The Thermionic Work Function of Oxide Coated Platinum

Abstract

Measurements of the thermionic work function of pure platinum coated with oxides of barium and strontium have been made simultaneously by two methods for the same segment of a uniformly heated filament. The theory of the measurements and the experimental arrangements are the same as used in an earlier experiment on the thermionic work function of pure tungsten. Filament temperatures accurate to \ifmmode\pm\else\textpm\fi{}5\ifmmode^\circ\else\textdegree\fi{}, were found from the resistance of the filament at 0\ifmmode^\circ\else\textdegree\fi{}C in conjunction with the temperature coefficients of resistance. (1) In the calorimetric method the equivalent voltage of the work function was computed from the sudden voltage change resulting from switching off the space current, due to the cooling effect of the emission. The determination was much more difficult than in the case of the tungsten filament, and measurements were made at the single temperature 1064\ifmmode^\circ\else\textdegree\fi{}K. At this temperature the work function $\ensuremath{\phi}$ was found to be equal to 1.79\ifmmode\pm\else\textpm\fi{}.03 volts. (2) In the temperature variation method it was found that, after the temperature had been changed suddenly from one value to another, the emission changed approximately exponentially from an initial value to a final steady value. The half value period of this change varied from a few seconds at high temperature to over a quarter of an hour at low temperature. Interpreting this phenomenon as due to a progressive and reversible change of the character of the filament with temperature, the initial emissions after temperature changes from 1064\ifmmode^\circ\else\textdegree\fi{}K were used to determine the $b$ constant of Richardson's equation corresponding to the equilibrium character of the filament at 1064\ifmmode^\circ\else\textdegree\fi{}K, and similar measurements were made for the $b$ constant corresponding to the character of the filament at 911\ifmmode^\circ\else\textdegree\fi{}K. The two determinations lead, through the relationship $\ensuremath{\phi}=\frac{\mathrm{bk}}{e}$, to 1.79 volts and 1.60 volts for the corresponding values of $\ensuremath{\phi}$. For 1064\ifmmode^\circ\else\textdegree\fi{}K, then, the two methods give values for $\ensuremath{\phi}$ in agreement. The measurements are, however, not sufficiently accurate to give any indication whether or not an electron within the metal possesses the thermal energy $\frac{3kT}{2}$. The various corrections made and possible errors are thoroughly discussed. It is pointed out that if the transition from the equilibrium state at one temperature to that at another had occurred so rapidly as to avoid observation, a disagreement of 25 per cent between the values of $\ensuremath{\phi}$ given by the two methods would have been obtained which might have been misinterpreted.