Search

The efficiency of silicon p‐n junction photovoltaic cells for solar energy conversion is higher in terrestrial use than in space use by an amount which can be as high as 25%. The power output in terrestrial use is nevertheless lower than in space use because of the lower solar intensity. The explanation for the higher efficiency is that, on traversal of the sunlight through the earth’s atmosphere, water vapor and carbon dioxide absorb energy primarily at wavelengths in the near infrared to which the solar cell is not sensitive. In the present paper a quantitative evaluation of this hypothesis is carried out by numerical determination of the total integrated power and of the number of photons/sec capable of creating hole‐electron pairs corresponding to known solar power distributions for air masses 0, 1, 2, and 3 and by translation of the results into lossless cell efficiencies. The efficiency for silicon cells at air mass 1 for the case of no absorption bands proves to be the same as for air mass 0, where...

The tables associated with the exact Langmuir space charge theory have been represented to a maximum relative error of 0.01% by approximation formulas which are suitable for use with digital computers. Application of the exact theory to a thermionic converter comprising two close‐spaced planar impregnated cathodes has permitted a critical evaluation of the approximate space charge theory of Nottingham. The influence of the electrode separation, the emitter and collector work functions and of the emitter temperature has been investigated. Spectral emittance measurements on a cathode surface at two wavelength has resulted in a more firmly based estimate of the radiation heat transfer between two impregnated cathodes and of the maximum efficiency of an ideal design.

An experimental study of thallous sulfide photo‐conductive cells is presented. It includes an investigation of the sensitizing oxidation reaction; cell characteristics as function of voltage, temperature, light intensity, wave‐length, modulating frequency, and composition; measurements of thermoeffect and optical absorption; x‐ray and electron diffraction studies, and an examination of the photo‐response to light flashes of very short duration. A theoretical discussion will be given as Part II in a subsequent paper.

An experimental study of the L cathode is presented, including the chemistry of carbonate decomposition and barium generation, the origin of BaO in the evaporant, the barium transport mechanism through the porous plug, and the factors determining cathode life. Carbonate decomposition is accompanied in low porosity cathodes by the deleterious reaction BaCO3+13W=13Ba3WO6+CO. Barium is generated in the cavity at an equilibrium pressure by the reaction 2BaO+13W=13Ba3WO6+Ba and is transported continuously through the porous tungsten predominantly by Knudsen flow. During transport the barium acquires oxygen from the tungsten, producing in the evaporant a large BaO content which helps to activate the surface. Termination of life coincides with completion of the BaO–W reaction and onset of the subsequent reaction 23Ba3WO6+13W=BaWO4+Ba, which is accompanied by release of a poisoning agent. Material decrease in evaporation rate and a corresponding increase in cathode life may be gained by addition of SrCO3.

A theory is formulated which is in substantial agreement with all facts reported in a previous paper on thallous sulfide photo‐conductive cells. The underlying physical picture is that light is absorbed in the Tl2S proper and in the presence of oxygen, positive ``holes'' and negative oxygen ions are formed. The ``holes'' and ions recombine at a rate proportional to the product of their concentrations. During their lifetime, the negative ions act as space charge compensators for the defect conduction and allow correspondingly higher currents to flow. This space charge compensating action, which is similar to the role played by positive ions in gas discharges, seems to be one essential mechanism of producing secondary photo‐effects.

A simple procedure for determining the ``instantaneous'' absolute evaporation rate of barium from dispenser cathodes is presented. The method entails exposing a clean tungsten wire to a stream of barium and noting the time required to reach maximum emission. The sensitivity of the wire as a detector of total barium is established on an absolute basis by chemical calibration. It has been found that the wire sensitivity is a function of the barium oxide content of the evaporant stream. This fact may be employed to determine the oxide composition of the evaporant from wire emission‐time measurements and a single chemical analysis. It is also possible to obtain approximate, instantaneous values of total barium and the fraction BaO at any time during tube life from wire emission‐time measurements only. Calibration data are presented for almost the full range of Ba–BaO compositions.

