2 PERKIN 2410 J. Chem. Soc., Perkin Trans. 2, 2000, 2410–2414 DOI: 10.1039/b006890j This journal is © The Royal Society of Chemistry 2000 Isolation and spectroscopic characterisation of C 60 F 17 CF 2 CF 3 and isomers of C 60 F 17 CF 3 ; insertion of :CF 2 into uorofullerene C–F bonds Olga V. Boltalina, a Peter B. Hitchcock, b Pavel A. Troshin, a Joan M. Street c and Roger Taylor* b a Chemistry Department, Moscow State University, Moscow 119899, Russia b The Chemistry Laboratory, CPES School, University of Sussex, Brighton, UK BN1 9QJ c Chemistry Department, The University, Southampton, UK SO17 1BJ Received (in Cambridge, UK) 23rd August 2000, Accepted 12th October 2000 First published as an Advance Article on the web 8th November 2000 The peruoroalkyluoro[60]fullerenes C 60 F 17 CF 2 CF 3 (C s ) and three isomers of C 60 F 17 CF 3 [C s (major, 65%), enantiomeric C 1 pair (minor, 35%)] have been separated by HPLC from the many products obtained by uorination of [60]fullerene by K 2 PtF 6 at ca. 465 °C. They have been characterised by 19 F NMR spectroscopy, and for the triuoromethyl compound, by single crystal X-ray structure determination. The triuoromethyl compound shows either extremely weak or no resonances for the CF 3 group in the 19 F NMR spectrum, which led to an earlier misidentication of the major isomer in this compound as a CF 2 derivative. Isolation of these compounds indicates that triuoromethyl compounds are formed under these conditions by insertion of :CF 2 groups (from fragmentation of other uorinated fullerene cages) into C–F bonds. Introduction Hydrogenation of fullerenes results in formation of methylene species from cage fragmentation, which add to other intact hydrogenated cages. This gives rise for example to the formation of species of 780 amu 1 [believed to be C 60 H 18 (CH 2 ) 3 , rather than C 60 H 60 ]. Fragmentation giving diuoromethylene species likewise accompanies uorination 2 (which like hydrogenation, is a radical addition process). The formation of methano- fullerenes (in which a methylene group is added across a fullerene ‘double’ bond) is well established, 3 hence the anal- ogous addition of diuoromethylene could be expected. Strong circumstantial evidence that diuoromethanofullerenes exist is obtained during mass spectrometry of C 60 (CF 3 H) n derivatives, which eliminate HF (20 amu) when the CF 3 and H addends are adjacent. 4 Recently, from the uorination of [60]fullerene by K 2 PtF 6 , we isolated by HPLC a fraction of 1112 amu from amongst the numerous uorofullerene products. 5 This is consistent with either C 60 F 18 CF 2 or C 60 F 17 CF 3 . Analysis by both 1 D and 2 D 19 F NMR spectroscopy showed the following features: 1. The fraction consisted of a mixture of C s (major) and C 1 (minor) components, in a 65 : 35 ratio. 2. Both components were based on the C 60 F 18 motif. 3. There were no resonances for the CF 3 group in the 19 F NMR spectrum, indicating that the major component could not be C 60 F 17 CF 3 , leaving C 60 F 18 CF 2 as the only feasible alternative. Moreover, addition of the CF 2 group across a double bond would give a three-membered ring (literature values are -143.2, -142.7 and -143.2 for CF 2 in such rings), 6 a peak at -143 ppm in the spectrum being attributed to this group. We have now isolated a further quantity of this material, obtained further 1 D and 2 D 19 F NMR spectra as well as a single crystal X-ray structure and nd, despite the NMR evidence, that it consists of three isomers of C 60 F 17 CF 3 viz. a major C s isomer and a minor enantiomeric pair of C 1 isomers. We have isolated also a species of 1162 amu which in principle could be either C 60 F 17 CF 2 CF 3 or C 60 F 18 (CF 2 ) 2 , but is unambiguously shown to be the former. We now describe the full characterisation of all of these components, propose a mechanism by which they are formed and draw attention to the major problem that arises in the 19 F NMR analysis of CF 3 -containing fullerenes. Experimental [60]Fullerene (ca. 300 mg) was uorinated at ca. 0.1 bar/465 °C. A dry toluene solution of the crude uorofullerene mixture was carefully ltered and puried by HPLC, (10 mm × 250 mm Cosmosil Buckyprep column) with toluene elution at a ow rate of 4.7 ml min -1 ( 1 ml min -1 for a 4.6 mm diameter column). This gave recovered [60]fullerene (ca. 75 mg), C 60 F 18 7 (ca. 200 mg) together with 20 other components in 2–10 mg yields. Mass spectra (EI 70 eV) The species eluting at 30 and 28 min gave parent ions of 1112 and 1162 amu (Fig. 1), respectively; the spectrum for the 1112 amu species was identical to that given previously. 1 Fig. 1 (C 60 F 17 C 2 F 5 ) solves an ongoing problem associated with uorofullerene mass spectra, namely the frequent appearance of fragmentation ions at 790 and 860 amu, which could arise from C 60 F 2 O 2 or C 60 CF 3 H, and C 60 (F 2 O 2 ) 2 or C 60 (CF 3 H) 2 , respectively. Whilst the oxides are rather improbable (given the parent compounds), the hydrogenated species often fail to show the expected M - 1 mass peak. These peaks are observed in the present case also, but the main fragmentation ion at 840 amu (which is not due to [70]fullerene because of the M + 1 isotope peak intensity and HPLC retention time) clearly arises from C 60 C 2 F 5 H, there being no mass-equivalent oxide. This fragment is notable in being able to undergo 1,2-elimination of HF (unlike C 60 CF 3 H), to give a peak of 820 amu (see Fig. 1). 19 F NMR spectra The peak data are given in Table 1 and the 1 D spectrum for the 1112 species is shown in Fig. 2. This is identical in the cage-F region to that obtained previously, 1 as is the 2 D spectrum (not shown).