4338
T. J. MALONE AND H. A. I\IIcGEE, JR.
Mass Spectrometric Investigations of the Synthegis, Stability, and Energetics of the Low-Temperature Oxygen Fluorides.
I. Dioxygen Difluoride
by T. J. Malone and H. A. McGee, Jr. School of Chemical EngheQTi?2Q,Georgia Institute of Technology, Atlanta, Georgia 30332 (Received July 20, 1965)
A mass spectrometric arrangement is described which permits the study of compounds that are stable only at cryogenic temperatures, The low-temperature oxygen fluorides were synthesized and their mass spectra were studied between 77 and 260°K,but neither the parent ion of O,F,, 03Fz,or O,F, nor the unambiguously assignable fragment ions of O,F, or 04F,were observed, However, A (OzF+)and A (OF+)weye measured to be 14.0 -;t 0.1 e.v. and 17.6 A 0.2 e,v,, respectively, at 13OOK. These values are consistent with other related data and indicate that these ions are being produced by electron impact fragmentation of OzFz. The data permit the development of the energetics of the OnF2 rnoleculc.
With the exception of OFz, the other known and pseudo-known oxygen fluorides ( O ~ F ~OF, , OZF, O@E: and 0 4 3 ' 2 ) constitute a family of compounds that exist as stable entities only so long as they are maintained at some very low temperature, This particular family, as well ax other instances of species exhibiting this unusual cryogenic stability, have been discussed in a recent review. The microwave spectra of OzFzhave indicated that the species is a nonplanar symmetric chain molecule, but no unequivocal direct observations of the other low-temperature oxygen fluorides have been reported. For example, OBFzis reported by nurneroud inveatigators to be a blood-red liquid at 90°K. and t o be evidently the maat powerful oxidizer known, but whether or not this red liquid is actually pure molecular 0,F2is quite unknown. Direct observations of these species are needed first to clarify their existence, and second, because of their significance in an exploration of that little known, but very interesting, realm of chemistry below about 150"K., which may reasonably be called cryochemistry. Mass spectrometry offers a means for the unambiguous observation of these molecules as well as a means to study their synthesis, their stability, their chemistry, and their energetics. Because of the unusual thermal instability at cryogenic temperatures, absolute temperature control is imperative at all times, and hence The Journal of Phyaical Chemistry
new equipment and procedures had t o be developed. A thermal gradient cryogenic inlet system has been developed in which a mixture of these highly reactive species can be synthesized, separated, and vaporized directly into the ionizing electron beam af a time-offlight mass spectrometer. Experimental Section Apparu~A ~ Bendix time-of-flight mas8 spectrometer, Model 12-107 with a Model 8-14-107 source, was used in these experiments, Although the rather unusual auxiliary equipment that was used is described elsewhere,Z a brief description of the cryogenic reactor and inlet system seems appropriate. Essentially, the system i5 designed such that a cold, gaseous sample can be injected directly into the ionizing electron beam of the spectrometer. Since the species are stable only a t very low temperatures and since the ion source is at room temperature, the necessary fast pumping to remove thermally cracked background gas from the source rapidly was provided by a 750 l./sec. diffusion pump system (National Research Corporation, Model HS4-750). (1) H. A. McGee, Jr., and W. J. Martin, Cryogenics, 2, 257 (1962). (2) H. A. McGee, Jr., T. J. Malone, and W. J. Martin, to be pub-
lished.
