Cs RECOM BINATION CROSS SECTIONS 151 that the recombination were nondissociative then the value of o. given above would have to be increased by a factor of 2. The importance of the dissociative re- combination mechanisms under appropriate conditions in the noble gases is widely accepted. " While it is probable that molecular cesium recombines dissocia- tively, it is not certain. The large values of cesium recombination observed by Dandurand and Holt" (3 — 12X10 r cm%ec) at around 300'K are due to molecular recombination although Dandurand and Holt attribute their observation to a nondissociative mechan- ism. Very recently Harris" reported a value of cesium molecular recombination of 10 ' cm'/sec in the tem- perature range 1520-1670'K. He definitely attributes his observation to dissociative recombination. Bates" has made a rough estimate of dissociative recombination and predicts values of the order of 10 ' cm%ec at T=250'K. The temperature dependence of the co- efficient that Bates gives is roughly as 1/gT. By Bates' "D. R. Bates, Atomic and Molecular Processes (Academic Press Inc. , New York, 1962), p. 266. "P. Dandurand and R. B. Holt, Phys. Rev. 82, 278 (1951). ~ L. P. Harris, J. Appl. Phys. 36, 1543 (1965). "D. R. Bates, Phys. Rev. 77, 718 (1950); 78, 492 (1950). estimate, then, the coefficient at T= 1380'K would be 4&&10- cm'/sec. Thus, both the recent experimental work of Harris and the estimate of Bates agree with the experimental value found by the beam method for the molecular recombination. The Dandurand and Holt values are, however, at least an order of Inagnitude larger than those found in this experiment. It seems likely, however, that the high values of recombination observed by all of the measurements on cesium mo- lecular ions are due to dissociative recombination. ACKNOWLEDGMENTS The authors wish to thank Professor B. Bederson of New York University for much helpful discussion and advice. We are indebted to Dr. S. Bloom of RCA Laboratories for his continued encouragement and his critical reading of the manuscript. The help of P. V. Goedertier and J. J. Thomas of RCA Laboratories during the earlier stages of the program is gratefully recognized. The complex assembly operations required in this experiment@'were ably performed by N. Klein and R. Chamberlain. P H YS I CAL REVIEW VOLUME 141, NUMBER 1 JANUARY 1966 Absolute Cross Sections for Single Ionization of Alkali Ions by Electron Impact. I. Description of Apparatus anti Li+ Results* W. C. LIKE'BERGER, ) J. W. HOOPER, AND E. W. McDANrzr. Georgia Institute of 2'eehaofogy, Atlarsta, Georgia (Received 19 August 1965) The absolute cross sections for the single ionization of Li+ ions by electron impact have been measured over the electron-energy range from below threshold (75.6 eV) to 800 eV. The measurements were performed with an ultrahigh-vacuum crossed-beam facility operating under continuous, rather than the usual time modulated, beam conditions. A detailed description of the apparatus and experimental method is presented. Numerous checks were performed to justify the use of the continuous-beam measurement technique de- veloped for this experiment, particular attention being paid to beam intensities, beam pro6les, space charge, signal-to-noise ratio, and ion-beam composition. The error in the measured cross sections is believed not to exceed +12% above 150 eV electron energy; it may be as large as &21% at 90 eV. Of this possible error, an amount +6% is considered systematic. I. INTRODUCTION ' ~ LASTIC and inelastic collisions involving electrons & and heavy particles are of great importance in astrophysics, upper atmospheric phenomena, thermo- nuclear research, plasma physics, and gaseous electron- ics. Experimental information concerning such collisions may be obtained either indirectly from swarm studies *This work was partially supported by the Controlled Thermo- nuclear Research Program of the U. S. Atomic Energy Commis- sion. $ The work reported here is a portion of a research program undertaken by one of us (WCL) in partial fu116llment of the re- quirements for the degree of Doctor of Philosophy at the Georgia Institute of Technology. or directly from beam experiments. The vast majority of direct-collision studies have involved the passage of a beam of projectiles through a target gas, and either detection of the reaction products formed in the gas or observation of changes in the composition of the emerg- ing beam. However, the single-beam approach is not applicable to the study of many of the reactions of greatest interest — for example, those between two species of charged particles. To obtain reliable results in such cases, it is usually necessary to study collisions occurring in intersecting beams. This paper describes the experimental apparatus and techniques employed for the measurement of the cross