)⋯S(1) interactions result in the formation of an S6 ring graph set motif. N(1)—H(1B)⋯O(2) and N(2)—H(2B)⋯O(1) interactions are responsible for the two-dimensional unit cell packing; details of these interactions are summarized in Table 2.

Table 2
Hydrogen bonds for the co-crystal (Å and °)

D–H⋯A	d(D–H)	d(H⋯A)	d(D⋯A)	<(DHA)
N(1)—H(1A)⋯S(1)^i	0.86	2.75	3.511 (3)	148.9
N(1)—H(1B)⋯O(2)^ii	0.86	2.01	2.867 (4)	176.6
N(2)—H(2A)⋯S(1)^i	0.86	2.42	3.253 (3)	164.4
N(2)—H(2B)⋯O(1)	0.86	1.96	2.821 (4)	174.2
C(4)—H(4)⋯O(2)^iii	0.93	2.49	3.336 (5)	151.3
Symmetry transformations used to generate equivalent atoms: (i) −x + 1, y − 1/2, −z; (ii) x + 1, y, z − 1; (iii) −x, y + 1/2, −z + 1.

Figure 4 (a) 21 axis rotation. (b) Parallel chain elongation along the b axis. (c) Unit cell packing diagram. (d) Intermolecular hydrogen bonding within the co-crystal, viewed along b axis to link neighbouring molecules to form a three-dimensional network.

A two-dimensional Hirshfeld surface (Spackman & Jayatilaka, 2009) has been generated using d_norm [Fig. 5(a)]. In the Hirshfeld surface, the question of `compact packing' arises (Spackman & Jayatilaka, 2009), as shown in the encircled area and empty s spaces (A) in the unit cell. This is answered by the expansion of crystal packing, revealing that the neighbouring molecule is encapsulating this gap [Fig. 5(b)]. The above situation is the main reason for analysing the API and co-former separately in qualitative and qualitative analyses, as they are involved in intermolecular interactions with neighbouring molecules.

Figure 5 (a) Two dimensional Hirshfeld surface generated for the co-crystal. (b) Two dimensional Hirshfeld surface generated to show the compact nature of the Hirshfeld surface.

Fig. 6 shows the intermolecular interactions of the API and co-former with the neighbouring moieties. S⋯H and the O⋯H are important interactions responsible for forming strong hydrogen bonds as indicated by the red spots on the Hirshfeld surface.

Figure 6 Intermolecular interactions with the two-dimensional Hirshfeld surface generated for the co-crystal.

Two-dimensional fingerprint plots (McKinnon et al., 2007) (FPs) for the API [Fig. 7(a)] show the percentage contribution of contacts towards the crystal packing in which conventionally H⋯H contributes the most, amounting to 65.5%. O⋯H was found to be the strongest interaction among them all as shown by the sharp spike pointing to the origin and a 17.9% contribution to the unit-cell packing, followed by a 10.2% contribution from C⋯H. S⋯H and N⋯H contribute 4.2 and 2.1%, respectively; usually FPs generate the percentage contribution as a reciprocal for both distances d_e and d_i. Here in the case of S⋯H and N⋯H for the API, cyan dots are only in the region of de, which means S and N atoms are present only outside of the generated surface. In the case of the co-former [Fig. 7(b)], the size of the molecule is small, which is why the percentage contribution of H⋯H is equal to S⋯H, i.e. a 36.3% contribution to the unit cell packing. The O⋯H contribution is 15.2%, followed by contributions of 6.6 and 5.5% for N⋯H and C⋯H. In Fig. 7(c), FPs for the asymmetric unit have been drawn, which reveals that H⋯H contacts are at a maximum of 60.3%, whereas O⋯H and S⋯H are 14.2 and 12.5%, respectively; in addition to these contacts, C⋯H and N⋯H show contribution of 9.3 and 3.4% to the unit cell packing. Two valuable tools of the Hirshfeld surface, namely the shape index and curvedness, reveal information about the π–π stacking and the probability of forming interactions with neighbouring molecules, respectively. The shape index [Fig. 8(a)] surface showed blurred patches which reveal the presence of weak π–π stacking of the co-former moiety with neighbouring molecules. Curvedness [Fig. 8(b)] shows the packing probabilities at various positions of the surface generated, these bumps and hollows on the surface reflect the packing behaviour. The ab initio electrostatic potential (Fig. 9) (Spackman et al., 2008) surface generated over the Hirshfeld surface shows the positive and negative potential sites of the molecule. In Fig. 9, blue dots over the surface relate the positions of the strong contacts to hydrogen-bond acceptors, whereas the red regions relate positions on the surface to hydrogen-bond donors (Fig. 9). The electrostatic potential map shows that the region of the surface where C=S of the TH moiety is more electronegative than the C=O attached to ring D.

Figure 7 (a) Two-dimensional FPs of the API showing the percentage contribution of all contacts to the crystal packing. (b) Two-dimensional FPs for the TH moiety showing the percentage contribution of contacts to the crystal packing. (c) Two-dimensional FPs of the asymmetric unit.

Figure 8 Shape index and curvedness mapped over the Hirshfeld surface.

Figure 9 (a) Electrostatic potential surfaces mapped over the Hirshfeld surface to view the electropositive and electronegative sites. (b) Packing mode of the electrostatic potential surface.