ENERGETICS O F LOW-YCEMPERATURE
OXYGEN
FLUORIDES
Referring to the schematic shown in Figure 1, the outer sleeve, A, slides i,hrough a double O-ring seal into a vacuum lock arrange.ment and then into the spectrometer, This permits the removal of the cryogenic assembly for adjustments or repairs without the necessity of breaking vacuum in the main spectrometer system. I n essence, the unit consists of a straight tubular reactor and condensation space with a thermostated chamber (B and C) at each end. The temperatures of the two chambers are controlled independently by accurately balancing the heat required to vaporize a finely controlled liquid nitrogen input stream against the heat added. by means of a resistance heater wound on the center posts, D and E. By varying the liquid nitrogen rate and the heater power, each chamber could be readily maintained at any temperature above the normal boiling point of nitrogen (77°K.). The temperatures were automatically controlled within %l.O"K. of the desired values by two Leeds and Northrup Speedomax H, AZAR recorder-controllers which operated the heaters in the simple on-off mode. A mixture of very reactive species can be either injected into or synthesized in the reactor space, F, while both B and C are a t some low temperature. Then by raising the temperature of B, a thermal gradient is imposed along the connecting tube, G, between B and C causing the species to be successively vaporized and recondensed at distances along G which depend upon the volatility of the individual species. Then by manipulating the temperatures of B and C the species may be successively transported through the channel, H, and into the spectrometer. The blade-shaped nose, I, which is at the same temperature as C, is inserted directly into the ion source so that the electron beam is actually in grazing, tangential incidence with the sample inlet hole, J. Hence, mass spectrometric analysis of the vapors is achieved without warmup above the temperature of C. Procedure. The syntheses were conducted in the assembly shown in Figure 1 by using a 1100-v., 60cycle annular discharge between the single electrode,
4339
K, and the grounded monel walls of the reactor space. The synthesis gas was an 0 2 and Bz mixture in the stoichiometric ratio corresponding to either O2F2 or to O@z, which yields O2F2 or OsF2 as the predominant product, re~pectively.~ The reactor pressure and temperature were 25-35 torr and 77"K., respectively. The mixture of fluorides was partially separated and vaporized directly into the spectrometer as described above. The appearance potentials of the ions of interest were determined using the Fox retarding potential difference (RPD) method,* with the energy scale calibrated immediately before and after each energy measurement using both argon and nitrogen, I n making the energy measurements on the highly reactive fluorine compounds, it was necessary for the ion source to become partially passivated before the syxtrometer would stabilize enough to make accurate and reproducible measurements. With continuous injection of the fluoride samples, the trap current decreased t o an inoperable level after 12-15 hr.
Results Mass Spectrum as a Punclion of Temperature. Several hundred traces of the mass spectrum were made over the temperature range of 77-200°K. in four OaF2 syntheses and four 02Fz syntheses. In addition, mass spectra as a function of temperature were obtained for the unreacted feed gas mixture of O2and Fz, as well as for SiF,, which turned out to be a particularly troublesome impurity in the product mixtures. Relative intensities of all ion currents were tabulated at each temperature. The limitations of the spectrometer and complicated features of the spectra combined to necessitate care in making mass assignments for the observed ions, Mass scales were prepared for each scan rate using m/e 16 (0 atom) and m/e 32 (02 molecule) as reference masses. Using only the measured distance between the reference masses as verified from several traces, the distance from m/e 16 to all other masses out to m/e 130 was calculated and plotted to yield completely unbiased scales. The use of these scales resulted in consistent and unambiguous mass assignments for 12 I m/e S 130 for more than 98% of all traces. The source of each ion was then determined from the observed relative intensities and reported cracking patterns of species known to be present as impurities in commercial fluorine and oxygen (i.e., OFz, CF,, CzFs, N2, HF, N F , COZ, SFa, andSiF4). I n
SCALE4HCHES
Figure 1. Schematic diagram of thermal gradient, cryogenic reactor, and inlet, assembly.
(3) A. D. Kirshenbaum and A. V. Grosse, J. Am. Chem. SOC.,81, 1277 (1959). (4)E C..Melton and W. H. Hamill, J. Chem. Phys., 41, 546 (1964).
Volume 69, Number 1.2 December 1966
4340
T. J. l
TEMPERATURE P K I
Figure 2, Eeprewentative variation of ion current intensities with temperature of the most interesting ions observed in the
OsFasynthesis experiments.
order to identify each impurity present in the OBand f feed gas and to determine the volatility, and hence the temperature dependence of the mass spectra, the unreacted feed gas was pumped through the system with both B and C at 77"K., whereupon all significant impurities were condensed. Spectra were then recorded at 2-3'K. intervals over the temperature range of 77-200°K. From the temperature required for the observation of the various condensed impurities of known vapor pressure, it was determined that a species exerted a vapor pressure of 0.01-0.1 torr in E (see Figure 1)when it was first detected in the spectrometer. Ozone Dfluoride. Figure 2 illustrates the general variation of the more interesting ion currents as a function of temperature for the experiments in which OaFz the principal reaction product. The cuweB give a composite representation of the ion current intensities observed in three of the 03Fzsynthesis runs. The intensity vs. temperature curves from the fourth 03Fzrun exhibited a somewhat different character that was similar to that of the OzFz synthesis experiments. Hence, it was concluded that OzF2was likely the principal reaction product in that experiment. I n Figure 2, three temperature ranges are apparent (80-95, 102-118, and 126-150°K.) within which significant changes in the mass spectra occur. These temperature ranges coincide with the melting point of 03Fz (83-84'1
