L. W. James

Bandgap and lattice constant of GaInAsP as a function of alloy composition

AbstractPublished data for the composition dependence of the room-temperature bandgap (Eg) and lattice constant (ao) in the pseudobinary GayIn1-yAs, GayIn1-yP, GaAsxPl-x, and InAsxPl-x systems have been used to derive the following equations for the quaternary GayInl-yAsx Pl-x, alloys:

Liquid Epitaxial Growth of GaAsSb and Its Use as a High‐Efficiency, Long‐Wavelength Threshold Photoemitter

\begin{gathered} a_o ({\AA}) = 5.87 + 0.18x - 0.42y + 0.02xy \hfill \\ E_g (eV) = 1.35 - x + 1.4y - 0.33xy - (0.758 - 0.28x)y(1 - y) \hfill \\ - (0.101 + 0.109y) x(1 - x). \hfill \\ \end{gathered}

Dependence on Crystalline Face of the Band Bending in Cs2 O‐Activated GaAs

Available experimental data are in excellent agreement with these equations.

Operation of III‐V Semiconductor Photocathodes in the Semitransparent Mode

For terrestrial applications, the figure of merit for photovoltaic solar energy conversion devices is watts output per dollar of cost. AlGaAs/GaAs heterojunction cells have a very favorable watts per dollar figure of merit when used at high values of sunlight concentration. An experimental 1/2−in.−diam cell was operated in air mass 1.4 sunlight with an output power density of 4.52 W/cm2 at an effective concentration of 312 suns with a power conversion efficiency of 17.5%. The same cell was operated at 200 °C with an output power density of 3.45 W/cm2 at a 14% efficiency. The efficiency of the cell was 23% with a fill factor of 0.85 at a lower concentration ratio which is obtainable using simple concentrator schemes.

Transferred electron photoemission from InP

The ternary‐phase diagram of GaAsSb has been calculated using Darkens quadratic formalism for a ternary liquid and assuming a regular solid solution. Liquid epitaxial layers of GaAsxSb1−x have been grown in the range 0.75>x>1 on {100} and {111} GaAs substrates. Results are in excellent agreement with the calculated phase diagram. Variation of bandgap with composition of the layer has been determined by transmission, photoemission, and x‐ray fluorescence experiments. The data were fitted to a curve of the form EG=A+Bx+Cx2, where A=0.725 eV, B=−0.32 eV, and C=1.02 eV. Graded bandgap layers have been obtained, with gradients of 700 eV/cm near the substrate interface and 25 eV/cm for thick layers. For use as high‐efficiency photoemitters, the samples were doped p type by the addition of elemental Zn to the melt. Cesium and oxygen surface layers were used to lower the work function. Quantum yields of 0.1%–0.2% at 1.06 μ were obtained. Field assisted photoemission in a graded bandgap sample has been calculated...

Organometallic‐sourced VPE AlGaAs/GaAs concentrator solar cells having conversion efficiencies of 19%

A detailed picture of the behavior of cesium oxide as a low work‐function coating on III‐V semiconductors and on silver has been obtained. Measurement of required cesium and oxygen exposure for optimum photoyield shows that the compound normally formed is close to CS2O, with variations in required exposure for very thin and very thick layers. By making simultaneous Kelvin work‐function, photoyield‐threshold, and thickness measurements, it was possible to establish that the CS2O, an n‐type semiconductor, forms a heterojunction or Schottky barrier with its substrate. This provides a band bending which produces a gradual lowering of the vacuum level with increasing thickness to an ultimate work function of 0.6 eV. The photoyield and dark current from the substrate are limited by the interfacial barrier at the heterojunction. This barrier is 1.00±0.05 eV for a silver substrate and 1.23±0.03 eV for GaSb. The band‐bending distance in the CS2O is about 50 A and the hot electron scattering distance is 9 A. These data have been used in an improved calculation of the maximum Γ escape probability and requisite CS2O thickness for electron emission from III‐V semiconductors of different bandgaps. Electron emission from CS2O induced by an oxygen overpressure was also measured. CSOH is compared with CS2O as a work‐function lowering coating.

Optimization of the InAsxP1−x–Cs2O Photocathode

Electron energy loss in the band‐bending region of the p‐type III–V semiconductor in a III–V photocathode is an important factor in determining the escape probability and the optimum doping. From measurements of photoelectric yield near threshold from Cs2O‐activated n‐type GaAs, the position of the Fermi level at the GaAs–Cs2O interface was determined for {110}, {100}, {111A}, and {111B} surfaces. Assuming the Fermi‐level position at the GaAs surface to be independent of doping, the band bending for p‐type GaAs is greatest for the {111A} face and least for the {111B} face. The measured escape probabilities of photoexcited electrons from different crystalline faces of optimally activated 5 × 1018/cm3 Zn‐doped liquid epitaxial GaAs correlate well with the band‐bending measurements. The {111B} sample has an escape probability of 0.489 and a luminous sensitivity of 1837 μA/lm.

Photoemission from cesium‐oxide‐activated InGaAsP

Calculations show that very nearly bulk quality material is required for high‐efficiency semitransparent III‐V photocathodes. For narrow‐band response, this can be obtained by epitaxially growing a thin layer of a semiconductor whose bandgap is slightly less than that of the substrate. Cathodes made by growing GaAsSb on GaAs have given quantum efficiencies comparable with front surface values, peaking out at 0.54% at 1.35 eV near the onset of absorption in the GaAs substrate. Preliminary results demonstrating semitransparent yield at 1.06 μ of 0.013% are also shown.