3.2. Thermogravimetric analysis

From TGA studies (Steed, 2013) it is evident that the co-crystal (EX:TH) demonstrated a thermal stability up to 176.18°C, with a percentage weight loss of 2.384%. When the temperature was increased from 176.18 to 334.21°C, frequent percentage weight loss up to 51.396% was observed for the co-crystal (EX:TH) (Fig. 10). DSC spectra of the co-crystal (Fig. 11) clearly show three endothermic peaks observed at 170.27, 198.2 and 231.50°C due to the fusion of the TH crystal, thio­urea:exemestane co-crystal (EX:TH) and the degradation of the thio­urea polymer, respectively.

Figure 10 TGA Spectra of co-crystal (EX:TH).

Figure 11 DSC Spectra of co-crystal (EX:TH).

3.3. Biological activity

EX has been reported as an anti-cancer agent against breast cancer, therefore, the synthesized co-crystal was first evaluated in vitro to observe any changes in its anti-cancer activity against the MCF-7 breast cancer cell line; however, it was found to be ineffective. On the other hand, the co-former (thio­urea) is reported as a tested standard to check urease inhibition activity in vitro (Kanwal et al., 2018). Therefore, both the API (EX) and the synthesized co-crystal (EX:TH) were evaluated for their urease enzyme inhibition activity (in vitro) and interesting results were obtained. EX was found to be inactive against the urease enzyme. However, its co-crystal showed potent urease inhibition activity (IC₅₀ = 3.86 ± 0.31µM) and was found to be several-fold more active than standard TH (IC₅₀ = 21.0 ± 1.45µM). These promising results clearly indicate that the significant change in activity can be attributed to the synergistic effects of both the API and co-former; however, further studies are required to be able to comment on that. Results of the biological activity evaluation are summarized in Table 3.

Table 3
Bioactivity of exemestane and the co-crystal

Sample	Urease inhibition activity (IC50 in µM ± SEM)	Anti-cancer activity MCF-7 cell line (IC50 in µM ± SEM)	Cytotoxicity against 3T3 normal fibroblast cell line (IC50 in µM ± SEM)
EX (API)	NA	Not evaluated	>30
Co-crystal (EX:TH)	3.86 ± 0.31	Inactive	>30
Standard	Thio­urea 21.0 ± 1.45	Doxorubicin 0.924 ± 0.01	Cyclo­heximide 0.26 ± 0.12

4. Conclusions

A co-crystal of the commercially available anti-cancer drug exemestane with thio­urea has been successfully synthesized with enhanced urease inhibition activity (in vitro) compared with that of the API (EX) and co-former (thio­urea) individually. The promising results of urease inhibition activity clearly demonstrate the possibility of the EX:TH (1:1) co-crystal as a new potential lead against peptic ulcers and uroli­thia­sis. The contribution of non-covalent interactions to the structural stability of the co-crystal was also evaluated quantitatively with the aid of Hirshfeld surface analysis, and clearly demonstrated the role of the amine and carbonyl functionalities of the co-former and API in forming various contacts in the crystal structure. Electrostatic potential analysis clearly identified the sites of the molecule that can be targeted for structural modification. The promising results of the present study clearly demonstrate the role of co-crystallization as an effective tool in the search for potential leads against various health disorders.

Supporting information

Computing details

(Exemestane)

Crystal data

C20H24O2	F(000) = 640
Mr = 296.39	Dx = 1.312 Mg m^−3
Orthorhombic, P212121	Mo Kα radiation, λ = 0.71073 Å
a = 9.959 (4) Å	Cell parameters from 1063 reflections
b = 11.731 (4) Å	θ = 2.6–19.4°
c = 12.842 (5) Å	µ = 0.08 mm^−1
V = 1500.2 (10) Å^3	T = 273 K
Z = 4	BLOCK, colorles

Data collection

Bruker APEX-II CCD diffractometer	Rint = 0.069
φ and ω scans	θmax = 28.4°, θmin = 2.4°
10477 measured reflections	h = −13→13
3736 independent reflections	k = −15→14
1742 reflections with I > 2σ(I)	l = −17→15

Refinement

Refinement on F^2	Hydrogen site location: inferred from neighbouring sites
Least-squares matrix: full	H-atom parameters constrained
R[F^2 > 2σ(F^2)] = 0.057	w = 1/[σ^2(Fo^2) + (0.1P)^2] where P = (Fo^2 + 2Fc^2)/3
wR(F^2) = 0.192	(Δ/σ)max < 0.001
S = 0.87	Δρmax = 0.12 e Å^−3
3736 reflections	Δρmin = −0.12 e Å^−3
202 parameters	Extinction correction: SHELXL-2014 (Sheldrick 2014), Fc^*=kFc[1+0.001xFc^2λ^3/sin(2θ)]^-1/4
0 restraints	Extinction coefficient: 0.015 (4)
Primary atom site location: Direct Method	Absolute structure: Flack x determined using 489 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons, Flack and Wagner, Acta Cryst. B69 (2013) 249-259).
Secondary atom site location: Fourier Synthesis Method	Absolute structure parameter: −1.2 (10)

