Tetraphenylarsonium iodide, a new determination

The reported crystal structures of Ph₄P⁺·I^- and Ph₄As⁺·I^- have been re-examined. An apparent instance of substitutional dimorphism could not be reproduced and, contrary to an earlier report, tetra­phenyl­arsonium iodide, [As(C₆H₅)₄]I or Ph₄As⁺·I^-, was found to be isostructural with the phosphorus compound. The cation and anion are both located on symmetry sites. The crystal packing involves linear chains of cations in fourfold phenyl embraces.

S1. Comment

A search of the Cambridge Structural Database (CSD, Version?; Allen, 2002) for Ph₄E⁺·I⁻ revealed three structure reports for Ph₄P⁺·I⁻ and Ph₄As⁺·I⁻, all in space group I4, and with cell dimensions as follows. For Ph₄P⁺·I⁻ with CSD refcode SATSUB [crystallized from CH₂Cl₂-hexane (Schweizer et al., 1989)], a = b = 11.979 and c = 6.981 Å. For Ph₄P⁺·I⁻ with CSD refcode SATSUB01 [crystallized from CH₃CN (Gorobinsky et al., 2001)], a = b = 11.967 and c = 6.974 Å. For Ph₄As⁺·I⁻ with CSD refcode TPHASI [crystallized from hot water (Mooney, 1940)], a = b = 12.194 and c = 7.085 Å. The agreement factors for these structure determinations are 0.037, 0.046 and 0.280, respectively. The crystal forms of both Ph₄P⁺·I⁻ and Ph₄As⁺·I⁻, obtained by us from various solvents and at a range of temperatures, are isostructural with that reported previously for Ph₄P⁺·I⁻. Only our determination of the structure of the arsonium salt, (I), is described here. \sch

The space-group for (I) found in the present work is the same as that found by Mooney. The cell dimensions differ slightly, but significantly. In (I), the cation necessarily has S₄ symmetry. The difference between the original structure and the present one lies in the (one) torsion angle of the phenyl groups in this cation, which is defined here as that described by the molecular S₄ axis and the As—C—C orientation. This angle is 88° for the original structure determination but 63.6 (2)° in our determination (Fig 1). In Ph₄As⁺·I₃⁻ (Runsink et al., 1972), the cation possesses twofold symmetry and the two independent torsion angles are 51.8 and 58.5° (relative to the molecular twofold axis).

The difference in molecular conformations in the original and the present determination of (I) is illustrated in Fig 2. The almost orthogonal conformation of phenyl rings in the original structure is what we have previously described as the `flipper' conformation (Dance & Scudder, 2000): the planes of a pair of phenyl rings are not twisted relative to each other. In the present work, the phenyl rings are different from the original structure, by concerted torsion through 24°. The crystal packing in our structure of (I) contains linear infinite chains of fourfold phenyl embraces (LIT4PE) in the c direction. Along this chain, adjacent cations form the orthogonal fourfold phenyl embrace, O4PE, in which there is a concerted cycle of four edge-to-face (EF) phenyl···phenyl interactions (Dance & Scudder, 1996a,b). In addition, there are EF interactions which link the chains laterally. The I atom is located in a hydrogenous environment, taking part in a total of 24 C—H···I interactions in the range 3.7–4.2 Å, each substantially longer than the sum of the van der Waals radii (3.2 Å) and with linearity in the range 103–126°, but sufficiently numerous to be considered a stabilizing influence.

We attempted to clarify whether our crystal of (I) was indeed a subtle dimorph of that produced by Mooney, in which case her crystal could be classified as an example of a `disappearing polymorph' (Dunitz & Bernstein, 1995; Bombicz et al., 2003). We have probed the existence of dimorphs for (I) from several points of view.

The original crystal structure, although carried out using photographic data, does not appear to be in error. The cell parameters were determined from powder diffraction data and are likely to be quite accurate. Mooney measured the crystal density and found it to be identical to that calculated (and significantly different from our measured and calculated density). There are no unexpectedly short bond lengths which might have resulted from misreading or misinterpreting coordinates. We analysed the crystal packing of both the original and the new crystal structures, looking for any unreasonably short intermolecular contacts which might suggest error. There were none. We took the original coordinates as a starting point for refinement using our diffraction data. Not surprisingly, the final result was identical to ours. There is no alternative solution for our data.

We tested the possibility that alternative crystal forms with different torsions of the phenyl rings might occur, or transform, as a result of temperature variation, by making powder diffraction measurements in the range 299–373 K. The reflections 132, 310, 330 and 112 were identified and monitored. The 2θ variations over the temperature range studied were as follows. For reflection 132, 34.98 (at 299 K) to 34.71° (at 373 K); for reflection 310, 23.08–22.99°; for reflection 330, 31.17–31.08°; for reflection 112, 27.86–27.64°, consistent with thermal expansion but not indicating a phase change over the temperature range studied. The calculated 2θ values from our (and Mooney's) crystal structures are: 132 34.99° (34.38°); 310 23.13° (23.06°); 330 31.20° (31.12°); 112 27.93° (27.21°). Our powder results are consistent with our crystal structure but significantly different from those computed from Mooney's structure. Temperatures were given simply as ° in the original CIF, leading to potential confusion. °C was assumed. Please check that the correct numbers have been converted to the SI Kelvin scale, here and in the Experimental section.

