RU83142U1 - OPTICAL MICROWAVE PULSE GENERATOR - Google Patents

Abstract

1. An optical generator containing a diode structure comprising an ν layer and electrodes made on opposite faces of the ν layer, as well as a bias voltage source electrically connected to the electrodes of the diode structure and a control light source, is optically coupled to one of its faces, according to utility model, said structure additionally contains two nanoscale monolayer dielectrics located symmetrically with respect to the ν-layer between the electrodes made in the form of optically transparent electrodes on the surface which are coated with antireflection and contain two dielectric guard rings with respect to the ν layer, each of which encloses a nanoscale dielectric monolayer, an optically transparent electrode and an antireflection coating along the perimeter, and the control light source is optically coupled to one of the antireflection coatings, while the ν layer is made based on a high-resistance direct-gap semiconductor crystal with intrinsic conductivity, for which the product of the equilibrium concentration of carriers by the length of the sample satisfies the condition: n0L <1012cm-2. ! 2. The optical generator according to claim 1, characterized in that it further comprises a second control light source parallel to the first and similar to it, the axis of radiation of which is perpendicular to the plane of the second antireflection coating.

Description

The proposed utility model relates to the field of microwave technology, specifically to solid-state optical sources for generating microwave oscillations of the photocurrent and can be used in the equipment of information processing systems for various purposes for optical generation and control of the spatio-temporal characteristics of the microwave signal.

A number of devices are known that are used to generate electric microwave pulses. Among these devices, along with tunnel diodes, microwave transistors, and avalanche-span diodes, are Gunn diodes (Gunn generators), which are homogeneous semiconductor structures with ohmic contacts, created on the basis of semiconductor crystals or epitaxial films that are different in their electrophysical characteristics. To control the parameters of microwave oscillations generated by Gunn diodes (frequency, amplitude, threshold generation voltage (generation field E _p (x)), efficiency, shape, etc.), various diode operating modes are used, as well as their switching circuits external circuit. In addition, in some cases, to control the parameters of the generated oscillations, in addition to the electric signal, an additional influence is applied to the diode of factors such as the magnetic and microwave fields, backlight, temperature, etc., which affects the spectral composition and parameters of the generated oscillations or can lead to to a change in the conditions of generation excitation [see: Levinstein M.E., Pozhela Yu.K., Shur M.S. Gann effect. - M .: Sov. Radio, 1975. - 288 p.].

One device in which the backlight allows controlling the excitation conditions and the parameters of the generated oscillations is the Gunn diode [see: Adams RF, Schulte HJ Optically triggered domain in GaAs Gunn diodes. - Appl. Phys. Lett., 15 (8), p. 265-267, 1969]. The active element of such a diode is a planar structure, including an n-GaAs epitaxial layer grown on a GaAs: Cr substrate with a dopant concentration N _D nonuniformly distributed over the area of this layer, varying from 2.35 × 10 ¹⁴ to 2.35 × 10 ¹⁵ cm ^-3 , and ohmic contacts made on opposite faces of the epitaxial layer at a distance

L = 100 μm from each other. In addition, this device includes a pulsed voltage source and a source of light pulses (λ = 0.63 μm, P _imp. = 1.5 mW), optically coupled to an n-GaAs layer, the voltage pulses and optical radiation supplied to the active element are synchronized with each other.

The optical control of the generation threshold field value E _p (x) and the generation frequency in the analog device is based on a change in the backlight of the internal electric field strength E (x) in the diode structure. In the absence of external illumination, the application of a pulsed voltage to the diode leads to the appearance of a uniformly distributed internal electric field E (x) inside the n-GaAs layer, the value of which is slightly lower than the threshold generation field E _p (x) in the domain nucleation region. Under these conditions, illumination of the n-GaAs layer in the anode region and most of its active region, with the exception of the region adjacent to the cathode, the light pulse synchronized in time with the applied voltage pulse, leads to an insignificant redistribution of the electric field strength inside the diode, its increase in cathode to the value of E _p (x), which leads to the excitation of a single domain in the diode structure.

The main disadvantages of such a device include the impossibility of exciting current oscillations only due to exposure to illumination, as well as the impossibility of optical control of the parameters of the generated oscillations.

The closest in technical essence and the achieved positive effect, the prototype is the device [see: Gontar V.M., Egiazaryan G.A., Rubin BC, Murygin V.I., Stafeev V.I. Negative differential conductivity in high-resistance gallium arsenide under illumination. - FTP, v.5, v.6, p.1061-1066, 1971], containing a diode structure made in the form of a multilayer n ⁺ -ν-n ⁺ structure, including a high-resistance ν-layer, created on the basis of a volumetric compensated single crystal n -GaAs with a concentration of free carriers n ₀ = 10 ⁷ -10 ⁸ cm ^-3 and impurity levels N _A (N _A ≈10 ¹⁷ ÷ 10 ¹⁸ cm ^-3 ), electrical contacts made on the outer surface of two opposite faces of the ν-layer in the form of an n ⁺ -layer. In addition, the device also includes a source of pulsed bias voltage, electrically coupled to the electrodes of the diode structure and a monochromatic control light source, optically coupled to one of the faces of the structure containing the n ⁺ layer, and the oscillation is generated at a temperature of T = 203 ° K.

