A mixed-valence iron complex of the antitumour drug 6-mercaptopurine: tris(6-mercaptopurine)iron(II) tetrachloroferrate(III) chloride

Single crystals of the title complex, tris(1,6-di­hydro-9H-purine-6-thione-N⁷,S)­iron(II) tetra­chloro­ferrate(III) chloride, [Fe(C₅H₄N₄S)₃][FeCl₄]Cl, were grown on the surface of solid 6-mercaptopurine monohydrate pellets in a solution of iron(III) chloride. The solution of the hexagonal structure required the application of twin refinement techniques. All the component ions lie on threefold rotation axes. The complex contains distorted octahedral [Fe(C₅H₄N₄S)₃]²⁺ cations with three N7/S6-chelating neutral 6-mercaptopurine ligands, tetrahedral [FeCl₄]^- anions with a mean Fe-Cl distance of 2.189 (1) Å, and free chloride ions.

S1. Comment

6-mercaptopurine (6-MP), the synthetic thio analogue of the naturally occurring purine derivative hypoxanthine, is a clinical agent routinely administered for the therapy of human leukaemia. 6-MP is converted intracellularly to the corresponding ribonucleotide, which inhibits purine biosynthesis (Elion, 1989). Metal complexes of 6-MP and related drugs are of special interest in view of their possibly enhanced therapeutic effect with respect to the free base, which may also be a consequence of the protection of the molecule from enzymatic biological degradation by metal complexation. Metal complexes of drug molecules could, in addition, be used as slow-release drugs (Farrell, 1989). Preliminary investigations by UV spectroscopy of the enzymatic degradation of 6-MP in aqueous solutions by xanthine oxidase, supported by catalase, show a significant decrease of the decomposition reaction rate of 6-MP in the form of the title iron complex, (I), compared with the corresponding rate of the free ligand. \sch

A review of metal complexes of sulfur-containing purine derivatives has recently been given by Dubler (1996). It has been shown that, depending on the protonation status of the ligand and on the hardness or softness of the corresponding metal atom, 6-MP variously coordinates monodentately through S6, N7 or N3, by chelating through N7/S6 and/or by bridging two metal atoms through S6, S6/N7 or N1/N7. As part of a program elucidating the coordination properties of oxo- and thiopurines (Zhu et al., 1998), we present here the synthesis and structure of the title complex, (I).

There are three building units in the structure (Fig.1): octahedral [Fe(C₅H₄N₄S)₃]²⁺ cations with three N7/S6 chelating neutral 6-MP ligands, tetrahedral [FeCl₄]⁻ anions and free chloride anions. The FeN₃S₃-octahedron is significantly distorted, with octahedral angles ranging from 83.39 (5) to 95.44 (5)°. The neutral 6-MP ligand is protonated at N1 and N9, but not at the coordinating N7 atom. A common feature of most coordinating purine bases is a slight non-planarity in the pseudoaromatic ring system. Small deviations from planarity occur also in Fe₂(6-MP)₃Cl₅, where the maximum distances of an atom from the best plane through the nine purine ring atoms are −0.029 (2) for N9 and 0.025 (2) Å for N7. The `bite distance' S6···N7 of the chelating 6-MP ligand is 3.179 (2) Å, compared with the corresponding distance of 3.342 (1) Å for the non-coordinating molecule in 6-mercaptopurine monohydrate (Sletten et al., 1969). This fact is in agreement with the observation that the decrease of the bite distance is more pronounced the smaller the metal atom is. The distance may vary from about 3.32 Å in Cd^II-complexes to 3.04 Å in Cu^II complexes (Dubler & Gyr, 1988).

The [Fe(C₅H₄N₄S)₃]²⁺ cations are separated from each other by the [FeCl₄]⁻ anions, and there is no evidence for stacking interactions between the purine planes. The mean Fe—Cl distance in the [Fe^IIICl₄]⁻ anion is 2.189 (1) Å. This distance clearly indicates the Fe^III oxidation state of the central atom, since for a tetrahedral [Fe^IICl₄]²⁻ anion the expected mean bonding distance is about 2.30 Å (query) (Pelizzi et al., 1977).

The [Fe(C₅H₄N₄S)₃]²⁺ cations and the free Cl⁻ anions are connected through a network of hydrogen bonds of the type N—H···Cl, where the Cl⁻ anion acts as a six-coordinated acceptor. In addition, weak interactions of the type C—H···S are observed (Table 2 and Fig. 2).