We analysed all other (cationic, anionic and neutral) Ph₄E structures in the CSD which crystallize in tetragonal space groups with E located on a 4 site and contiguous Ph₄E along c separated by 6.8–7.2 Å. These structures all have the molecular packing type which incorporates LIT4PE chains with E–E (the tetragonal axial length c) about 7 Å. The distribution of phenyl-ring torsion angles is shown in Fig 3. The black bars correspond to our determination of the structure of (I) and that of Mooney, with the latter being the outlier at 88°. This unusual result is the only one that might lead to some suspicion about the accuracy of the earlier determination.

Despite considering evidence of various types, we leave open to question the existence or otherwise of dimorphs of (I).

S2. Experimental

Commercial Ph₄P⁺·I⁻ was recrystallized by slow cooling of a hot aqueous solution, to give colourless needles which were shown (by X-ray diffraction) to have unit-cell parameters and atomic coordinates identical to those published for Ph₄P⁺·I⁻ crystallized from CH₂Cl₂-hexane. Commercial Ph₄As⁺·Cl⁻·H₂O was converted into Ph₄As⁺·I⁻ by dissolving equimolar amounts of NaI and Ph₄As⁺·Cl⁻·H₂O in MeOH, and then precipitating NaCl by the addition of acetone. The mixture was filtered and the mixed solvent was removed on a rotary evaporator. Dichloromethane was added to the crude solid product, the mixture was filtered and the solvent was removed from the filtrate on the rotary evaporator. The final solid product was crystallized from both hot water (as colourless needles; analysis found: C 56.28, H 3.99%; calculated: C 56.49, H 3.95%) and CH₂Cl₂-hexane (Ratio?; as colourless prisms). These gave crystals with identical unit-cell parameters, and preliminary structure determinations on both crystals indicated that they had the same atomic coordinates. The full X-ray analysis was carried out on the crystals of (I) obtained from CH₂Cl₂-hexane. A site population analysis during the data analysis confirmed that the material was at least 99% pure on a molar basis; no evidence was found for the presence of Cl⁻. Powder diffraction data for (I) were measured using Cu Kα radiation (λ 1.5406 Å) on a Bruker D5005 X-ray powder diffractometer at 16 temperatures in the range 299–373 K.

S3. Refinement

As and I atoms were refined anisotropically, while the phenyl group was refined with thermal motion described by a 15-parameter TLX group (where T is the translation tensor, L is the libration tensor and X is the origin of libration, set at the As atom). The phenyl ring was modelled as a planar group with refineable mm2 symmetry. The H atoms of the cation were included in calculated positions and were included in the group thermal for the phenyl ring. The enantiomer was confirmed, the alternative giving R = 0.039 and wR = 0.054.

Computing details

Data collection: CAD-4 Software (Enraf-Nonius, 1989); cell refinement: CAD-4 Software; data reduction: local program; program(s) used to solve structure: SIR92 (Altomare et al., 1994); program(s) used to refine structure: RAELS (Rae, 2000); molecular graphics: ORTEPII (Johnson, 1976) and CrystalMaker (Palmer, 2003); software used to prepare material for publication: local programs.

Figures

	Fig. 1. A view of (I), showing the atom-numbering scheme and with displacement ellipsoids drawn at the 30% probability level.
	Fig. 2. The conformations of the phenyl rings in Ph4As^+ as found in (a) the original (Mooney, 1940) and (b) the present determination of the crystal structure of (I). One ring is coloured black to emphasize the difference in the ring orientations.
	Fig. 3. Histogram of phenyl-ring torsion angles (°) for species Ph4E with S4 symmetry in tetragonal space groups in the CSD. The black outlier is from the original determination (Mooney, 1940) of (I), and the other black entry is for the present determination of (I).

tetraphenylarsonium iodide

Crystal data

[As(C6H5)4]I	Dx = 1.66 Mg m^−3 Dm = 1.66 Mg m^−3 Dm measured by flotation in what
Mr = 510.3	Mo Kα radiation, λ = 0.71073 Å
Tetragonal, I4	Cell parameters from 10 reflections
a = 12.174 (2) Å	θ = 13–15°
c = 6.885 (2) Å	µ = 3.16 mm^−1
V = 1020.4 (4) Å^3	T = 294 K
Z = 2	Prism, colourless
F(000) = 500.0	0.24 × 0.12 × 0.08 mm

