Thc sorptions of hydrogen and of oxygen on freshly prepared and on gas-treated cadmium oxide were measured from 0 to 250 "C. The data suggest that hydrogen chemisorption predominates near 26", where it is the precursor of water formed on the surface. Surface reduction occursat temperat~~res as low as 50". At least two types of oxygen sorption exist. The isobar has a min~mum near 50". Chcmisorption occurs below 5.0" and the sorption kinetics indicates the presence of two types. The incorporation of oxygen into the bulk may account for the observed increased sorption above 150".
SORPTION O F H Y D R O G E S A X D O X Y G E S BY CADMIUh/I OXIDET h e relatively large deviations fro111 stoichionletry of cadmium oxide and its behavior as degenerate electronic se~lliconductor have prompted a considerable amount of 1vork on the electronic properties of that substance. Some effects of oxygen pressure on conductivity or thermoelectric power have been noted, but quantitative sorption experinlents have not been reported. I n view of the inlportance of surface effects to se~lliconductor behavior, and the possibility of comparison of data \\;it11 those of zinc oxide, we have underta1;en the present study.
EXPERIMENTAL DETAILSGas sorptions were measured by following the decrease in pressure within a high-vacuum system of conventional design equipped with mercury and dibutylphthalate manometers. Hydrogen was purified by diffusion through hot palladium. Tanli oxygen was passed over hot CuO and through a trap filled with glass wool a t -1(36°. I-Ieliurn and nitrogen were purified by passage through glass-wool-filled traps a t -106O. Water was degasscd by alternate freezing and thawing ilz vaczro.A rcproduc~blc adsorbent was prepared by the decomposition of Cd0.1.6!5H10 to C d 0 0 . s~ i n vaczro (1). A sa~llple of the hydroxide was placed in the sample chamber of the adsorption system and evacuated a t room te~nperature. !-\ furnace, preheated to 500°, was then placed around the sample chamber. A pressure of approximately 10-6 rnm Hg was reached within about 15. min. 'I'he furnace was removed after 40 min, a second preheated furnace \va> installed, and the sample \ -. , as then held a t some previously choscn temperature.Dead spaces were measured by the expansion of helium. The nitrogen R.E.T. surface area (2) of the oxide prepared as described remained constant a t 5.2 m v g in vacuo for 12 h. 'The analysis of several samples yieldcd thc f o r m~~l a C d 0 0 . s~ and indicated that the dccomposition of the hydroxide was co~nplctc (1). Amounts of gas sorbed are given as 1111 N.T.P./g Cd0u.s~.T o obtain information about sorption on gas-treated as well as on freshly prepared surfaces, several series of experiments were made. Each serics was isothermal, and \vas made with a freshly prepared sanlple weighing -1 g. I n one series, a n adsorbent sample was prepared as described, and was then exposed to hydrogen a t an initial pressure of 206 111111 I-Ig. T h e isothermal gas sorption was follo~ved for sollle time. Each of the ...