The action of the prototype device is based on the formation of recombination (concentration) nonlinearity in the n ⁺ -ν-n ⁺ diode structure under the influence of pulse voltage and lighting. A bias voltage is applied to the electrodes of the diode structure, providing an internal field strength E (x) of 1 to 5 kV / cm in the ν layer. Illumination of the surface of one of its faces with white light of intensity I from 1 to 4 × 10 ³ lux (P _{imn. Ill.} ~ 6 mW) leads to the generation of nonequilibrium photocarriers in its contact area. Because Since the semiconductor component of the n ⁺ -ν-n ⁺ structure contains a large concentration of impurity levels N _A , then at high internal electric field strengths E (x) and external illumination, a situation is realized in which an equilibrium shift occurs in the ν-layer of the diode structure side of the strong predominance of the processes of recombination and capture of nonequilibrium carriers over the processes of their thermal and optical generation (in a compensated semiconductor, this occurs due to an increase in the capture cross section and / or changes in capture coefficient with increasing electric field strength, etc.). As a result, with an increase in the internal field dE (x), due to the intense capture of carriers at deep impurity levels of the ν layer, the change in the concentration of current carriers dn in the conduction band of the diode structure occurs much faster than the change in their mobility dμ, with an increase in the field, i.e. . dn / dE (x) »dμdE (x), which leads to a significant change in the ratio between carrier concentrations. The concentration of free carriers, which determine the conductivity of the diode structure, drops sharply, and the concentration of carriers trapped at impurity levels increases. Correspondingly, this leads to a decrease in the conductivity of the diode structure and, as a result, to the appearance of negative differential conductivity (NDC) dn / dE (x) <0 in it, upon the occurrence of which a sharp falling section appears on the current – voltage characteristic (I – V) of the diode structure . As the field increases, the conductivity in the diode structure decreases until the kinetics of the generation-recombination processes is restored in its semiconductor component, causing a shift in the direction of the dominance of the generation processes over recombination (and / or capture). This kind of “reverse” shift, accompanied by an increase in conductivity in the compensated semiconductor, occurs due to the complete filling of the impurity levels, or when the value of the internal field E (x) reaches a value comparable to the field of the impurity level E * (for GaAs E * ~ 10 ⁴ ÷ 10 ⁵ V / cm). In the first case, this occurs due to a decrease in intensity

capture taking place as deep impurity levels are filled, in the second - with a sharp increase in the processes of carrier autoionization from deep impurity levels (backward ejection). In both the first and second cases (when both mechanisms are implemented together in a compensated semiconductor), the conductivity increases in the diode structure until the differential electron mobility again becomes positive dn / dE> 0 (in this case, on the I – V characteristic of the diode the structure corresponds to an increasing portion and a saturation portion), after which the formation of a recombination domain occurs in the diode structure, the periodic formation of which leads to the appearance of intrinsic low-frequency current generation oscillation circuit comprising a diode structure. At the same time, high illumination intensities (I ~ 10 ⁴ lux) under conditions of large internal fields E (x) can lead to the appearance of subharmonics in the output spectrum of generated oscillations in the Hann diodes created on the basis of compensated semiconductors. Under conditions of strong illumination, their appearance is associated with the formation in the conduction band of the ν layer of an excess photogenerated charge of nonequilibrium carriers, which can no longer be captured by filled impurity levels. The appearance of such a charge in the conduction band causes the joint generation of two types of oscillations: stable natural low-frequency oscillations and unstable higher-frequency ones.

Moreover, both under lighting conditions and without lighting, the frequency of the generated recombination domains is low f ~ 10 ³ ÷ 10 ⁴ Hz, since it is determined not only by the low carrier velocity υ ~ 2.5 × 10 ^-2 cm / s, but also by the speed the occurrence of relatively slow metabolic processes associated with the capture of carriers at impurity levels and their release into the conduction band.

Thus, the prototype device provides, when illuminated at low temperatures (T = 203 ° K), only its own low-frequency generation of electrical pulses and at high illumination intensities and internal fields, unstable higher-frequency generation (f≤10 ⁷ Hz in the form of a subharmonic). In addition, the frequency of self-generation in transit mode in the prototype device changes by 1-2 octaves when the backlight intensity changes or the magnitude of the voltage applied to the structure changes.

At present, there are no optical devices for generating microwave pulses of the photocurrent that provide the generation of microwave pulses at room temperature

only due to the influence of illumination, as well as allowing optically and electrically control the parameters of the generated oscillations under various modes of generation and the magnitude of the threshold generation voltage (threshold generation field).

The utility model is aimed at solving the problem of creating an optical microwave pulse generator, which allows the generation of microwave pulses only due to the influence of illumination, optical and electrical control of the parameters of the generated oscillations under different generation modes, and also increase the frequency of the optically generated oscillations.

To solve the problem in an optical generator containing a diode structure, including the ν-layer and electrodes made on opposite faces of the ν-layer, as well as a bias voltage source electrically connected to the electrodes of the diode structure, and a control light source optically coupled to one of of its faces, according to a utility model, said structure additionally contains two nanoscale monolayer dielectrics located symmetrically with respect to the ν-layer between electrodes made in the form of optically transparent electrodes, on the surface of which an antireflection coating is made, and contains two dielectric guard rings relative to the ν layer, each of which covers a perimeter of a nanoscale dielectric monolayer, an optically transparent electrode and an antireflection coating, and the control light source is optically coupled to one of the antireflection coatings, in this case, the ν layer is made on the basis of a high-resistance direct-gap semiconductor crystal having intrinsic conductivity, for which the product of the equilibrium carrier density at the sample length satisfies the condition: n ₀ L <December ¹⁰ cm ^-2.

To solve the problem of increasing the frequency of generated oscillations, the optical generator further comprises a second control light source parallel to the first and similar to it, whose radiation axis is perpendicular to the plane of the second antireflection coating.