Data collection

Enraf-Nonius CAD-4 diffractometer	Rint = 0.008
ω/2θ scans	θmax = 25°
Absorption correction: analytical (De Meulenaer & Tompa, 1965)	h = 0→14
Tmin = 0.70, Tmax = 0.77	k = 0→14
539 measured reflections	l = 0→8
532 independent reflections	1 standard reflections every 30 min
499 reflections with I > 2σ(I)	intensity decay: none

Refinement

Refinement on F	w = 1/[σ^2(F) + 0.0004F^2]
R[F^2 > 2σ(F^2)] = 0.022	(Δ/σ)max = 0.001
wR(F^2) = 0.030	Δρmax = 0.25 e Å^−3
S = 1.13	Δρmin = −0.54 e Å^−3
499 reflections	Absolute structure: Anomalous scattering
28 parameters	Absolute structure parameter: 0.13 (3)
H-atom parameters constrained

Special details

Refinement. The phenyl was refined as a planar group with mm2 symmetry, and a 12-parameter TL rigid-body thermal parameter with the centre of libration at the As atom was used for its thermal motion. As and I were refined with single atom anisotropic thermal parameters. The enantiomer was confirmed, the alternative giving R 0.039, wR 0.054.

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

	x	y	z	Uiso*/Ueq
As1	0.5000	0.5000	0.5000	0.0323 (2)
C1	0.5720 (3)	0.6049 (3)	0.6615 (5)	0.0362 (8)
C2	0.5416 (3)	0.7143 (3)	0.6605 (6)	0.0481 (7)
C3	0.5917 (3)	0.7869 (3)	0.7882 (6)	0.059 (1)
C4	0.6709 (3)	0.7494 (3)	0.9142 (6)	0.060 (1)
C5	0.7013 (3)	0.6407 (4)	0.9155 (6)	0.066 (1)
C6	0.6516 (3)	0.5675 (3)	0.7883 (5)	0.0522 (9)
I1	1.0000	0.5000	0.7500	0.0782 (3)
HC2	0.4839	0.7409	0.5685	0.064
HC3	0.5702	0.8662	0.7886	0.080
HC4	0.7068	0.8020	1.0060	0.074
HC5	0.7590	0.6144	1.0078	0.095
HC6	0.6733	0.4883	0.7884	0.069

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
As1	0.0319 (3)	0.0319 (3)	0.0331 (4)	0.0000	0.0000	0.0000
C1	0.035 (1)	0.040 (1)	0.034 (1)	−0.0043 (9)	0.004 (1)	0.001 (1)
C2	0.040 (1)	0.041 (1)	0.063 (2)	−0.0011 (9)	0.000 (1)	−0.009 (1)
C3	0.051 (2)	0.052 (1)	0.073 (2)	−0.008 (1)	0.005 (2)	−0.021 (1)
C4	0.070 (2)	0.064 (1)	0.047 (2)	−0.025 (1)	0.001 (1)	−0.009 (1)
C5	0.083 (2)	0.062 (2)	0.052 (2)	−0.028 (1)	−0.025 (2)	0.011 (1)
C6	0.059 (2)	0.048 (1)	0.049 (2)	−0.014 (1)	−0.018 (1)	0.010 (1)
I1	0.0718 (4)	0.0718 (4)	0.0911 (7)	0.0000	0.0000	0.0000

Geometric parameters (Å, º)

As1—C1	1.907 (4)	C4—C5	1.375 (4)
As1—C1^i	1.907 (4)	C5—C6	1.387 (5)
As1—C1^ii	1.907 (4)	C2—HC2	1.0
As1—C1^iii	1.907 (4)	C3—HC3	1.0
C1—C2	1.382 (4)	C4—HC4	1.0
C1—C6	1.382 (4)	C5—HC5	1.0
C2—C3	1.387 (5)	C6—HC6	1.0
C3—C4	1.375 (4)
C1—As1—C1^i	108.7 (2)	C4—C5—C6	119.8 (4)
C1—As1—C1^ii	109.9 (1)	C1—C6—C5	119.5 (3)
C1—As1—C1^iii	109.9 (1)	C1—C2—HC2	120.2
C1^i—As1—C1^ii	109.9 (1)	C1—C6—HC6	120.2
C1^i—As1—C1^iii	109.9 (1)	C2—C3—HC3	120.2
C1^ii—As1—C1^iii	108.7 (2)	C3—C4—HC4	119.6
As1—C1—C2	121.3 (3)	C3—C2—HC2	120.2
As1—C1—C6	118.0 (3)	C4—C5—HC5	120.2
C2—C1—C6	120.6 (4)	C4—C3—HC3	120.2
C1—C2—C3	119.5 (3)	C5—C6—HC6	120.2
C2—C3—C4	119.8 (4)	C5—C4—HC4	119.6
C3—C4—C5	120.8 (5)	C6—C5—HC5	120.2

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