Special details

Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
O1	0.1953 (4)	0.7080 (4)	0.6229 (3)	0.1029 (14)
C1	0.2574 (4)	0.6152 (4)	0.3751 (4)	0.0546 (11)
C2	0.2003 (5)	0.6668 (4)	0.4512 (4)	0.0636 (13)
H2	0.119769	0.704747	0.439205	0.076*
O2	0.8306 (3)	0.6236 (4)	0.0299 (3)	0.0870 (12)
C3	0.2544 (6)	0.6680 (5)	0.5512 (5)	0.0710 (15)
C4	0.3816 (5)	0.6204 (6)	0.5637 (4)	0.0773 (16)
H4	0.424378	0.626002	0.627875	0.093*
C5	0.4403 (5)	0.5695 (6)	0.4893 (4)	0.0763 (16)
H5	0.526408	0.541825	0.501284	0.092*
C6	0.3825 (4)	0.5515 (4)	0.3869 (4)	0.0555 (12)
C10	0.1997 (4)	0.6169 (4)	0.2735 (4)	0.0593 (12)
C9	0.2928 (4)	0.6413 (4)	0.1890 (4)	0.0602 (13)
H9A	0.313046	0.722171	0.188866	0.072*
H9B	0.249977	0.623041	0.123221	0.072*
C8	0.4192 (4)	0.5765 (4)	0.1976 (3)	0.0503 (11)
H8	0.400274	0.495626	0.185235	0.060*
C7	0.4787 (4)	0.5892 (4)	0.3029 (3)	0.0540 (12)
H7	0.493228	0.671072	0.313410	0.065*
C13	0.6463 (4)	0.5541 (4)	0.1248 (4)	0.0571 (12)
C12	0.7069 (4)	0.5703 (5)	0.2272 (4)	0.0648 (13)
H12A	0.729756	0.650074	0.236556	0.078*
H12B	0.789055	0.526147	0.231809	0.078*
C11	0.6132 (4)	0.5342 (5)	0.3105 (4)	0.0692 (15)
H11A	0.652454	0.552425	0.377551	0.083*
H11B	0.602051	0.452131	0.307081	0.083*
C15	0.4829 (5)	0.6191 (5)	0.0082 (4)	0.0651 (13)
H15A	0.444168	0.547417	−0.014613	0.078*
H15B	0.420949	0.680591	−0.007201	0.078*
C14	0.5174 (4)	0.6159 (4)	0.1198 (4)	0.0530 (11)
H14	0.538294	0.694882	0.138823	0.064*
C18	0.3531 (5)	0.4280 (4)	0.3780 (4)	0.0764 (15)
H18A	0.313533	0.412663	0.311291	0.115*
H18B	0.291823	0.405946	0.432047	0.115*
H18C	0.434994	0.385496	0.384871	0.115*
C20	0.6347 (5)	0.4309 (5)	0.0969 (5)	0.0777 (16)
H20A	0.609072	0.423883	0.025099	0.117*
H20B	0.568000	0.395466	0.140055	0.117*
H20C	0.719715	0.394020	0.107537	0.117*
C19	0.0737 (5)	0.6007 (6)	0.2579 (5)	0.0904 (19)
H19A	0.016665	0.587370	0.313947	0.109*
H19B	0.039439	0.602441	0.190525	0.109*
C17	0.7146 (5)	0.6074 (5)	0.0373 (4)	0.0670 (14)
C16	0.6151 (5)	0.6389 (5)	−0.0399 (5)	0.0790 (17)
H16A	0.625207	0.592598	−0.101960	0.095*
H16B	0.625010	0.718420	−0.059000	0.095*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
O1	0.113 (3)	0.107 (3)	0.089 (3)	0.012 (3)	0.025 (3)	−0.028 (3)
C1	0.047 (2)	0.051 (3)	0.066 (3)	0.000 (2)	0.008 (2)	0.005 (2)
C2	0.054 (3)	0.059 (3)	0.078 (4)	0.001 (2)	0.009 (3)	0.000 (3)
O2	0.056 (2)	0.098 (3)	0.107 (3)	−0.012 (2)	0.029 (2)	−0.015 (2)
C3	0.078 (3)	0.065 (3)	0.070 (4)	−0.004 (3)	0.020 (3)	−0.010 (3)
C4	0.073 (3)	0.098 (4)	0.061 (3)	−0.001 (4)	0.006 (3)	0.000 (3)
C5	0.064 (3)	0.099 (4)	0.065 (3)	0.014 (3)	0.001 (3)	0.010 (3)
C6	0.052 (2)	0.056 (3)	0.058 (3)	0.006 (2)	0.006 (2)	0.004 (2)
C10	0.044 (2)	0.062 (3)	0.072 (3)	0.003 (2)	0.002 (2)	0.002 (2)
C9	0.044 (2)	0.072 (3)	0.064 (3)	0.006 (2)	0.003 (2)	0.008 (2)
C8	0.043 (2)	0.049 (3)	0.059 (3)	−0.0019 (19)	−0.001 (2)	0.000 (2)
C7	0.044 (2)	0.058 (3)	0.059 (3)	0.003 (2)	0.001 (2)	0.001 (2)
C13	0.046 (2)	0.057 (3)	0.069 (3)	0.000 (2)	0.008 (2)	−0.006 (2)
C12	0.041 (2)	0.082 (4)	0.071 (3)	0.007 (2)	0.003 (2)	−0.005 (3)
C11	0.045 (2)	0.093 (4)	0.069 (3)	0.016 (3)	−0.002 (2)	0.003 (3)
C15	0.061 (3)	0.077 (3)	0.058 (3)	−0.003 (3)	0.006 (2)	0.005 (3)
C14	0.045 (2)	0.053 (3)	0.060 (3)	−0.003 (2)	0.005 (2)	0.000 (2)
C18	0.082 (3)	0.058 (3)	0.089 (4)	0.007 (3)	0.026 (3)	0.013 (3)
C20	0.072 (3)	0.057 (3)	0.104 (4)	0.008 (3)	0.011 (3)	−0.009 (3)
C19	0.049 (3)	0.138 (6)	0.084 (4)	−0.002 (3)	0.000 (3)	0.004 (4)
C17	0.058 (3)	0.066 (3)	0.077 (3)	−0.005 (3)	0.018 (3)	−0.020 (3)
C16	0.079 (4)	0.085 (4)	0.073 (4)	−0.008 (3)	0.018 (3)	0.003 (3)