The essence of the proposed technical solution is based on a new type of current instability discovered by the author - volumetric photocurrent drift domain instability. Experimental studies of the photoelectric properties of homogeneous structures with non-injecting contacts, created on the basis of direct-gap high-resistance semiconductors with intrinsic or close to intrinsic

conductivity, showed that for any polarity of the applied voltage V ₀ and illumination with its own light with quantum energy hν≥E _g , where hν is the quantum energy, E _g is the semiconductor band gap, the flow of a monopolar photocurrent limited by the space charge (PFET), leads to spatial rearrangement of the electric field in the volume of such a structure. As a result, for the duration of illumination in the volume of its semiconductor component, a non-illuminated electrode forms a static region of a “strong electric field”, which leads to a change in the initial distribution of the electric field, the magnitude and final profile of which are determined by the power of the illuminating pulse and the magnitude of the voltage applied to the structure. In this case, the magnitude and inhomogeneous distribution profile of the electric field in the structure does not change while maintaining the power (intensity) of the affected lighting and the magnitude of the applied voltage. After the cessation of lighting, an inhomogeneous electric field in the structure spontaneously returns to its initial dark homogeneous state during the swelling of the accumulated charge of free photocarriers.

In the course of further studies, it was found that at bias voltages that provide an inhomogeneous field in one of the parts of the structure The occurrence of FTOPZ leads to the formation of an NDC in the structure and, as a consequence, to the appearance of a photocurrent domain instability of a drift species for the duration of the photocurrent in the structure. Moreover, in the absence of external illumination, the appearance of generation in such a structure does not occur at any values of the applied voltage, up to electrical breakdown of the structure. In addition, it was experimentally established that such characteristics of oscillations excited by lighting as frequency, period, amplitude, shape, etc., as well as the generation modes, differ significantly both from oscillations excited in Gunn diodes under conditions of external illumination, and from current oscillations excited in classical Gunn diodes by an electric or microwave field.

An analysis of the results of experimental studies from the standpoint of nonlinear dynamics shows that spatial photostimulated rearrangement of the electric field underlies both the formation of NDCs under the influence of illumination and the appearance of photocurrent drift domain instability. Analytically spatial photostimulated rearrangement of the electric field is described as a stationary linear or quasilinear response arising in a tunnel

MIS system for the duration of the lighting, which leads to a change in the value of several macro- and microparameters characterizing the tunnel MIS system. However, this kind of response does not cause irreversible changes in it, related to both the equilibrium state in the system as a whole and the state of the electric field, because for the duration of the illumination, the MIS tunnel system in a self-consistent manner is capable of stably maintaining its new quasi-equilibrium state for an arbitrarily long time [see: Yu. N. Perepelitsyn Methods of optical modulation of light fluxes based on spatial photostimulated restructuring of the electric field. - In Sat: Directions for the development of laser and millimeter systems and tools in information transmission technology and medicine. / Proceedings of the NTTC /, Voronezh, 1995, p.1-14].

At the same time, the generation of microwave oscillations of the photocurrent in a high-resistance tunneling MIS system should be considered as a new nonlinear form of the photoelectric response that spontaneously arises in it for the duration of the illumination after, due to optical and / or electrical influence, at one or several values characterizing the equilibrium the state of a macroscopically homogeneous tunneling MIS system, there is a significant excess of certain values at which the tunneling MIS system is no longer capable of self-consistency in this way, it is stable to maintain its former equilibrium or quasi-equilibrium state. As a result, the state of quasiequilibrium in the volume of a macroscopically homogeneous tunneling MIS system becomes unstable and a spontaneous transition to another sharply inhomogeneous state occurs in it. Such a transition is accompanied by a qualitative change in several of its macro- and microparameters. First of all, a change occurs in one of its main internal macroparameters - the distribution of the electric field, and its change occurs spasmodically: from the initially nonuniformly distributed throughout the entire diode structure, to another sharply inhomogeneous, spatially narrowly localized in one of its parts (in the form of a domain strong electric field). In addition to the change in the field distribution profile, the jump-like transition is also accompanied by another qualitative change. If earlier, during the formation of a linear or quasilinear response under conditions of constant illumination and bias voltage, a nonuniform but stationary field distribution profile was formed in the tunnel MIS system then with a nonlinear response under the same conditions, a spontaneous periodic formation of a dynamic response occurs in it, which is in the form of a photoelectric

domain (or several at once) periodically moves from one contact to another.

The main external macro parameters that can be used to characterize the state of the diode system include: L ₁ is the distance between the contacts, I is the intensity (power P) of the illumination, ρ is the semiconductor resistivity, V ₀ is the applied voltage, N _A , N _D is the concentration of acceptor and donor levels, n ₀ is the equilibrium concentration, T is the temperature, etc. The number of internal macro- and microparameters includes: E (x) = V ₀ / L ₁ - value of the internal field, P ₀ - threshold power ( or the intensity I ₀₎ of the control cally signal providing optical injection unipolar free charge photocarriers in which a diode structure is performed Kramer criterion: n ₀ L ₁ ≥10 ¹² cm ^-2, and, accordingly, there is the occurrence of generation, V _n (x) - lasing threshold voltage E _n (x) = V _p (x) / L ₁ - threshold generation field, - non-uniform electric field, - carrier mobility in an inhomogeneous field outside the domain, J _f - photocurrent density, - drift velocity of the main carriers in an inhomogeneous field , - mobility of the main carriers in an inhomogeneous field , τ _n , τ _p is the lifetime of the main and minority carriers, d is the thickness of the nanoscale dielectric monolayer, γ is the transparency coefficient of the tunnel barrier, etc.

The results of experimental studies show that the external and internal macro- and microparameters, as well as the relationships between them, have a complex and interrelated effect not only on the conditions for the occurrence of photocurrent drift domain instability, but also on the final characteristics of the microwave oscillations of the photocurrent (amplitude, domain shape, frequency, period, coherence, etc.). However, in spite of this, a number of relations can be distinguished between the main macro- and microparameters at which a nonlinear photoelectric response of a certain kind (generation modes) stably manifests itself, i.e. domains are generated that have certain spatio-temporal, power and frequency characteristics, which ensures the achievement of a positive effect.

