The ultimate goal of all development of photosensitive detectors is to find a detector capable of detecting single photons with high efficiency. Furthermore, the photon shall not only be detected as a photon somewhere. We want to know where it was with high precision in space, often down to a few micrometers. We want to know when it was there, preferably with a precision of less than a nanosecond. We want to know where and when for each individual photon in a high flux of photons. Sometimes we even want to know the polarization of each photon. Position-sensitive gaseous photomultipliers filled with photosensitive vapours are capable of all of this. It is a challenging task. A single photon is the weakest light there is. For UV and visible light the energy in the photon is so low that it can barely emit a single electron through photoelectric effect with a gas. This photoelectron has practically no kinetic energy when it is released. A single electron at rest is the weakest possible electrical signal there is, so the detector must be able to amplify this extremely weak signal without any noise. We will here describe the history of photosensitive gaseous detectors, their applications and what the state of the art technology is today.
Top1. Introduction
A breakthrough in the development of instruments for detecting individual photons was the invention of the first position sensitive gaseous photomultiplier. As often happens in science, the first prototypes of these detectors were developed by two independent groups, one at CERN (Séguinot, 1977) and the other one in P.L. Kapitza laboratory in Moscow (Bogomolov, 1978). The two historical papers describing these first instruments were submitted in 1976 only two months apart. Both teams used the newly invented multiwire proportional chamber (MWPC) as amplification structure (Charpak, 1968). The breakthrough was to fill the MWPC with vapours with low ionization potential, which made the detector photosensitive. A photosensitive gas had been used long time ago in a single-wire counter (Chubb, 1955). However, this detector had no position resolution. It could just tell whether there was light or not. The big step was the capability of the MWPC to determine a two dimensional position of each avalanche and hence the position of each absorbed photon. These photomultipliers became the first UV imaging detectors.
The photosensitive MWPC has several other important features. For example, one can build a photosensitive MWPC detector with a very large sensitive area (~m2). As it is operated at atmospheric pressure there are no mechanical constraints on the size of the window. The detector can be used with a continuous flow of gas, which makes it less sensitive to outgassing, oxygen leaking in and other impurities in the gas.
There were some important differences between the MWPC designs used by the two groups (Séguinot, 1977; Bogomolov, 1978). In the first experiments Séguinot et al. used a small standard MWPC with an anode wire diameter of 20 μm. The detector was either filled with a mixture of argon (Ar) with 11 Torr of benzene vapour or argon +99Torr of CO2+11Torr of benzene vapour. In both cases the total pressure was 1 atmosphere. The ionization threshold of benzene (Ei) is 9.15 eV, meaning that UV photons with a wavelength shorter than 137 nm will be able to ionize the gas and liberate a photoelectron, which in turn can trigger an electron avalanche in the MWPC.
The gas gains in these particular gas mixtures were limited by photon feedback. There are many more UV photons produced by de-excitation of atoms in the electron avalanche than secondary electrons. Hence, if the photosensitive vapour is sensitive to the wavelength of these secondary photons and the absorption length is long enough, the secondary UV photons will trigger secondary avalanches. The maximum achievable gas gain with the two gas mixtures was 104 and 105, respectively.
All early MWPC designs had this drawback. At high gas gains the photon feedback frequently caused discharges to appear at the dielectric surfaces supporting the anode wires and the cathode mesh. Later, to avoid feedback and discharge problems the CERN group introduced a multistep photosensitive chamber. It was a MWPC combined with one or two parallel plate avalanche chambers made of meshes (Charpak, 1979). This device was successfully used for detection of Cherenkov photons and was the foundation for the first RICH detectors. Eventually, the MWPC was modified and made capable to operate at high gas gains without photon feedback and without a preamplification stage.