Geometric parameters (Å, º)

O1—C3	1.190 (6)	C13—C12	1.459 (6)
C1—C2	1.283 (6)	C13—C14	1.476 (6)
C1—C10	1.426 (7)	C13—C20	1.494 (7)
C1—C6	1.460 (6)	C12—C11	1.481 (6)
C2—C3	1.392 (8)	C12—H12A	0.9700
C2—H2	0.9300	C12—H12B	0.9700
O2—C17	1.174 (5)	C11—H11A	0.9700
C3—C4	1.395 (7)	C11—H11B	0.9700
C4—C5	1.268 (7)	C15—C16	1.473 (7)
C4—H4	0.9300	C15—C14	1.474 (7)
C5—C6	1.451 (7)	C15—H15A	0.9700
C5—H5	0.9300	C15—H15B	0.9700
C6—C18	1.482 (7)	C14—H14	0.9800
C6—C7	1.509 (6)	C18—H18A	0.9600
C10—C19	1.285 (6)	C18—H18B	0.9600
C10—C9	1.456 (6)	C18—H18C	0.9600
C9—C8	1.475 (6)	C20—H20A	0.9600
C9—H9A	0.9700	C20—H20B	0.9600
C9—H9B	0.9700	C20—H20C	0.9600
C8—C14	1.473 (6)	C19—H19A	0.9300
C8—C7	1.484 (6)	C19—H19B	0.9300
C8—H8	0.9800	C17—C16	1.449 (7)
C7—C11	1.490 (6)	C16—H16A	0.9700
C7—H7	0.9800	C16—H16B	0.9700
C13—C17	1.454 (7)
C2—C1—C10	120.8 (4)	C13—C12—C11	110.6 (4)
C2—C1—C6	122.8 (5)	C13—C12—H12A	109.5
C10—C1—C6	116.5 (4)	C11—C12—H12A	109.5
C1—C2—C3	122.4 (5)	C13—C12—H12B	109.5
C1—C2—H2	118.8	C11—C12—H12B	109.5
C3—C2—H2	118.8	H12A—C12—H12B	108.1
O1—C3—C2	121.8 (6)	C12—C11—C7	113.3 (4)
O1—C3—C4	121.2 (6)	C12—C11—H11A	108.9
C2—C3—C4	117.0 (5)	C7—C11—H11A	108.9
C5—C4—C3	121.3 (5)	C12—C11—H11B	108.9
C5—C4—H4	119.3	C7—C11—H11B	108.9
C3—C4—H4	119.3	H11A—C11—H11B	107.7
C4—C5—C6	124.6 (5)	C16—C15—C14	101.7 (4)
C4—C5—H5	117.7	C16—C15—H15A	111.4
C6—C5—H5	117.7	C14—C15—H15A	111.4
C5—C6—C1	111.0 (4)	C16—C15—H15B	111.4
C5—C6—C18	106.9 (4)	C14—C15—H15B	111.4
C1—C6—C18	108.9 (4)	H15A—C15—H15B	109.3
C5—C6—C7	110.7 (4)	C8—C14—C15	120.8 (4)
C1—C6—C7	108.5 (4)	C8—C14—C13	113.2 (4)
C18—C6—C7	110.9 (4)	C15—C14—C13	104.9 (4)
C19—C10—C1	122.3 (5)	C8—C14—H14	105.6
C19—C10—C9	122.3 (5)	C15—C14—H14	105.6
C1—C10—C9	115.4 (4)	C13—C14—H14	105.6
C10—C9—C8	112.7 (4)	C6—C18—H18A	109.5
C10—C9—H9A	109.0	C6—C18—H18B	109.5
C8—C9—H9A	109.0	H18A—C18—H18B	109.5
C10—C9—H9B	109.0	C6—C18—H18C	109.5
C8—C9—H9B	109.0	H18A—C18—H18C	109.5
H9A—C9—H9B	107.8	H18B—C18—H18C	109.5
C14—C8—C9	110.7 (4)	C13—C20—H20A	109.5
C14—C8—C7	108.8 (4)	C13—C20—H20B	109.5
C9—C8—C7	110.9 (4)	H20A—C20—H20B	109.5
C14—C8—H8	108.8	C13—C20—H20C	109.5
C9—C8—H8	108.8	H20A—C20—H20C	109.5
C7—C8—H8	108.8	H20B—C20—H20C	109.5
C8—C7—C11	112.0 (4)	C10—C19—H19A	120.0
C8—C7—C6	111.6 (4)	C10—C19—H19B	120.0
C11—C7—C6	113.4 (4)	H19A—C19—H19B	120.0
C8—C7—H7	106.4	O2—C17—C16	125.1 (5)
C11—C7—H7	106.4	O2—C17—C13	126.3 (5)
C6—C7—H7	106.4	C16—C17—C13	108.6 (4)
C17—C13—C12	116.5 (4)	C17—C16—C15	106.5 (4)
C17—C13—C14	99.3 (4)	C17—C16—H16A	110.4
C12—C13—C14	109.6 (4)	C15—C16—H16A	110.4
C17—C13—C20	105.5 (4)	C17—C16—H16B	110.4
C12—C13—C20	112.0 (4)	C15—C16—H16B	110.4
C14—C13—C20	113.4 (4)	H16A—C16—H16B	108.6
C10—C1—C2—C3	−179.3 (5)	C5—C6—C7—C11	55.0 (6)
C6—C1—C2—C3	1.2 (8)	C1—C6—C7—C11	177.0 (4)
C1—C2—C3—O1	−173.7 (5)	C18—C6—C7—C11	−63.5 (6)
C1—C2—C3—C4	6.5 (8)	C17—C13—C12—C11	167.6 (4)
O1—C3—C4—C5	174.2 (6)	C14—C13—C12—C11	55.9 (5)
C2—C3—C4—C5	−5.9 (9)	C20—C13—C12—C11	−70.9 (5)
C3—C4—C5—C6	−2.3 (10)	C13—C12—C11—C7	−53.3 (6)
C4—C5—C6—C1	9.2 (8)	C8—C7—C11—C12	51.9 (6)
C4—C5—C6—C18	−109.4 (7)	C6—C7—C11—C12	179.3 (4)
C4—C5—C6—C7	129.7 (6)	C9—C8—C14—C15	−54.5 (6)
C2—C1—C6—C5	−8.5 (6)	C7—C8—C14—C15	−176.6 (4)
C10—C1—C6—C5	172.0 (4)	C9—C8—C14—C13	179.9 (4)
C2—C1—C6—C18	108.9 (5)	C7—C8—C14—C13	57.8 (5)
C10—C1—C6—C18	−70.6 (5)	C16—C15—C14—C8	−168.4 (5)
C2—C1—C6—C7	−130.3 (5)	C16—C15—C14—C13	−39.1 (5)
C10—C1—C6—C7	50.2 (5)	C17—C13—C14—C8	177.0 (4)
C2—C1—C10—C19	−44.4 (8)	C12—C13—C14—C8	−60.4 (5)
C6—C1—C10—C19	135.1 (6)	C20—C13—C14—C8	65.5 (6)
C2—C1—C10—C9	133.9 (5)	C17—C13—C14—C15	43.3 (5)
C6—C1—C10—C9	−46.6 (6)	C12—C13—C14—C15	165.9 (4)
C19—C10—C9—C8	−136.2 (6)	C20—C13—C14—C15	−68.2 (5)
C1—C10—C9—C8	45.4 (6)	C12—C13—C17—O2	30.9 (8)
C10—C9—C8—C14	−171.7 (4)	C14—C13—C17—O2	148.3 (5)
C10—C9—C8—C7	−50.9 (6)	C20—C13—C17—O2	−94.0 (7)
C14—C8—C7—C11	−52.3 (5)	C12—C13—C17—C16	−148.6 (5)
C9—C8—C7—C11	−174.3 (4)	C14—C13—C17—C16	−31.1 (5)
C14—C8—C7—C6	179.3 (4)	C20—C13—C17—C16	86.5 (5)
C9—C8—C7—C6	57.3 (5)	O2—C17—C16—C15	−171.5 (6)
C5—C6—C7—C8	−177.4 (4)	C13—C17—C16—C15	8.0 (6)
C1—C6—C7—C8	−55.4 (5)	C14—C15—C16—C17	18.9 (6)
C18—C6—C7—C8	64.1 (5)