S2. Experimental

A solution of FeCl₃·6H₂O (270 mg, 1 mmol) in ethanol (20 ml) and 1 N HCl (1 ml) was overlaid on a solid pill of 6-MP·H₂O and stored at a temperature of 333 K in a closed vessel. After 2 d red-brown hexagonal prismatic crystals of (I) suitable for X-ray analysis had grown on the surface of the pill. Crystals of the same composition were also synthesized by heating a solution of 6-MP·H₂O (681 mg, 4 mmol) and FeCl₃·6H₂O (1081 mg, 4 mmol) in 1 N HCl (2 ml) and ethanol (100 ml). The reaction mixture was filtered and kept at a temperature of 333 K for crystallization. Within a few hours, a red-brown microcrystalline product, (I), could be separated from the solution.

S3. Refinement

The structure had been refined previously (ref?) with SHELX76 (Sheldrick, 1976) on F_o based on room-temperature measurements (three different crystals were used), with relatively bad R values of about 0.08. Improvement of refinement techniques (e.g. SHELXL97) and the possibility of user-friendly twin refinements with SHELXL97 (Sheldrick, 1997; Herbst-Irmer & Sheldrick, 1998) led us to remeasure one of the example crystals at 173 (1) K with an image-plate detector system (Stoe IPDS) in order to resolve this old problem structure. Repeating the refinement of the known model with the new data set showed the same result for R1 = 0.09, wR2 = 0.143, based on F_o² using all unique reflections. Application of the twin matrix (010, 100, 00–1) led to a remarkable improvement of the final R values, as shown in the Experimental section; the twinning ratio was 0.286:0.714. The positions of the H atoms were determined from difference electron-density maps but they were finally calculated after each cycle of refinement using a riding model.

Computing details

Data collection: IPDS (Stoe & Cie, 1999); cell refinement: IPDS; data reduction: X-RED in IPDS; program(s) used to solve structure: SHELXS97 (Sheldrick, 1990); program(s) used to refine structure: SHELXL97 (Sheldrick, 1997); molecular graphics: PLATON (Spek, 1990) and PLUTON (Spek, 1991); software used to prepare material for publication: SHELXL97.

Figures

	Fig. 1. The PLATON (Spek, 1990) displacement ellipsoid plot of (I) at the 50% probability level with the atom-numbering scheme. H atoms are drawn as small spheres of arbitrary radii.
	Fig. 2. The PLUTON (Spek, 1991) packing diagram for (I) showing the hydrogen-bonding interactions between the 6-MP ligands and the Cl^− ion.

tris(1,6-dihydro-9H-purine-6-thione-N⁷,S)iron(II) tetrachloroferrate(III) chloride

Crystal data

[Fe(C5H4N4S)3][FeCl4]Cl	Dx = 1.820 Mg m^−3
Mr = 745.50	Mo Kα radiation, λ = 0.71073 Å
Trigonal, P63	Cell parameters from 7998 reflections
Hall symbol: P 6c	θ = 3.6–30.4°
a = 10.3499 (6) Å	µ = 1.82 mm^−1
c = 14.6652 (11) Å	T = 173 K
V = 1360.48 (15) Å^3	Prism, red-brown
Z = 2	0.28 × 0.25 × 0.24 mm
F(000) = 742

Data collection

STOE IPDS diffractometer	2717 independent reflections
Radiation source: fine-focus sealed tube	2484 reflections with I > 2σ(I)
Graphite monochromator	Rint = 0.031
ϕ rotation scan	θmax = 30.4°, θmin = 3.6°
Absorption correction: numerical (Coppens et al., 1965)	h = −12→14
Tmin = 0.630, Tmax = 0.669	k = −14→14
16156 measured reflections	l = −20→20

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)] = 0.021	w = 1/[σ^2(Fo^2) + (0.01P)^2] where P = (Fo^2 + 2Fc^2)/3
wR(F^2) = 0.038	(Δ/σ)max < 0.001
S = 1.10	Δρmax = 0.26 e Å^−3
2717 reflections	Δρmin = −0.24 e Å^−3
114 parameters	Extinction correction: SHELXL97 (Sheldrick, 1997), Fc^*=kFc[1+0.001xFc^2λ^3/sin(2θ)]^-1/4
1 restraint	Extinction coefficient: 0.0107 (4)
Primary atom site location: structure-invariant direct methods	Absolute structure: Flack (1983); 1306 Friedel pairs
Secondary atom site location: difference Fourier map	Absolute structure parameter: −0.05 (2)

Special details

Experimental. The crystal represents a twin with twin law (010, 100, 00–1), refined with the twin option of SHELXL97.

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. The weighted R-factor wR and goodness of fit S are based on F², conventional R-factors R are based on F, with F set to zero for negative F². The threshold expression of F² > σ(F²) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F² are statistically about twice as large as those based on F, and R- factors based on ALL data will be even larger.

