1,1-Bis[4-(trifluoromethyl)phenyl]germetane

The internal C—Ge—C bond angle in the germacyclobutane ring of the title compound, C17H14F6Ge or [Ge(C3H6)(C7H4F3)2], is 77.8 (3)°. The –CF3 groups display rotational disorder [occupancies 0.604 (14):0.396 (14) and 0.410 (6):0.411 (6):0.179 (3)] and the germacyclobutane ring also shows disorder [occupancies 0.604 (14):0.396 (14)].

We thank the Natural Sciences and Engineering Research Council (NSERC) of Canada for financial support and for a postgraduate scholarship to LAH. supporting information Acta Cryst. (2009)

S1. Comment
Photolysis of 1,1-diarylgermetanes results in two competing modes of cycloreversion: (2 + 2) to yield the corresponding 1,1-diarylgermene (R 2 Ge=CH 2 ) and ethylene, and (3 + 1) to yield the corresponding diarylgermylene (R 2 Ge) and cyclopropane (Leigh et al., 2008). While the quantum yield for decomposition of the germetane remains roughly constant regardless of aromatic substitutent (Φ ca 0.10), decreasing the electron density at germanium by aromatic ring substitution favors formation of the germylene and cyclopropane (Leigh et al., 2008). The molecular structure of (I) is shown in Figure 1. The internal C-Ge-C bond angle is 77.8 (3)°.
To our knowledge, there are no other reported crystal structures of germacyclobutanes (i.e. those in which the three carbon atoms of the four membered ring are all saturated) with which to compare these data. There are, however, several reported crystal structures of germacyclobutenes to which limited comparisons can be made (Tokitoh et al., 1995;Eichler et al., 1999;Meiners et al., 2002;Tajima et al., 2005). The endocyclic Ge-C bond distances are quite similar, despite the carbon-carbon double bond within the ring, and the endocyclic C-Ge-C angle is slightly smaller in these germacyclobutene molecules, which is to be expected. The 4-membered germacyclobutane ring of the title compound is not planar.
The angle between the plane made by the carbon atoms in the ring (C15-C17) and the plane containing the germanium and two adjacent carbon atoms in the ring (C15, C17) is 24 (2)°. The aromatic rings are nearly perpendicular, with an angle of 85.8 (1)° between them.
The trifluoromethyl groups are disordered and have been refined as such (see refinement details). The two CF 3 groups interact with each other and the germacyclobutane ring of neighbouring molecules (see the packing diagram - Figure 2).
The non-planarity of the germacyclobutane ring results in disorder that has been refined over two occupancy sites.

S2. Experimental
Compound (I) was synthesized as described elsewhere (Leigh et al., 2008). Single crystals for x-ray diffraction were obtained via repeated recrystallizations from methanol.

S3. Refinement
The data were collected at -100°C on a single-crystal mounted in a cryoloop. X-ray crystallographic analysis was performed at the McMaster Analytical X-Ray (MAX) Diffraction Facility. Hydrogen atoms were treated as riding upon their parent atoms with C-H distances of 0.95 Å (aromatic) and 0.96 Å (CH 2 ) and with U iso (H) = 1.2U eq (C). Fluorine atoms at C7/C7A were split over two positions, with approximately 60:40 occupancy. Fluorine atoms at C14/C14A/C14′ were disordered over three positions (41:41:18 occupancy) with one thermal parameter (0.06). The propyl moiety of the germacycle is disordered over two sites (60:40 occupancy) and a restraint is applied so that the groups can twist away supporting information from each other but not leave the sphere of germanium.

Figure 1
The molecular structures of (I) drawn with 50% probability ellipsoids. H atoms not shown. Both trifluoromethyl groups and the (CH 2 ) 3 group are disordered but the disorder is not shown for clarity.

Figure 2
Packing diagram of (I) in the space group P 2 1 /n. Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.