(Exemestane-thiourea)

Crystal data

C20H24O2·CH4N2S	F(000) = 400
Mr = 372.51	Dx = 1.378 Mg m^−3
Monoclinic, P21	Mo Kα radiation, λ = 0.71073 Å
a = 10.3797 (14) Å	Cell parameters from 1442 reflections
b = 8.3706 (11) Å	θ = 2.4–21.9°
c = 10.7275 (14) Å	µ = 0.20 mm^−1
β = 105.568 (3)°	T = 293 K
V = 897.9 (2) Å^3	BLOCK, colorles
Z = 2	0.17 × 0.15 × 0.07 mm

Data collection

CCD area detector diffractometer	Rint = 0.033
Radiation source: sealed tube	θmax = 28.4°, θmin = 2.0°
phi and ω scans	h = −13→13
6563 measured reflections	k = −6→11
3315 independent reflections	l = −14→14
2402 reflections with I > 2σ(I)

Refinement

Refinement on F^2	Secondary atom site location: Fourier Synthesis Method
Least-squares matrix: full	Hydrogen site location: mixed
R[F^2 > 2σ(F^2)] = 0.047	H atoms treated by a mixture of independent and constrained refinement
wR(F^2) = 0.110	w = 1/[σ^2(Fo^2) + (0.0489P)^2 + 0.0449P] where P = (Fo^2 + 2Fc^2)/3
S = 1.02	(Δ/σ)max < 0.001
3315 reflections	Δρmax = 0.17 e Å^−3
243 parameters	Δρmin = −0.20 e Å^−3
1 restraint	Absolute structure: Flack x determined using 578 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons, Flack and Wagner, Acta Cryst. B69 (2013) 249-259).
Primary atom site location: Direct method	Absolute structure parameter: 0.05 (7)

