Conventional S20 multialkali photocathodes have a wide wavelength coverage from < 200 to > 850 nm, but their high transparency and the surface work function result in low quantum efficiencies at longer wavelengths. Theoretical modelling of the photon and excited electron interactions that define the cathode performance provides a realistic prediction of the measured response. The theory emphasizes that the basic light absorption is strongly sensitive to the cathode thickness, wavelength, polarization and incident angle. Parameters can be selected which predict that even at long wavelengths (e.g. 900 nm), absorption may be increased from ∼1% to ∼100%. Cathode topographies can be designed to exploit these responses and offer increased absorption at the longer wavelengths. Alternative designs, which include waveguiding of light within the cathode window, or in structured surfaces, can similarly lead to almost total absorption of the incident light by increasing the number of interactions. These concepts of optimal incidence and waveguiding have been both theoretically modelled and demonstrated in newly fabricated cathode designs. The methods have variously reached quantum efficiencies in excess of 50% at wavelengths in the range from 200 to > 750 nm under different operational conditions. The improvement factors relative to normal incidence on planar cathodes increase for longer wavelengths, and examples of 20–50 times by ∼900 nm were noted. Whilst the absolute S20 efficiency values at long wavelengths are still small, the improvements offer a usable sensitivity even beyond 1 µm, as demonstrated by spectroscopy data up to at least 1140 nm.

No abstract

A model for conductivity storage in CdS is proposed. It involves a storage level from which electrons are excited while the collection of these electrons by a receiving level is accompanied by local crystal deformation allowing the quasi-equilibrium state, which stored conductivity represents, to be set up. Published literature dealing with the sensitivity of CdS electrical properties to thermal or mechanical stress lends credibility to the model.Various techniques of photoelectronic analysis are used to obtain values of 5 × 10−16 cm2 for the photon-capture cross section and 10−30 cm2 for the recapture cross section of a stored conductivity carrier. Storage level densities ranged from 1016 to 1019 cm−3 with various crystals.

The efficiency of S20 multialkali photocathodes is a function of thickness, wavelength, polarization and angle of incidence of light into the cathode. Therefore no single set of conditions can maximize performance over the entire spectral range. Consequently, two prototype photocathodes have been made with a gradation in thickness so that during monochromatic analysis the optical beam can be addressed to different thickness regions at preferred angles of incidence. This potentially enables a spectrum to be recorded under optimal conditions at a single interaction event. Relative to normal commercial S20 photomultipliers the quantum efficiency (QE) has been significantly raised across the entire spectral range. The current data were primarily obtained between 450 and 800 nm and at 450 nm it resulted in values of at least 65% QE, which is the highest value ever cited at this wavelength. Signals at longer wavelengths, for example at 750 and 800 nm, were recorded with up to 20 and 10% QE, respectively. Once again these are new record values that match performance from multiple interactions in waveguide cathodes. The data from this new design of photocathode underline the potential for improvements in efficiency for non-normal incidence in graded thickness photocathodes and indicate that current S20 technology could be significantly enhanced. Alternative enhancement methods are mentioned, particularly for spectrally dispersed signals. The enhancements are compared with data for a standard high quality S20 photocathode.

In order to increase the transmission and speed of response of fibreoptic faceplates in imaging tubes a method is reported which used metallization only of the interfibre glass regions. This offers maximum transmission of the fibres whilst supporting high conductivity, as required for fast response times. An initial metal layer was laser ablated with light directed through the fibre array to provide a self-aligning process. Transmission has been increased threefold compared with earlier metal layer methods, which is of particular value for long wavelength sensitivity where the cathode efficiency is low. By allowing the use of more conductive pathways the more critical values of the response times have been reduced from the nanosecond to the picosecond range for these imaging tubes.