Introduction
The development of micro – nano electronics and new technological possibilities for the manufacture of complex semiconductor structures stimulate further study of new optical and photovoltaic phenomena in active film elements.
Currently, oxides and nitrides of semiconductors and semiconductor films grown on their surfaces are widely used in the manufacture of multichannel photovoltaic converters and other active elements of microelectronics circuits, and in particular, optoelectronics. In this case, it is possible to obtain high-quality and dielectric layers of semiconductors with deep levels. At the same time, it is easier and cheaper to use polycrystalline films sprayed on amorphous substrates rather than epitaxial ones.
CdTe semiconductor films are an important material for the creation of photodetector devices based on heterostructures operating in the near (up to 3 microns) and far (8—14 microns) The IR range. It is of interest to obtain heterostructures based on photosensitive layers with different types of conductivity. A promising p-type material doped with silver and copper, which give an acceptor level in the forbidden zone with a long lifetime of non-main charge carriers [1—14].
The aim of the work is to study new photovoltaic properties of active CdTe thin films and heterostructures in a system with SiO2-Si under conditions of specific external influences.
The results of experimental studies of the photovoltaic properties of textures from sprayed layers of CdTe – SiO2 – Si, etc., allow the development of new devices based on polycrystalline films with controlled properties.
Below we investigate the photosensitivity of the CdTe – SiO2 – Si structure, which can be used, for example, as a metal – silicon nitride oxide semiconductor (MNP) – a transistor with a polarized dielectric [1,2], which allows electrical rewriting of information.
Experimental results
Polycrystalline (grain sizes are 0.05—0.1 microns) CdTe films were obtained on the surface of SiO2 – Si. CdTe and Ag and Cu impurities evaporated in a vacuum of 10—5 mmHg from separate evaporators onto the heated oxidized surface of Si. The relative arrangement of the layers of the CdTe – SiO2 – Si structure and the ohmic contacts to them is schematically shown in Fig.1. In such a structure, photosensitivity is controlled by external influences, such as an electric field or corona discharge, which change the built-in field in the dielectric. In this case, we have a «reverse» field – effect transistor of the CdTe – SiO2 – Si type, when the control charge is located under the semiconductor layer, and its surface remains open.
Fig.1 The relative position of the layers of the CdTe – SiO2 – Si structure. 1,2 – contacts; 3 – filtering contacts.
Currently, electrification using a corona discharge is the main method of sensitizing photovoltaic layers in industrial electrography [3].
An experimental setup was used for corona electrification of the studied structures, the block diagram of which is presented in [4]. Electrification occurs due to deposition of positive or negative ions in a corona discharge on the surface of the layer. Corona discharge occurs if the voltage between the metallized surface of the Al layer and the electrode exceeds 6 kV, when the field embedded in the structure reached 100 V. The spectra of the short – circuit current charged in this way in the CdTe – SiO2 – Si structure were studied depending on the magnitude of the external corona discharge and showed that in the static mode a shift of the spectra to the short-wave region is observed (Fig.2). It turned out that in such a structure, the photosensitivity of the layer can be controlled by the action of an external corona discharge potential (using the «field effect» method), which, as it turns out below, induce embedded electric charges in the dielectric.
In Fig.2. The spectral dependences of the short-circuit current (Icz) of the CdTe layer for various values of the corona discharge intensity, which were carried out by contact (2) and electric probe contact (3) to the surface of the CdTe semiconductor, are presented. It can be seen that in the absence of external influences in the Icz (v) spectra, an inversion of the Icz sign is observed in the vicinity of the light quantum energy value equal to hν= 1.21eV (curve 1) the inclusion of the surface corona discharge potential between the CdTe layer and silicon leads to a significant change in the spectral sensitivity of the short-circuit current (Icz). When the surface potential changes within its value from 0 to 100 V, the inversion position of the short-circuit current sign will mix into the short-wave region of the spectrum. In this case, the maximum photosensitivity of the Icz will be mixed into the short-wavelength region of the spectrum in the range from 0.93 eV to 1.5 eV. The position of the maximum value of the Icz increases by more than 1000 times at 70 angstems (curve 3).
Fig.2. Spectral dependences of the Icz for the CdTe-SiO2-Si structure on the magnitude of the corona discharge potential: jcr = 0 V (curve 1), 40 V (2),70 V (3). The inset shows the photosensitivity spectra of the impurity region of light absorption on a logarithmic scale.
Discussion of the results
For a qualitative description of the physical nature of the transfer phenomenon occurring in the CdTe – SiO2 – Si (semiconductor – oxide – semiconductor, i.e. POP) structure when a voltage is applied to it, consider a model in which a stationary current consists of a stream of electrons tunneling from the conduction band of a semiconductor into a deep level located in the oxide (and including the trap at the interface). Since the thickness of the silicon oxide in the structure under consideration is 0.4 microns, according to our estimates, the first contribution to the total flow is insignificant (less than 25%).
Tunneling of current carriers from the CdTe film into deep levels of silicon oxide leads to a change in the filling of the surface state. The latter, depending on the magnitude of the built-in charge, modifies the potential relief of the structure. So that the photogeneration rate will depend on the magnitude of the built-in charge, i.e. on the magnitude of the corona discharge potential in the structure. This means that the magnitude of the photo-EMF will be determined by the degree of asymmetry of the potential relief.
For a qualitative description of the physical nature of the kinetic phenomenon in the structure of semiconductor CdTe – oxide semiconductor SiO2 – semiconductor Si, a model based on the theory of a TIR (metal-dielectric-semiconductor) transistor can be considered. In this case, we mean that in a thick (0.4 µm) oxide layer, the main mechanism of current flow is determined by the Fowler – Nordheim model [5] and the corresponding current is denoted as
where i is the emission current density, E is the electric field strength, φ is the output operation, functions a and b depend on the geometry and operation of the output, for example, the degree of asymmetry, height, and width of the potential barrier. The current carrier flow should occur: a) due to the