Special details

Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
S1	0.66432 (10)	0.40820 (14)	0.10510 (9)	0.0608 (3)
O1	0.3665 (3)	0.3877 (5)	0.2447 (3)	0.0893 (11)
O2	−0.1536 (3)	0.2406 (4)	0.8737 (3)	0.0703 (9)
N1	0.6055 (3)	0.1756 (4)	−0.0531 (3)	0.0540 (8)
H1A	0.554094	0.099371	−0.090876	0.065*
H1B	0.679599	0.193610	−0.072029	0.065*
N2	0.4599 (3)	0.2283 (4)	0.0542 (3)	0.0495 (8)
H2A	0.412016	0.150760	0.013897	0.059*
H2B	0.432806	0.283062	0.110113	0.059*
C1	0.3531 (4)	0.2725 (5)	0.5472 (4)	0.0547 (10)
H1	0.369851	0.181426	0.598746	0.066*
C2	0.3707 (4)	0.2640 (6)	0.4341 (4)	0.0608 (11)
H2	0.400132	0.168225	0.407393	0.073*
C3	0.3472 (3)	0.3937 (7)	0.3490 (4)	0.0606 (11)
C4	0.2981 (4)	0.5321 (5)	0.3921 (4)	0.0532 (10)
H4	0.276925	0.618924	0.336286	0.064*
C5	0.2808 (3)	0.5446 (4)	0.5071 (3)	0.0387 (8)
C6	0.2280 (3)	0.6858 (5)	0.5471 (3)	0.0431 (8)
C7	0.1180 (4)	0.6612 (4)	0.6047 (3)	0.0467 (9)
H7A	0.037138	0.641275	0.536215	0.056*
H7B	0.104403	0.758455	0.648664	0.056*
C8	0.1388 (3)	0.5280 (4)	0.6973 (3)	0.0373 (8)
H8	0.208121	0.558520	0.775383	0.045*
C9	0.1831 (3)	0.3826 (4)	0.6420 (3)	0.0366 (7)
H9	0.112870	0.357837	0.563022	0.044*
C10	0.3091 (3)	0.4138 (5)	0.6002 (3)	0.0398 (7)
C11	0.1937 (3)	0.2420 (5)	0.7272 (4)	0.0483 (9)
H11A	0.268250	0.257489	0.803125	0.058*
H11B	0.212805	0.148520	0.681834	0.058*
C12	0.0715 (3)	0.2117 (4)	0.7693 (4)	0.0486 (9)
H12A	−0.000501	0.178968	0.695729	0.058*
H12B	0.087914	0.125983	0.832494	0.058*
C13	0.0330 (3)	0.3557 (4)	0.8262 (3)	0.0430 (8)
C14	0.0163 (3)	0.4893 (4)	0.7336 (3)	0.0409 (8)
H14	−0.047556	0.451856	0.654134	0.049*
C15	−0.0989 (4)	0.3532 (5)	0.8503 (4)	0.0503 (10)
C16	−0.1527 (4)	0.5144 (6)	0.8344 (5)	0.0672 (12)
H16A	−0.162350	0.555166	0.916133	0.081*
H16B	−0.239642	0.515972	0.771717	0.081*
C17	−0.0556 (4)	0.6126 (5)	0.7884 (4)	0.0561 (10)
H17A	−0.100772	0.688497	0.722867	0.067*
H17B	0.004667	0.669560	0.859187	0.067*
C18	0.4216 (4)	0.4639 (5)	0.7125 (3)	0.0541 (11)
H18A	0.401998	0.566684	0.742777	0.081*
H18B	0.502666	0.470201	0.685951	0.081*
H18C	0.432249	0.387084	0.780986	0.081*
C19	0.2751 (5)	0.8231 (6)	0.5350 (5)	0.0621 (11)
C20	0.1284 (3)	0.3930 (6)	0.9545 (3)	0.0566 (10)
H20A	0.096357	0.483712	0.992164	0.085*
H20B	0.214875	0.416740	0.942805	0.085*
H20C	0.135129	0.302591	1.010965	0.085*
C21	0.5717 (3)	0.2621 (4)	0.0310 (3)	0.0415 (8)
H19A	0.347 (4)	0.833 (4)	0.496 (3)	0.050*
H19B	0.244 (3)	0.908 (5)	0.564 (3)	0.050*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
S1	0.0648 (6)	0.0517 (6)	0.0602 (6)	−0.0056 (5)	0.0069 (4)	−0.0019 (6)
O1	0.096 (2)	0.125 (3)	0.0578 (17)	0.021 (2)	0.0412 (16)	−0.014 (2)
O2	0.0671 (18)	0.074 (2)	0.084 (2)	−0.0160 (16)	0.0453 (16)	−0.0047 (18)
N1	0.0525 (17)	0.052 (2)	0.063 (2)	0.0019 (15)	0.0258 (15)	−0.0040 (17)
N2	0.0480 (17)	0.053 (2)	0.0510 (17)	0.0003 (14)	0.0198 (14)	−0.0032 (16)
C1	0.052 (2)	0.043 (2)	0.079 (3)	0.0076 (17)	0.033 (2)	0.004 (2)
C2	0.060 (2)	0.055 (3)	0.078 (3)	0.005 (2)	0.036 (2)	−0.016 (2)
C3	0.048 (2)	0.084 (3)	0.055 (2)	0.004 (2)	0.0219 (17)	−0.017 (3)
C4	0.051 (2)	0.067 (3)	0.045 (2)	0.009 (2)	0.0189 (17)	0.0050 (19)
C5	0.0358 (17)	0.042 (2)	0.0397 (19)	0.0002 (15)	0.0119 (14)	0.0025 (16)
C6	0.052 (2)	0.041 (2)	0.0376 (18)	0.0038 (17)	0.0132 (15)	0.0011 (16)
C7	0.057 (2)	0.040 (2)	0.048 (2)	0.0067 (17)	0.0236 (16)	0.0039 (17)
C8	0.0412 (17)	0.039 (2)	0.0326 (17)	0.0019 (15)	0.0115 (14)	−0.0004 (15)
C9	0.0374 (16)	0.037 (2)	0.0373 (16)	0.0015 (14)	0.0128 (13)	−0.0016 (15)
C10	0.0404 (16)	0.0378 (18)	0.0432 (17)	0.0026 (16)	0.0147 (13)	0.0011 (17)
C11	0.055 (2)	0.037 (2)	0.060 (2)	0.0050 (16)	0.0279 (18)	0.0063 (18)
C12	0.055 (2)	0.040 (2)	0.055 (2)	−0.0054 (17)	0.0237 (17)	0.0026 (18)
C13	0.0442 (18)	0.047 (2)	0.0408 (17)	−0.0031 (15)	0.0158 (15)	0.0001 (16)
C14	0.0410 (18)	0.045 (2)	0.0375 (18)	0.0015 (15)	0.0126 (15)	−0.0032 (16)
C15	0.049 (2)	0.060 (3)	0.047 (2)	−0.0076 (19)	0.0211 (16)	−0.0050 (19)
C16	0.060 (3)	0.073 (3)	0.080 (3)	0.012 (2)	0.039 (2)	0.006 (3)
C17	0.059 (2)	0.055 (3)	0.064 (3)	0.0079 (19)	0.033 (2)	−0.001 (2)
C18	0.042 (2)	0.065 (3)	0.051 (2)	−0.0072 (17)	0.0064 (16)	0.0100 (19)
C19	0.075 (3)	0.042 (3)	0.078 (3)	0.004 (2)	0.036 (2)	0.003 (2)
C20	0.061 (2)	0.066 (3)	0.0420 (18)	−0.008 (2)	0.0111 (16)	0.002 (2)
C21	0.0457 (19)	0.041 (2)	0.0367 (18)	0.0110 (16)	0.0093 (15)	0.0094 (16)