One of the essential features of the claimed device, ensuring the achievement of a positive effect, is the implementation of the ν-layer structure based on high-resistance direct-gap semiconductor with intrinsic or close to intrinsic conductivity. The need for use in a semiconductor component

The diode structure of such a semiconductor is essential for several reasons.

First, to implement the utility model, it is necessary that, under illumination with control light and at any polarity of the applied voltage, the conductivity of the diode structure is completely determined only by its photocurrent component σ _f , where σ _f is the conductivity of the diode structure under illumination. It is possible to provide this condition by fulfilling the ν layer based on a high-resistance intrinsic semiconductor and creating non-injecting contacts in the diode structure, which make it possible to exclude the electrical injection of carriers from the diode structure electrodes into the semiconductor volume with an applied bias voltage V ⁰ and in the absence of lighting. For high-resistance semiconductors, electrically non-injecting contacts are realized only on the basis of Au or Pt metal layers. Such layers are optically transparent for the spectral range of controlled light and, at the same time, provide the highest possible metal to semiconductor work function ratio, which, accordingly, ensures the minimum value of over-barrier emission even at high applied voltages and almost completely eliminates the electrical injection of carriers into the semiconductor bulk. In addition, the presence of non-injecting contacts is a necessary but not sufficient condition for the current in the diode structure to be determined only by the photocurrent component σph. This can only be achieved in the only way - by performing an ν layer based on a high-resistance intrinsic semiconductor. This is due to the fact that only in such a semiconductor under the conditions of an external applied voltage does the illumination lead to the formation of a unipolar space charge of free carriers, which cannot be compensated in the semiconductor by a charge of another sign during its drift transfer from the generation site to the non-illuminated electrode, t .to. the compensating charge cannot come from the external circuit, and it is practically absent in the volume of the semiconductor component. Together, this leads to the fact that in the range of internal fields E (x) (0.1 ≤ E _p (x) ≤ 3), the dark conductivity in the diode structure σ _is negligible, and when illuminated, its conductivity is completely determined by its photocurrent component σ _f . (As estimates show, when changing the intensity of the dark electric field E (x) ~ 0,1≤E _n (x) ≤3 value of _the dark current i varies in a range from 10 ^-9 to 10 ^-7 A. When power pulsed light P ~ 10-20 mW,

the average current in the structure is i _f ~ 10 ^-3 ÷ 10 ^-2 A, i.e. when illuminated, the average change in conductivity is <σ _f / σ by _that. > in the diode structure occurs from 10 ³ to 10 ⁴ times).

Secondly, the generation of microwave oscillations under the influence of an electric field can occur only in diode structures for which the Kramer criterion is fulfilled - n ₀ L≥10 ¹² cm ^-2 , where n ₀ is the equilibrium concentration, L is the length of the sample. The use of an intrinsic direct-gap semiconductor with a low equilibrium carrier concentration n ₀ <10 ⁷ cm ^-3 in the semiconductor component leads to the fact that the Kraemer criterion in such structures is not initially satisfied (n ₀ L ₁ "10 ¹² cm ^-2 ), since at the dark field E (x) ≥E _p (x) and above the equilibrium concentration of carriers contained in the ν layer of the structure is not enough for the emergence of a domain. Therefore, in such a structure, drift domain instability only under the influence of an electric field cannot arise at any values of the external applied voltage. At the same time, the effect of lighting leads to the appearance in the diode structure of the photocurrent i _f , the charge of which ensures the fulfillment of the Kramer criterion and, accordingly, the appearance of microwave generation only when exposed to lighting.

Thirdly, the implementation of the ν-layer in the diode structure based on a high-resistance intrinsic semiconductor, initially containing not only a low concentration of equilibrium carriers along, but in which there are also no deep and shallow impurity levels, allows for illumination and for any internal fields E (x) to exclude the occurrence of metabolic processes in it and, thereby, change the mechanism of the formation of domain instability. Those. switch from the recombination nonlinearity mechanism underlying the formation of NDCs in the multilayer n ⁺ -ν-n ⁺ diode structure to the drift one, in which the NDC and, correspondingly, domain formation occurs only due to a change in the mobility of free photocarriers (dμ / dE) with growth of an inhomogeneous field rather than due to a change in their concentration with increasing field (dμ / dE), as is the case in the prototype, due to the relatively slow exchange processes in the high-resistance ν layer of the compensated semiconductor associated with the capture of carriers to impurity levels and their reverse emission into conduction band.

At the same time, in order to achieve the task in the utility model, it is not enough just to change the mechanism of the formation of domain instability in the diode structure and to provide photocurrent conductivity when illuminated in it.

To implement various modes of generation, optical and electrical control of the parameters of the generated oscillations, as well as to increase the frequency of the optically generated oscillations, it is also necessary to provide in the diode structure:

- the flow of a through photocurrent;

- spatial restructuring of the electric field in the flow of the through photocurrent (both in the case of generation of oscillations of the photocurrent and in their absence), moreover, at the end of the illumination, the relaxation of the photocurrent pulse should occur spontaneously and in a minimum time;

- a one-to-one correspondence between the spatio-temporal and energy characteristics of the control light pulse and the corresponding characteristics of the photocurrent pulse arising from the action of the control optical pulse.

It is possible to ensure the fulfillment of the above conditions at the same time only if the nanoscale dielectric monolayer is satisfied between the ν layer and the metal electrode, and a positive effect will be achieved if the nanoscale dielectric monolayer is performed on each of the two opposite faces of the direct-gap intrinsic semiconductor, and nanoscale monolayers of the dielectric will be identical in their electrophysical characteristics. As experimental studies have shown, the introduction of symmetric nanoscale dielectric monolayers into the diode structure is necessary for a number of reasons.