The Simon two‐band semiconductor model is investigated further theoretically to determine whether a finite upper bound can be imposed on the thermoelectric figure of merit. It is shown that limitations on the adjustability of the Fermi level and use of reasonable physical values for the band parameters do lead to a finite upper bound for the two‐band model which is of the same order of magnitude as that derived earlier by Rittner for a one‐band model. The former can be severalfold higher than the latter but only at pathologically high values of the effective mass ratio mh/me. Other factors which favor a high value of the upper bound are discussed.

Emission and evaporation characteristics of a porous tungsten cathode impregnated with the composition 5BaO·2Al2O3 are presented and are interpreted in terms of the cathode mechanism. Barium necessary for activation is generated by the reaction, 23 Ba3Al2O6+13 W=13 BaWO4+23 BaAl2O4+Ba, and is transported through partially clogged pores, the length of which increases with time, predominantly via Knudsen flow. During transport, oxygen is acquired from the tungsten, leading to a substantial content of BaO in the evaporant. The BaAl2O4 component of the impregnant is inert. Emission is substantially lower than that of an L cathode, presumably because of release of a poisoning agent accompanying the activator.

A new electrolytic selenium photo‐cell is described and its characteristics given as a function of operating voltage, temperature, external resistance, and time. Data are included indicative of its response to unmodulated, modulated, and monochromatic light; comparison is made with cells of similar type or possessing similar characteristics; and a method of compensating for the dark current of photo‐cells is shown.An electronic circuit is also described which can greatly improve the voltage output of the photo‐element. It is generally applicable to photo‐conductive and photo‐voltaic cells.

The figure of merit for a single stage Peltier heat pump is optimized in the region of partial Fermi degeneracy. A more rigorous treatment is presented of the properties of a cascade operating at maximum coefficient of performance. Finally, the heat leak in the absence of current flow is called to attention as a serious problem in devices with excess capacity.

A study has been made of the inhibition of emission from barium‐activated tungsten dispenser cathodes and adjoining molybdenum electrodes which can be produced by carburization. X‐ray diffraction and metallographic techniques have been employed coupled with emission measurements as a function of time, temperature, and arrival rate of Ba+BaO. Performance characteristics including the extent of the emission suppression and its time dependence are presented. It is concluded that the emission inhibition results from greatly reduced activator coverage of the surface of the carbide relative to that of the metal owing to much weaker adsorption forces. Return of emission coincides with disappearance of the carbide layer from the surface via diffusion.

Previous work is reviewed, and the Mott‐Littleton method of treating the rigid lattice case is extended with some modification to higher order approximations. Numerical results are given for their hypothetical monatomic crystal and for NaCl.

Hitherto unpublished data on life tests and evaporation rate of total barium from type B impregnated cathodes are presented. These data conflict with the recently deduced conclusion of Forman that the cathode surface is covered with less than a monlayer of adsorbed barium and that the barium coverage decreases with time. However, they are in agreement with the cathode mechanism proposed by Rittner and co‐workers based upon extensive experimental studies. The Forman erroneous conclusion is traced to an inadequate oxygen content of the monolayer in synthesized cathode surfaces intended to simulate the real cathode.

Measurements are presented of (a) the emission density from clean and oxygen covered molybdenum surfaces as a function of the arrival rate of Ba or of Ba+BaO and of (b) the migration length of Ba and of Ba+BaO over oxygen covered molybdenum surfaces. The emission density is comparable to that obtained from similarly activated tungsten surfaces, but the minimum arrival rate of total barium required for complete surface coverage is much higher than for the tungsten case. The migration length is also comparable for the two systems.

In my recent paper "Concerning the Theory of Photoconductivity in Infrared-sensitive Semiconducting Films" (Science, 1950, 111) page 685, line 7, column 2, should read ". . . to essentially pictures (1) or (2). . . ." rather than ". . . pictures (10) or (5). . . " On page 687, line 5, column, 1, the words "fixed," "positive," and "electrons" should be deleted.

Instances in which the work function M does not exceed the ionization potential I, and the cesium pressure is too low to permit envocation of volume ionization as a result of substantial temperature rise in plasma due to Joule losses, are discussed. Methods of circumventing these difficulties by postulating the presence of regions of sufficiently high work function on the otherwise low workfunction emitter surface or by generalizing the concept of contact ionization by using the Saha-Langmuir theory are compared. The latter approach is shown to be more desirable. (L.N.N.)