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

	x	y	z	Uiso*/Ueq
Fe2	2/3	1/3	0.40989 (4)	0.02359 (12)
Cl1	2/3	1/3	−0.16182 (5)	0.01632 (15)
Cl2	2/3	1/3	0.55907 (7)	0.0501 (3)
Cl3	0.43881 (8)	0.25114 (9)	0.36101 (5)	0.0479 (2)
Fe1	0	0	0.15255 (3)	0.01517 (9)
S6	−0.23600 (6)	−0.12766 (8)	0.25433 (3)	0.01852 (10)
N3	−0.4658 (3)	−0.5425 (3)	0.07443 (13)	0.0312 (5)
N7	−0.1220 (2)	−0.1982 (2)	0.06987 (11)	0.0177 (4)
C8	−0.1196 (3)	−0.2637 (2)	−0.00670 (13)	0.0200 (4)
H8	−0.0390	−0.2212	−0.0487	0.024*
N1	−0.4545 (2)	−0.4048 (2)	0.20667 (13)	0.0254 (4)
H1	−0.5018	−0.4051	0.2566	0.030*
C6	−0.3192 (3)	−0.2824 (3)	0.18914 (14)	0.0177 (4)
N9	−0.2433 (2)	−0.3970 (2)	−0.01909 (12)	0.0221 (4)
H9	−0.2627	−0.4564	−0.0663	0.026*
C5	−0.2565 (2)	−0.2976 (3)	0.10827 (14)	0.0168 (5)
C4	−0.3339 (3)	−0.4234 (3)	0.05506 (15)	0.0215 (5)
C2	−0.5206 (3)	−0.5268 (3)	0.1512 (2)	0.0344 (6)
H2	−0.6144	−0.6071	0.1699	0.041*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
Fe2	0.02305 (17)	0.02305 (17)	0.0247 (2)	0.01152 (8)	0	0
Cl1	0.0183 (2)	0.0183 (2)	0.0123 (3)	0.00916 (12)	0	0
Cl2	0.0625 (5)	0.0625 (5)	0.0252 (4)	0.0312 (2)	0	0
Cl3	0.0291 (4)	0.0554 (5)	0.0562 (4)	0.0188 (4)	−0.0125 (3)	−0.0111 (4)
Fe1	0.01680 (13)	0.01680 (13)	0.01192 (18)	0.00840 (6)	0	0
S6	0.0209 (2)	0.0208 (3)	0.01322 (16)	0.0099 (3)	0.0010 (2)	−0.0031 (3)
N3	0.0252 (13)	0.0259 (13)	0.0347 (10)	0.0070 (9)	0.0001 (12)	−0.0135 (12)
N7	0.0211 (11)	0.0191 (9)	0.0130 (7)	0.0102 (8)	−0.0003 (7)	−0.0012 (7)
C8	0.0256 (14)	0.0255 (12)	0.0132 (7)	0.0159 (11)	0.0010 (10)	−0.0017 (8)
N1	0.0229 (11)	0.0257 (12)	0.0213 (8)	0.0075 (10)	0.0042 (8)	−0.0056 (9)
C6	0.0180 (11)	0.0200 (11)	0.0155 (9)	0.0099 (10)	−0.0020 (9)	−0.0016 (8)
N9	0.0269 (10)	0.0276 (11)	0.0159 (7)	0.0167 (9)	−0.0029 (8)	−0.0098 (9)
C5	0.0186 (11)	0.0158 (11)	0.0171 (10)	0.0094 (9)	−0.0028 (8)	−0.0032 (8)
C4	0.0230 (11)	0.0207 (12)	0.0238 (10)	0.0133 (11)	−0.0044 (10)	−0.0076 (9)
C2	0.0223 (14)	0.0254 (16)	0.0439 (11)	0.0034 (12)	0.0022 (13)	−0.0077 (13)

Geometric parameters (Å, º)