Firstly, the presence in the diode structure of symmetric nanoscale dielectric monolayers leads to the fact that for any polarity of the applied voltage in each of the near-surface regions of the semiconductor component, an equality is formed between the flow of minority carriers coming from the semiconductor volume, which form the inversion layer, and flowing through the dielectric layer by the flow of carriers, the outflow of which destroys the inversion layer (leakage current through the dielectric), i.e. their presence leads to the establishment of a through-current flow regime in the diode structure. Moreover, the study of the features of the flow of through current in such structures showed that its value depends on the value of the surface potential φ _s and in the absence of illumination, the through dark current i is _that. proportional to the rate of thermal generation of carriers.

As a result, this leads to the fact that in the absence of illumination in a diode structure containing nanoscale dielectric monolayers with high tunnel transparency, even at low bias voltages the magnitude of the dark current i _topics. ceases to be limited to a nanoscale dielectric monolayer and depends only on the rate of interband generation of electron-hole pairs, which determines the rate of inflow of minority carriers of both signs from the semiconductor volume to the corresponding interfaces. Under these conditions, the combined action of two factors — the low equilibrium carrier concentration n ₀ and the complete outflow of thermally generated charge carriers of both signs into the external circuit from the volume of the intrinsic semiconductor — leads to the fact that even at a bias voltage the depletion region extends to the entire thickness of the ν layer (L _0b > L ₁ ), as a result of which almost all the applied voltage is uniformly distributed on the semiconductor component of the structure, and only a small part of it falls on the nanodielectric layers. (As estimates show, in a GaAs structure, the voltage V _d incident on a nanoscale monolayer of a dielectric with a thickness <d> of 5 to 40 nm is 0.01-0.03 V ₀ ). Further increase in voltage applied to the diode structure causes only an increase in the surface potential φ _s since in this case the Fermi level in the metal and the edge of the semiconductor conduction band are shifted simultaneously. Therefore, at bias voltages V ₀ ~ V _p (x) and above, the currents above the barrier emission (i.e., the currents of the main carriers coming from both electrical contacts into the semiconductor volume) remain negligible, since their value is proportional to e- ^φs and, accordingly, they cannot provide an increase in the carrier concentration in the volume of the ν layer to a value at which, according to the Kramer criterion, electrical generation of microwave current oscillations without illumination can occur in the diode structure.

Secondly, the outflow of carriers from the volume of the semiconductor to the external circuit in the absence of illumination provides at a threshold bias voltage V _p (x) in the diode structure a minimum value of the dark current. When illuminated with a control light, this makes it possible to obtain the largest possible ratio of σ _f / σ _themes. and thereby achieve in it not only high photosensitivity, but also expand the region of spectral sensitivity of the control light.

Thirdly, studies of the features of the flow of the photocurrent in GaAs diode structures created on the basis of the intrinsic semiconductor showed that, under illumination and any polarity of the applied voltage, the presence of nanodielectric monolayers leads to the fact that the magnitude of the through photocurrent in them is determined by the balance that is established in each of its surface regions between the photocarriers flowing from the semiconductor volume to the corresponding nanodielectric layer and the photocarriers flowing through the nanosize The black dielectric monolayer. The speed of establishing the balance and the magnitude of the photocurrent flowing through the nanoscale monolayer of the dielectric are determined by two quantities: the intensity of the illumination and the transparency of the tunnel barrier. The first of them, the intensity of the acting illumination, which determines the concentration of minority photocarriers flowing from the semiconductor volume to the corresponding electrodes of the diode structure, is directly proportional to the intensity of the acting illumination. The second, which determines the rate of establishment of the through-flow photocurrent flow and the magnitude of the photocurrent flowing through the nanoscale dielectric monolayer, depends on the probability of carrier tunneling through the potential barrier, is determined by the structural perfection of the layer, the presence of deep and shallow impurities in it, as well as the concentration and speed of free photocarriers, electric field strength at the layer boundary and other factors. Accordingly, this leads to the fact that the achievement of the task for the utility model is provided only with certain electrophysical characteristics of nanodielectric monolayers. It was experimentally established that such layers should be homogeneous monolayers of nanodielectric, which should have the same thickness <d> from 5 to 40 nm over the entire face area, have a low density of structural defects and contain a minimum concentration of deep capture traps in the band gap of the monolayer material and at the same time, have high adhesion to the surface of the ν layer and resistance to breakdown. A study of the features of the flow of the photocurrent in diode structures containing nanoscale dielectric monolayers with the characteristics noted above showed that such layers have a high but finite tunnel transparency, and the transparency of the tunnel layer, which determines the magnitude of the through photocurrent, varies linearly and reversibly in wide aisles with changing intensity impact lighting

and / or bias voltage. As a result, these features of the flow through the photocurrent lead to the following:

- when the exposure to illumination in the flow-through photocurrent mode due to the partial accumulation of charge of free photocarriers at the non-illuminated electrode of the diode structure, the spatial field of the electric field occurs, and for each magnitude of the bias voltage and / or intensity of the illumination at the interface of the non-illuminated face, its values of the through photocurrent are dynamically set and electric field strength;

- when exposed to lighting, an unambiguous correspondence is realized between the spatio-temporal and energy characteristics of the control light pulse and the corresponding characteristics of the photocurrent pulse;

- during the generation of photocurrent pulses and after the end of illumination in the diode structure, the minimum time for spontaneous swelling of the knee charge and the minimum time for the domain to go into the external circuit are ensured, because due to the high speed of the photo-carriers, they are transported through the layer without being captured onto impurity levels of a nanoscale dielectric monolayer.