Geometric parameters (Å, º)

S1—C21	1.625 (4)	C9—C10	1.514 (4)
O1—C3	1.189 (4)	C9—H9	0.9800
O2—C15	1.161 (4)	C10—C18	1.496 (5)
N1—C21	1.277 (4)	C11—C12	1.477 (5)
N1—H1A	0.8600	C11—H11A	0.9700
N1—H1B	0.8600	C11—H11B	0.9700
N2—C21	1.282 (4)	C12—C13	1.454 (5)
N2—H2A	0.8600	C12—H12A	0.9700
N2—H2B	0.8600	C12—H12B	0.9700
C1—C2	1.277 (5)	C13—C15	1.460 (5)
C1—C10	1.439 (5)	C13—C14	1.475 (5)
C1—H1	0.9300	C13—C20	1.498 (5)
C2—C3	1.397 (7)	C14—C17	1.484 (5)
C2—H2	0.9300	C14—H14	0.9800
C3—C4	1.394 (6)	C15—C16	1.453 (6)
C4—C5	1.297 (5)	C16—C17	1.484 (6)
C4—H4	0.9300	C16—H16A	0.9700
C5—C6	1.417 (5)	C16—H16B	0.9700
C5—C10	1.458 (5)	C17—H17A	0.9700
C6—C19	1.269 (6)	C17—H17B	0.9700
C6—C7	1.451 (5)	C18—H18A	0.9600
C7—C8	1.470 (5)	C18—H18B	0.9600
C7—H7A	0.9700	C18—H18C	0.9600
C7—H7B	0.9700	C19—H19A	0.95 (3)
C8—C14	1.463 (5)	C19—H19B	0.87 (4)
C8—C9	1.480 (4)	C20—H20A	0.9600
C8—H8	0.9800	C20—H20B	0.9600
C9—C11	1.476 (5)	C20—H20C	0.9600
C21—N1—H1A	120.0	H11A—C11—H11B	107.7
C21—N1—H1B	120.0	C13—C12—C11	109.9 (3)
H1A—N1—H1B	120.0	C13—C12—H12A	109.7
C21—N2—H2A	120.0	C11—C12—H12A	109.7
C21—N2—H2B	120.0	C13—C12—H12B	109.7
H2A—N2—H2B	120.0	C11—C12—H12B	109.7
C2—C1—C10	124.2 (4)	H12A—C12—H12B	108.2
C2—C1—H1	117.9	C12—C13—C15	116.0 (3)
C10—C1—H1	117.9	C12—C13—C14	110.0 (3)
C1—C2—C3	122.2 (4)	C15—C13—C14	100.6 (3)
C1—C2—H2	118.9	C12—C13—C20	111.7 (3)
C3—C2—H2	118.9	C15—C13—C20	105.1 (3)
O1—C3—C4	121.1 (5)	C14—C13—C20	113.0 (3)
O1—C3—C2	122.5 (5)	C8—C14—C13	113.2 (3)
C4—C3—C2	116.4 (3)	C8—C14—C17	121.1 (3)
C5—C4—C3	122.6 (4)	C13—C14—C17	104.0 (3)
C5—C4—H4	118.7	C8—C14—H14	105.8
C3—C4—H4	118.7	C13—C14—H14	105.8
C4—C5—C6	121.3 (4)	C17—C14—H14	105.8
C4—C5—C10	122.5 (4)	O2—C15—C16	125.8 (4)
C6—C5—C10	116.2 (3)	O2—C15—C13	125.8 (4)
C19—C6—C5	122.2 (4)	C16—C15—C13	108.4 (3)
C19—C6—C7	122.7 (4)	C15—C16—C17	106.2 (3)
C5—C6—C7	115.0 (3)	C15—C16—H16A	110.5
C6—C7—C8	113.9 (3)	C17—C16—H16A	110.5
C6—C7—H7A	108.8	C15—C16—H16B	110.5
C8—C7—H7A	108.8	C17—C16—H16B	110.5
C6—C7—H7B	108.8	H16A—C16—H16B	108.7
C8—C7—H7B	108.8	C14—C17—C16	101.9 (3)
H7A—C7—H7B	107.7	C14—C17—H17A	111.4
C14—C8—C7	111.5 (3)	C16—C17—H17A	111.4
C14—C8—C9	108.1 (3)	C14—C17—H17B	111.4
C7—C8—C9	111.0 (3)	C16—C17—H17B	111.4
C14—C8—H8	108.7	H17A—C17—H17B	109.3
C7—C8—H8	108.7	C10—C18—H18A	109.5
C9—C8—H8	108.7	C10—C18—H18B	109.5
C11—C9—C8	112.8 (3)	H18A—C18—H18B	109.5
C11—C9—C10	113.2 (3)	C10—C18—H18C	109.5
