The crystal structure of [Ir02((C6H5)2PCH2CH2P(c6H5)2)21 [PF,] has been redetermined from single-crystal X-ray diffraction data. The crystals were monoclinic of space group P21/n with four molecules in the unit cell of a = 17.18(1) A. [17.24 (3)], b = 16.46 (1) 8, [16.35 ( 3 ) ] , c = 16.97 (1) 8, [16.98 (3)], and 0 = 95.02 (5) ' [96.64 (2)Oj (parameters from the previous X-ray determination in square brackets). The R of 6.7% from 2177 observed reflections in the earlier refinement has been reduced to 4.4% from 4007 observed reflections in this work. The most striking variation in this redetermination compared with the earlier work is (i) an 0-0 bond length of 1.52 ( I ) A (cf. 1.625 (23) A) and (ii) chemically equivalent sets Qf Ir-P bonds have similar bond lengths and are 2.308 (3) and 2.342 (3) A (cf. 2.283 (7) and 2.452 (9) 8,) (equatorial) and 2.353 (3) and 2.366 (3) 8, (cf. 2.336 (7) and 2.349 (7) A) (axial). A study of crystal decomposition caused by either heating or prolonged irradiation shows that these large differences are caused by systematic changes in the reflection intensities brought about by the decomposition.From results of X-ray determination^^,^ on rhodium and iridium dioxygen complexes, one of the inferences drawn was that increasing the donor strength of the ligands favors longer 0-0 bonds. One crucial structure in this series was [Ir02(DPE)2]PF6 which in the original refinement3 had two unusual parameters: a long 0-0 bond length of 1.625 (23) A and two chemically equivalent Ir-P bonds differing greatly in length, Le., 2.283 (7) and 2.452 (9) A. Our crystallographic investigation of the complexes [M02L4]X (M = Rh, Ir, L = PMelPh, X = BPh4; M = Ir, L2 = bis(diphenylphosphinomethane), X = PF6, C104)4 and [ R~O~( A S M~~P~)~] C I O~~~~ together with others of iridicobalt,' and platinum8 have, however, found the 0-0 bond to be in a range approximately between 1.41 and 1.52 A, implying that the 0-0 bond length is essentially independent of the metal and ligand involved. These results also suggested that the two unusual parameters in the [Ir02(D-PE)2]PF6 determination were erroneous, and a redetermination of the structure was therefore undertaken. Experimental Section Crystal Data. [Ir02{P(C6H5)2CH2CH2P(CsH5)2~2] [PF,]: M = 1166.0; monoclinic; space group P21/n; a = 17.18, b = 16.46, c = 16.97 (hO.01) A; 0 = 95.02 (hO.05)O; V = 4781.6 A3; 2 = 4; d , = 1.61, d , = 1.62 g ~m -~; Mo K a X 0.7107 A, ~( M o K a ) = 31.93 cm-I.Collection of Intensity Data. The compound was prepared by the published m e t h~d .~ Recrystallization from dichloromethane-etha-no1 yielded orange prisms, many of which were twinned or of poor quality for crystallographic purposes. Intensity data were collected from a crystal (No. 1) of dimensions 0.28 X 0.32 X 0.40 mm on a Philips PW 1100 four-circle diffractometer for 0 between 3 and 20° with graphite-monochromated Mo K a radiation. Each reflec-C,, 0.4228 (6) C,,, 0.4742 (6) -0.0190 (7) C, , , -0.0163 (7) C,,, 0.0462 (6) C,, 0.4625 (6) C,,, 0.5422 (7) C,,, 0.5684 (8) C,,, 0...