As a result, under conditions of a constantly applied voltage V ₀ ~ V _p (x), the action of a light signal, power P _{0 of} which provides optical injection of a monopolar charge, at which the Kraemer criterion n ₀ L ₁ ≥10 ¹² cm ^-2 (threshold power), the volume of the ν layer of the diode structure leads to the formation of an inhomogeneous electric field . Reaching them value in any region of the ν-layer of the diode structure leads to the appearance of NDCs in this region (due to the “heating” of a part of free photocarriers in a strong field and their transfer from the lower subzone n _I to the upper n _II ) and, as a result, to the formation of a dipole layer depleted of free photo carriers. The appearance of a dipole layer with increased resistance causes a rapid redistribution of stress in the ν layer of the structure: the voltage incident on the dipole layer increases rapidly, and decreases in the rest of the region of the ν layer. Such a dynamic redistribution of the voltage V _{0 is} accompanied by a decrease in the density of the photocurrent flowing in the structure J _f and an increase in the electric field in the dipole layer, which occurs until it reaches a value at which the differential electron mobility in the diode structure again becomes positive dn / dE > 0. After this, the electron velocities outside the domain and in the dipole layer are compared. Charge accumulation

in the dipole layer, the formed dipole charge layer in the form of a narrowly localized region of a strong electric field (electric domain) drifts in an inhomogeneous field at a rate approximately equal to the velocity of carriers outside the domain, from the place of its formation in the ν layer to the anode contact. Having reached the anode, the domain goes into the external circuit, which is accompanied by the appearance of photocurrent fluctuations in the circuit containing the sample. Moreover, its departure into the external circuit leads to a quick recovery of the inhomogeneous distribution of the electric field and the photocurrent density that existed in the structure before the formation of the domain, and the process of its spontaneous formation is repeated again.

In the transit mode of generation, the frequency of the generated oscillations f that occur at threshold values of the applied voltage V _p (x) and the power of pulsed illumination P ₀ depends only on the length of the sample L ₁ and is determined by the expression: f = 1 / T, where T is the period of oscillation . Accordingly, the period of the generated photocurrent oscillations that occur when V ₀ and / or P ₀ changes is determined by the ratio:

and the distribution of an inhomogeneous electric field and the density of the monopolar photocurrent J _f in the diode structure are determined by the expressions (2) and (3), respectively:

and

where: L ₁ is the distance between the contacts, l _dom. - domain width - carrier mobility in an inhomogeneous field, - carrier mobility in an inhomogeneous field outside the domain, drift velocity of carriers in an inhomogeneous field, χ - relative permittivity, V _ν - voltage is distributed in the region outside the domain structure, E _dom. - electric field strength in the domain.

In addition to the above essential features, also an important feature of the device is the implementation of protective dielectric rings with a width of 250-300

microns on two opposite illuminated faces of the diode structure. Moreover, each of the guard rings should be made along the perimeter of each of the illuminated faces and cover the nanodielectric layer, an optically transparent metal electrode and an antireflective coating, made on the surface of the face. The necessity of introducing the indicated layers into the design of the diode structure is due to the fact that, in a number of modes, the generation of photocurrent oscillations occurs under conditions of large internal electric fields, and during generation in addition to the internal field another domain field is added. (For example, in a GaAs-based diode structure when lighting forms a single domain with picosecond fronts with an average domain width μm the field strength in the domain can reach V / cm, which is comparable to the breakdown voltage, which for GaAs is W / cm). All this leads to the fact that at a depth of complete absorption of control light from 1 to 5 μm in the diode structure, the magnitude of the surface photocurrents sharply increases. Accordingly, at a high density of surface states N _S contained on the surface of the side faces after processing the single crystal, under conditions of high illumination intensities I ₀ and at a high bias voltage V _0, an increase in the surface component of the photocurrent can lead to electrode shorting along the surface of any of the side faces of the structure. Accordingly, the implementation of a dielectric guard ring on the surface of the faces along the perimeter makes it possible to provide "electrical isolation" between the faces of the diode structure containing optically transparent metal electrodes and thereby exclude the possibility of electrical breakdown of the structure. As dielectric layers providing dielectric isolation between the electrodes of the diode structure, dielectric films can be used, for example, SiO ₂ , Si ₃ N ₄ , Al ₂ O ₃ films with a thickness of 5-10 μm, widely used in microelectronics.

In addition, significantly reduce the magnitude of the surface photocurrents when exposed to a control optical signal allows the implementation of an antireflection coating. Its application on top of optically transparent electrodes is also necessary because the inclusion of this layer can significantly reduce the intensity of the control optical signal acting on the structure due to the almost complete suppression of light reflection from the illuminated face of the structure. (For example, for GaAs containing only optically transparent on a polished surface

metal electrode, reflection loss is η≈40-45%). Therefore, the implementation of an antireflection coating makes it possible not only to reduce surface photocurrents, but also the threshold power of the control optical signal, at which photocurrent pulse generation occurs. As an antireflection coating, a Si ₃ N ₄ film can be used, the thickness of which is a quarter of the wavelength of light illuminating the diode structure.