C8—C9—C10	111.2 (3)	H18A—C18—H18C	109.5
C11—C9—H9	106.3	H18B—C18—H18C	109.5
C8—C9—H9	106.3	C6—C19—H19A	119 (2)
C10—C9—H9	106.3	C6—C19—H19B	120 (3)
C1—C10—C5	112.0 (3)	H19A—C19—H19B	120 (4)
C1—C10—C18	106.7 (3)	C13—C20—H20A	109.5
C5—C10—C18	108.1 (3)	C13—C20—H20B	109.5
C1—C10—C9	111.5 (3)	H20A—C20—H20B	109.5
C5—C10—C9	107.6 (3)	C13—C20—H20C	109.5
C18—C10—C9	111.1 (3)	H20A—C20—H20C	109.5
C9—C11—C12	113.5 (3)	H20B—C20—H20C	109.5
C9—C11—H11A	108.9	N1—C21—N2	116.5 (3)
C12—C11—H11A	108.9	N1—C21—S1	122.2 (3)
C9—C11—H11B	108.9	N2—C21—S1	121.2 (3)
C12—C11—H11B	108.9
C10—C1—C2—C3	−0.5 (6)	C8—C9—C10—C5	−57.4 (3)
C1—C2—C3—O1	178.1 (4)	C11—C9—C10—C18	−67.4 (4)
C1—C2—C3—C4	−2.5 (6)	C8—C9—C10—C18	60.8 (4)
O1—C3—C4—C5	−177.3 (4)	C8—C9—C11—C12	52.1 (4)
C2—C3—C4—C5	3.3 (6)	C10—C9—C11—C12	179.5 (3)
C3—C4—C5—C6	−178.6 (4)	C9—C11—C12—C13	−52.9 (4)
C3—C4—C5—C10	−1.1 (6)	C11—C12—C13—C15	169.0 (3)
C4—C5—C6—C19	−50.9 (5)	C11—C12—C13—C14	55.7 (4)
C10—C5—C6—C19	131.5 (4)	C11—C12—C13—C20	−70.6 (4)
C4—C5—C6—C7	130.1 (4)	C7—C8—C14—C13	179.8 (3)
C10—C5—C6—C7	−47.5 (4)	C9—C8—C14—C13	57.6 (4)
C19—C6—C7—C8	−135.2 (4)	C7—C8—C14—C17	−55.8 (4)
C5—C6—C7—C8	43.8 (4)	C9—C8—C14—C17	−178.0 (3)
C6—C7—C8—C14	−169.2 (3)	C12—C13—C14—C8	−61.0 (4)
C6—C7—C8—C9	−48.7 (4)	C15—C13—C14—C8	176.1 (3)
C14—C8—C9—C11	−52.3 (4)	C20—C13—C14—C8	64.6 (4)
C7—C8—C9—C11	−174.9 (3)	C12—C13—C14—C17	165.7 (3)
C14—C8—C9—C10	179.2 (3)	C15—C13—C14—C17	42.8 (4)
C7—C8—C9—C10	56.7 (3)	C20—C13—C14—C17	−68.7 (4)
C2—C1—C10—C5	2.6 (5)	C12—C13—C15—O2	29.5 (6)
C2—C1—C10—C18	−115.5 (4)	C14—C13—C15—O2	148.1 (4)
C2—C1—C10—C9	123.2 (4)	C20—C13—C15—O2	−94.4 (5)
C4—C5—C10—C1	−1.8 (5)	C12—C13—C15—C16	−147.9 (4)
C6—C5—C10—C1	175.8 (3)	C14—C13—C15—C16	−29.3 (4)
C4—C5—C10—C18	115.4 (4)	C20—C13—C15—C16	88.1 (4)
C6—C5—C10—C18	−67.0 (4)	O2—C15—C16—C17	−172.5 (4)
C4—C5—C10—C9	−124.5 (3)	C13—C15—C16—C17	5.0 (5)
C6—C5—C10—C9	53.0 (4)	C8—C14—C17—C16	−168.7 (3)
C11—C9—C10—C1	51.4 (4)	C13—C14—C17—C16	−40.1 (4)
C8—C9—C10—C1	179.6 (3)	C15—C16—C17—C14	21.4 (5)
C11—C9—C10—C5	174.5 (3)

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
N1—H1A···S1^i	0.86	2.75	3.511 (4)	149
N1—H1B···O2^ii	0.86	2.01	2.867 (4)	177
N2—H2A···S1^i	0.86	2.42	3.252 (3)	164
N2—H2B···O1	0.86	1.96	2.821 (4)	174
C4—H4···O2^iii	0.93	2.49	3.336 (5)	151

Symmetry codes: (i) −x+1, y−1/2, −z; (ii) x+1, y, z−1; (iii) −x, y+1/2, −z+1.