Fe2—Cl2	2.1878 (11)	N7—C8	1.318 (3)
Fe2—Cl3^i	2.1892 (7)	N7—C5	1.371 (3)
Fe2—Cl3^ii	2.1892 (7)	C8—N9	1.345 (3)
Fe2—Cl3	2.1892 (7)	C8—H8	0.9500
Fe1—N7^iii	2.1638 (19)	N1—C6	1.363 (3)
Fe1—N7	2.1638 (19)	N1—C2	1.364 (3)
Fe1—N7^iv	2.1638 (19)	N1—H1	0.8800
Fe1—S6^iv	2.5908 (6)	C6—C5	1.398 (3)
Fe1—S6	2.5908 (6)	N9—C4	1.371 (3)
Fe1—S6^iii	2.5909 (6)	N9—H9	0.8800
S6—C6	1.686 (2)	C5—C4	1.380 (3)
N3—C2	1.307 (3)	C2—H2	0.9500
N3—C4	1.335 (4)
Cl2—Fe2—Cl3^i	109.11 (3)	C8—N7—Fe1	144.36 (17)
Cl2—Fe2—Cl3^ii	109.11 (3)	C5—N7—Fe1	111.92 (13)
Cl3^i—Fe2—Cl3^ii	109.83 (2)	N7—C8—N9	112.9 (2)
Cl2—Fe2—Cl3	109.11 (3)	N7—C8—H8	123.5
Cl3^i—Fe2—Cl3	109.83 (2)	N9—C8—H8	123.5
Cl3^ii—Fe2—Cl3	109.83 (2)	C6—N1—C2	123.7 (2)
N7^iii—Fe1—N7	91.66 (7)	C6—N1—H1	118.2
N7^iii—Fe1—N7^iv	91.66 (7)	C2—N1—H1	118.2
N7—Fe1—N7^iv	91.66 (7)	N1—C6—C5	112.0 (2)
N7^iii—Fe1—S6^iv	171.45 (6)	N1—C6—S6	125.13 (17)
N7—Fe1—S6^iv	95.44 (5)	C5—C6—S6	122.85 (19)
N7^iv—Fe1—S6^iv	83.39 (5)	C8—N9—C4	107.62 (18)
N7^iii—Fe1—S6	95.44 (5)	C8—N9—H9	126.2
N7—Fe1—S6	83.39 (5)	C4—N9—H9	126.2
N7^iv—Fe1—S6	171.45 (6)	N7—C5—C4	111.59 (19)
S6^iv—Fe1—S6	90.122 (19)	N7—C5—C6	128.2 (2)
N7^iii—Fe1—S6^iii	83.39 (5)	C4—C5—C6	120.2 (2)
N7—Fe1—S6^iii	171.45 (6)	N3—C4—N9	129.4 (2)
N7^iv—Fe1—S6^iii	95.44 (5)	N3—C4—C5	126.5 (2)
S6^iv—Fe1—S6^iii	90.122 (19)	N9—C4—C5	104.1 (2)
S6—Fe1—S6^iii	90.123 (19)	N3—C2—N1	125.5 (2)
C6—S6—Fe1	93.51 (8)	N3—C2—H2	117.2
C2—N3—C4	112.1 (2)	N1—C2—H2	117.2
C8—N7—C5	103.72 (19)
N7^iii—Fe1—S6—C6	93.98 (10)	Fe1—S6—C6—C5	−3.42 (18)
N7—Fe1—S6—C6	2.93 (9)	N7—C8—N9—C4	−0.9 (3)
N7^iv—Fe1—S6—C6	−52.0 (4)	C8—N7—C5—C4	0.7 (3)
S6^iv—Fe1—S6—C6	−92.53 (8)	Fe1—N7—C5—C4	179.81 (15)
S6^iii—Fe1—S6—C6	177.34 (8)	C8—N7—C5—C6	−177.7 (2)
N7^iii—Fe1—N7—C8	80.6 (3)	Fe1—N7—C5—C6	1.4 (3)
N7^iv—Fe1—N7—C8	−11.1 (3)	N1—C6—C5—N7	−179.4 (2)
S6^iv—Fe1—N7—C8	−94.6 (3)	S6—C6—C5—N7	2.0 (3)
S6—Fe1—N7—C8	175.9 (3)	N1—C6—C5—C4	2.4 (3)
S6^iii—Fe1—N7—C8	135.1 (3)	S6—C6—C5—C4	−176.23 (19)
N7^iii—Fe1—N7—C5	−97.90 (18)	C2—N3—C4—N9	−179.0 (3)
N7^iv—Fe1—N7—C5	170.39 (16)	C2—N3—C4—C5	2.1 (4)
S6^iv—Fe1—N7—C5	86.87 (15)	C8—N9—C4—N3	−177.9 (3)
S6—Fe1—N7—C5	−2.62 (14)	C8—N9—C4—C5	1.3 (3)
S6^iii—Fe1—N7—C5	−43.5 (4)	N7—C5—C4—N3	177.9 (2)
C5—N7—C8—N9	0.1 (3)	C6—C5—C4—N3	−3.6 (4)
Fe1—N7—C8—N9	−178.45 (19)	N7—C5—C4—N9	−1.2 (3)
C2—N1—C6—C5	−0.4 (3)	C6—C5—C4—N9	177.29 (19)
C2—N1—C6—S6	178.2 (2)	C4—N3—C2—N1	0.2 (4)
Fe1—S6—C6—N1	178.1 (2)	C6—N1—C2—N3	−1.0 (5)

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

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
N1—H1···Cl1^v	0.88	2.48	3.273 (2)	150
N9—H9···Cl1^vi	0.88	2.38	3.231 (2)	164
C2—H2···S6^vii	0.95	2.78	3.693 (3)	161

Symmetry codes: (v) x−y−1, x−1, z+1/2; (vi) x−1, y−1, z; (vii) −x+y−1, −x−1, z.