Thus, the combined effect of all the above essential features leads to the fact that the new population, in comparison with the prototype, has a number of new properties:

- the formation of a variable negative negative photoconductivity, leading to the formation of a volumetric photocurrent domain instability of a drift type only due to exposure to light;

- the formation by illumination of the inhomogeneous profile of the distribution of the electric field, which also remains inhomogeneous in the ν region of the structure outside the domain when a domain occurs and is transferred through the structure;

- parametric dependence of the magnitude of the NDC, the intensity of the inhomogeneous electric field and fields outside the domain , as well as the nature of their distribution from the main external and internal macro- and microparameters characterizing the diode structure

To increase the frequency of generation in the domain suppression mode, the microwave microwave optical generator additionally contains a second source of control optical pulses located parallel to the first and similar to it, the axis of radiation of which is perpendicular to the second face of the diode structure containing a nanoscale dielectric monolayer and an antireflective coating. In this case, both sources of control light must ensure the formation of control pulses both independently of each other and together.

The essence of the utility model is illustrated by drawings, in which Fig. 1 shows a general diagram of an optical microwave generator that generates photocurrent pulses for various modes of generation and illumination by rectangular pulses of light 16, 17, 18, as well as complex control pulses (triangular 19, sawtooth 20, sinusoidal 21, meander 22), in Fig.2 - separately, shows the design of the diode structure, Fig.3 shows a photograph of the waveform of the pulse

photocurrent generated under simultaneous illumination by a control pulse of rectangular shape 16 (see Fig. 1, position 16, static domain mode), Fig. 4 shows a photograph of the waveform of a photocurrent pulse generated by simultaneous illumination by a control pulse of rectangular shape 16 and a second pulse rectangular control light 17 (see Fig. 1, position 16, 17, domain suppression mode), Fig. 5 shows a photograph of an oscillogram of a single photocurrent pulse generated when a light pulse illuminates, angular shape 16 (see figure 1, position 16, trigger mode), Fig.6 shows a photograph of the waveform of the photocurrent pulses generated when illuminated by a rectangular light pulse 18 (see Fig.1, position 18, span generation mode ), Fig. 7 is a photograph of the waveform of the photocurrent pulse generated when a triangular shape light illuminates with a light pulse 19 (see Fig. 1, position 19), Fig. 8 is a photograph of a waveform of a photocurrent pulse generated by a sawtooth shape illuminating with a light pulse 20 (cm. figure 1, position - 20), figure 9 presents a photograph of the waveform of the photocurrent pulses generated when the light pulse illuminates a sinusoidal shape 18, (see figure 1, position - 21), figure 10 shows a photograph of the waveform of the photocurrent pulses, generated when illuminated by a periodic light pulse (see Fig. 1, position 22, meander), Fig. 11 shows a photograph of the waveform of photocurrent pulses resulting from the addition of two simultaneously generated photocurrent pulses when illuminated by a rectangular light pulse 18 (see Fig. .1, position - 18, runout mode), Fig. 12 shows a photograph of the waveform of the photocurrent pulses generated when a light pulse of rectangular shape 18 illuminates (see Fig. 1, position - 18, multi-domain mode), Fig. 13 is shown in on a double logarithmic scale, a graph of the average time-time current-voltage characteristics of the diode structure containing the section where the frequency of the generated oscillations changes: I - when a rectangular light 18 illuminates with constant power and changes the magnitude of the bias voltage; II - with a constant value of the bias voltage and a change in the power of the rectangular pulse of the control light 18 (see Fig. 1, position 18, the mode of tuning the frequency of the generated oscillations), Fig. 14 shows a photograph of the waveform of the photocurrent pulse generated by illuminating with a rectangular light pulse 18 at a constant threshold bias voltage

and increasing the power of the control optical pulse (see figure 1, position 18, aperiodic generation mode). All photographs show the time division scale (μs / sm, ns / sm) and the sensitivity (mV / sm, V / sm) of the oscilloscope, at which the shape and time and energy characteristics of the generated photocurrent pulses were recorded.

In figure 1:

1 - ν layer;

2 - nanoscale monolayer dielectric;

3 - optically transparent electrode;

4 - antireflection coating;

5 - dielectric security ring;

6 - electrical contact;

7 - source of bias voltage;

8 - load resistance;

9 - recording device;

10 - source of control light;

11 - optical system;

12 - the second source of control light;

13 - the second optical system;

14 - axis of radiation from a source of control light;

15 - axis of radiation from a second source of control light;

16 is an optical pulse of a rectangular shape formed by a control light source;

17 - a rectangular optical pulse generated by a second control light source;

18 is an optical pulse of a rectangular shape formed by a control light source;

19 is an optical pulse of a triangular shape formed by a control light source;

20 is an sawtooth-shaped optical pulse generated by a control light source;

21 is an optical pulse of a sinusoidal shape formed by a control light source;

22 - periodic optical pulse of a rectangular shape (meander), formed by a control light source.

The diode structure (see Fig. 2) includes an ν layer of a high-resistance direct-gap semiconductor crystal having intrinsic conductivity 1, two nanoscale monolayer dielectrics 2, similar to each other and located symmetrically with respect to the ν layer 1 between the ν layer 1 and optically transparent electrodes 3, two optically transparent electrodes 3, similar to each other, each of which is made on the outer surface of a nanoscale monolayer of dielectric 2, two antireflective coatings 4 similar to each other, each of which is nan hay on the outside of the optically transparent metal electrode 3, two guard rings 5 similar to each other, each of which covers a perimeter nanoscale monolayer of dielectric 2, an optically transparent electrode 3 and an antireflection coating 4, two electrical contacts 6, similar to each other and located symmetrically on one of the faces of the ν layer 1, perpendicular to the face containing a nanoscale monolayer of dielectric 2, an optically transparent electrode 3 and an antireflection coating 4, each of which is electrically connected to optically transparent electrodes 3 and a bias voltage source 7, a load resistance 8 electrically connected to a recording device 9, a bias voltage source 7 and one of the electrical contacts 6.

The axis of radiation 14 from the source of control light 10 is formed by the optical system 11 and is perpendicular to the plane of the antireflective coating 4, and the axis of radiation 15 from the second source of control light 12 is formed by the optical system 13 and perpendicular to the plane of the second antireflection 4.

