Semiconductor Detector

in nuclear physics, a device for detection of ionizing radiation; its principal component is a semiconductor single crystal. A semiconductor detector operates similarly to an ionization chamber. In the semiconductor detector, however, ionization occurs in a layer of the crystal rather than in a gas-filled chamber. A semiconductor detector is a semiconductor diode to which a reverse (cutoff) voltage of ~ 10² volts is applied. The layer of the semiconductor that is adjacent to the boundary of the p-n junction has a space charge and is depleted of the charge carriers (conduction electrons and holes); the layer consequently has high specific resistivity. A charged particle that penetrates this layer produces additional (nonequilibrium) electron-hole pairs. Under the influence of the electric field, these pairs are reabsorbed by migrating to the electrodes of the detector. As a result, an electric pulse is generated in the external circuit of the detector. This pulse is subsequently amplified and recorded (see Figure 1).

Figure 1. Semiconductor detectors: (a) Si surface-barrier detector, (b) drift-type Ge-Li planar detector, (c) Ge-Li coaxial detector; (n) semiconductor region with n-type conductivity, (p) semiconductor region with p-type conductivity, (i) semiconductor region with intrinsic conductivity; the sensitive volume is indicated by hatching

The charge collected on the electrodes of a semiconductor detector is proportional to the energy released by the particle in passing through the depletion layer, which is the sensitive volume of the detector. Therefore, if the particle is completely decelerated within the sensitive volume, the detector can be used as a spectrometer. The mean energy required to form one electron-hole pair in a semiconductor is low: 3.8 electron volts (eV) for Se and —2.9 eV for Ge. As a result, if a material of high density is used, a spectrometer can be obtained with a high resolution: ~0.1 percent for an energy of ~1 million electron volts (MeV). If the particle is fully decelerated within the sensitive volume, the detection efficiency is ~ 100 percent. The high mobility of the charge carriers in Ge and Si permits the charge to be collected in ~ 10 nanoseconds; thus making for the high time resolution of semiconductor detectors.

The first semiconductor detectors, which were constructed in 1956 and 1957, used surface-barrier junctions (Schottky diode) or alloyed p-n junctions in Ge. They had to be cooled in order to reduce the level of the noise resulting from the reverse current; and their sensitive volume was quite shallow. Such detectors did not come into wide use. In the 1960’s, semiconductor detectors based on a surface-barrier junction in Si (Figure 1, a) found practical application. For a surface-barrier detector the depth W of the sensitive volume is determined by the magnitude of the cutoff voltage V:

where p is the resistivity of the semiconductor in ohm-cm. For surface barrier junctions in Si with p = 10⁴ ohm-cm at V = (1–2) × 10² volts, W = 1 mm. These semiconductor detectors exhibit low noise levels at room temperature and are used for recording short-range particles and for measuring specific energy losses dℰ/dx.

Semiconductor detectors of the p-i-n type (Figure 1, b) were developed in 1970 and 1971 to record long-range particles. An Li impurity is introduced into a p-type Si crystal. Li ions move through the p-region of the junction under the action of an electric field. These ions compensate the acceptors and thus create a wide sensitive region of intrinsic conductivity: The depth of this region is determined by the depth of diffusion of the Li ions and can reach 5 mm. Such drift-type Si-Li detectors are used to record, for example, protons with energies up to 25 MeV, deuterons with energies up to 20 MeV, and electrons with energies up to 2 MeV.

The next step in the development of semiconductor detectors was a return to Ge, which has a greater atomic number Z than Si and is therefore more effective for recording gamma radiation. Ge-Li planar semiconductor detectors of the drift type are used in recording γ-quanta with energies of several hundred kiloelectron volts. To register γ-quanta with energies up to 10 MeV, coaxial Ge-Li detectors are used (Figure 1, c); their sensitive volume can reach 100 cm³. The efficiency of recording γ-quanta with energies < 1 MeV is a few tens of percent. This efficiency falls to 0.1–0.01 percent for energies > 10 MeV. For high-energy particles whose range exceeds the sensitive volume, semiconductor detectors not only record the passage of the particle but permit the specific energy losses dℰ/dx to be determined. Position-sensitive semiconductor detectors even permit the x-coordinate of the particle to be determined.

The disadvantages of semiconductor detectors are a low detection efficiency for high-energy γ-quanta, a deterioration of resolution for loads of more than 10⁴ particles per second, and a limited service life of the detector at high doses of radiation because of the accumulation of radiation defects. The small dimensions of available single crystals (diameter ~ 3 cm, volume ~ 100 cm³) limit the use of semiconductor detectors in a number of areas.

The next stage in the development of semiconductor detectors is connected with the production of superpure large-sized semiconductor single crystals and the possibility of using GaAs, SiC, and CdTe. Extensive use is made of semiconductor detectors in nuclear physics, the physics of elementary particles, chemistry, geology, medicine, and industry.