ORGANIC MASS SPECTROMETRY, VOL. 29, 499-504 (1994) Novel Use of an Isotope Separator to Determine the Position of Fluorine-18 in Labelled Victor W. Pike,* Stephen L. Waters and Franklin I. Aigbirhio PET Methodology Group, MRC Clinical Sciences Centre, Royal Postgraduate Medical School, Hammersmith Hospit zy 11, Ducane Road, zyxwvutsrqp London W12 zyxwvutsrq ONN, UK John Makepeace Richard J. N. Tanner Glaxo Group Research Ltd, Ware, Hertfordshire SG12 ODP, UK National Physical Laboratory, Teddington, Middlesex TWl 1 OLW, UK A novel technique is described for measuring the site selectivity of methods for labelling the major CFC-alternative, 1,1,1,2-tetrafluoroethane (HFA 134a), with fluorine-18 zyxwvu = 109.7 min). The carbon-carbon bond in radioflu+ rinated HFA 134a is broken in the ion source of an isotope separator. Radioactivity associated with the ion beam of the [CF,"F] +' fragment zyxwvuts (m/z = 68) zyxwvut is collected, measured and divided by the integrated mass of the simulta- neously collected ion beam for the ICF,] +' fragment (m/z = 69) to give the 'specific radioactivity' (in nCi nmol- ') of the radiolabel in the 1-position. Similarly, the 'specific radioactivity' of the radiolabel in the 2-positir tn is calculated from the measured radioacticity of the ion beam from the [CH2'8F] +' fragment zyxw (m/z = 32) and t$e integrated mass of the simultaneously collected ion beam from the [ CH,F] +* fragment (m/z = 33). The selectiviiy of the labelling procedure for a particular position is then given by the decay-corrected ratio of specific radiw activity at that position to the sum of specific radioactivities. The labelling of HFA 134a by the reaction of I''Flfluoride with trifluoroethylene was found to have 97% selectivity for the CF, group, whereas labelling by tbe reaction of I "Flfluoride with 2,2,2-trifluoroethyl p-toluenesulphonate was found to have 91% selectivity for the CH,F group. This information is of value for tracer studies of the fate of HFA 1344s in man following its inhalatiun as a drug propellant. The described technique is of potentially wider value for determining the position of fluorine- 18 in labelled poiyfluorinated molecules. INTRODUCTION Damage to the Earth's ozone layer from commercial chlorofluorocarbons (CFCs) is of immense environ- mental concern and there is now international agree- ment to phase out their manufacture and use. This has led to major efforts to develop substitutes for their main applications as refrigerants and coolants.'-3 An impor- tant small-scale use of these compounds is as propel- lants for drugs to be given by inhalation. The hydrofluorocarbon, l,l, 1,2-tetrafluoroethane (HFA 134a), is now in large-scale production as a substitute for CFCs in all their major application^,'-^ since the absence of a chlorine atom in this molecule abolishes its ozone depletion potential. This compound is also being considered as a substitute drug propellant. In view of this potential, it is important to understand the fate of HFA 134a after its inhalation by man. Recently, we have addressed this question using a radiotracer tech- nique, involving the labelling of HFA 134a with cyclotron-produced fluorine-18, a short lived (tl,2 = 109.7 min) positron-emitting radioisotope zyxwvut (p' = 97%) that can be detected in vivo by, for example, sensitive whole-body counting4 or positron emission tom- ography (PET).5 CCC 0030-493X/94/090499-06 zyxwvut 0 1994 by John Wiley & Sons, Ltd. Of major interest is the metabolic fate of HFA 134a. Studies in animals suggest that an oxidative pathway to trifluoroacetate and a reductive pathway to fluoride might exist in man.6 Hence, in order to elucidate any possible metabolism of HFA 134a in man, we developed labelling procedures that were intended to be specific for labelling at either carbon atom.'s8 Given the short half-life of fluorine-18 and the inert nature of HFA 134a, it seemed impractical to verify the position of label by developing a degradative chemical technique. Hence we considered a physical technique for this purpose, namely the use of the 'isotope separator' at the National Physical Laboratory (NPL). With this instru- ment we were able to measure the specific radioactivity of fragments from each end of the molecule and to cal- culate the site selectivity of our labelling procedures. This technique is described in this paper. EXPERIMENTAL Materials 1,1,1,2-Tetrafluoroethane (HFA 134a; 97%) was obtained from Fluorochem or as a kind gift from ICI Chemicals and Polymers (Dr R. L. Powell). Received I0 March 1994 Revised 12 May 1994 Accepted 15 May 1994