An optical microwave pulse generator operates as follows. A constant voltage is applied to the optically transparent metal electrodes 3 through the electrical contacts 6 from the bias voltage source 7, the value of which provides a uniform dark electric field in the ν-layer of the diode structure 1 corresponding to the generation threshold field E _p (x). Illumination of the structure by any of the pulses of the control light, the spatio-temporal and energy characteristics of which are formed by the control light source 10 and the optical system 11, leads to the generation of electron-hole pairs and to their separation at the interface of the illuminated face of the ν layer 1 — nanoscale monolayer of dielectric 2 . In this case, one type of media flows into the external circuit through the illuminated

the electrode, and the second drifts under the influence of an electric field to the electrode opposite to the illuminated one, where it partially accumulates in the volume of the ν layer at the interface of the ν layer 1 — nanoscale dielectric monolayers 2. As a result, the partial accumulation of the monopolar charge of free photocarriers at the interface leads for the duration of the illumination to the spatial restructuring of the electric field in the volume of the ν layer 1, which leads to an increase in the electric field strength at the unlit face of the diode structure and, accordingly about, in a decrease in the illuminated face. (For example, when a potential of negative polarity is applied to an optically transparent electrode of a face illuminated by a control light pulse, an increase in the electric field strength in the volume of a high-impedance direct-gap semiconductor crystal occurs in ν layer 1 at the interface near the unlit electrode, to which a potential of positive polarity is applied). In turn, an increase in electric field strength in the near-contact region of the ν layer 1 at the non-illuminated electrode to a value exceeding E _p (x), leads to the formation of NDC in this region and, as a result, to the generation of photocurrent pulses. At the same time, since the final value and the spatio-temporal profile of the inhomogeneous electric field , which is established during illumination in the ν layer 1 of the diode structure, are mutually unambiguously determined by the combination of external and internal macro- and microparameters characterizing the diode structure, then at certain ratios between the value of the voltage V ₀ applied to the diode structure and the spatio-temporal and power characteristics of the control light pulse P various generation modes are implemented in which the optical generator provides the generation of photocurrent pulses of a certain type, and regardless of the mode The generation mechanism of the strong field domains in the optical generator is the same and corresponds to that described above.

Static domain mode (see Fig. 3) - the generation mode in which a static domain is formed is realized in a diode structure with a constant bias voltage V ₀ ≥2V _p (x) and illumination with a rectangular optical pulse 16 formed by a control light source 10 of power P = P ₀ , the duration of which is T _{imp. con.} ≤2t _prol. where t _prol . - time of flight of the main photocarriers through the diode structure.

The domain suppression mode (see Fig. 4) is the generation mode in which, immediately after the formation of the domain, it is extinguished, i.e. the domain does not move in the diode

structure from the region of its formation to the anode contact, which ensures the maximum generation frequency. This mode is implemented in a diode structure with a constant bias voltage V ₀ ≥2V _p (x) and simultaneous illumination by a rectangular optical pulse generated by the control light source 16 and a rectangular optical pulse formed by the second control light source 17, which have the same duration τ _{imp . con.} = t _prol. and power P = P ₀ .

The trigger generation mode (see Fig. 5) is the generation mode in which a single pulse is generated and transferred in the volume of the diode structure from the formation region to the anode contact. This mode is implemented in a diode structure with a constant bias voltage V ₀ ≥V _p (x) and illumination with a single rectangular optical pulse 18, formed by a control light source 10, the duration of which is τ _{imp. con.} »T _prol , with the power of the control optical pulse P = P ₀ .

Span generation mode (see Fig.6) is a generation mode in which high-frequency generation occurs with a constant oscillation period close to the time of flight of carriers, and the generation duration is determined by the duration of the control optical pulse. This mode is implemented in a diode structure with a constant bias voltage V ₀ ≥2V _p (x) and illumination with a rectangular rectangular optical pulse 18 formed by a control light source 10 having a constant power P »P ₀ , when the minimum control optical signal duration τ _{imp is} exceeded _{. con.} > 3t _prol .

The mode of generating pulses of a photocurrent of complex shape is the mode of generation in which the envelope of the generated pulses of the photocurrent is one-to-one repeats the shape of a control optical pulse of complex shape. This mode is implemented in a diode structure when it is illuminated by an optical pulse of any shape formed by a control light source 10: a triangular shape 19 (see Fig. 7), a sawtooth shape 20 (see Fig. 8), a sinusoidal shape 21 (see Fig. 9), periodic pulses of rectangular shape 22 (meander) (see figure 10). In this case, the generation of photocurrent pulses of any complex shape occurs at a constant bias voltage V ₀ = V _p (x) and illumination by light pulses 19, 20, 21, 22 with a power of P = 2 P ₀ , the duration of which significantly exceeds the transit time of the photocarriers τ _{imp. con.} "T _prol .

A multi-domain flyby mode (see Figs. 11, 12) —a generation mode in which several photocurrent pulses are generated simultaneously (two, three, four, beats, etc.) —are implemented in a diode structure with a constant bias voltage V ₀ ≥ 1.5 V _p (x) and illumination with a rectangular optical pulse 18 formed by a control light source 10 having a constant power P≥2P ₀ , with a control optical pulse duration of τ _{imp. con.} "T _prol. . At the indicated values of V ₀ and P, a two-domain span mode is implemented in the diode structure, in which an increase in voltage in the range from 1.65 to 1.8 V _p (x) or in the power of a control light pulse in the range of 2.4-2.6 P ₀ to the occurrence of beats (see 11). The three-domain and four-domain generation mode is realized with a further increase in the bias voltage V ₀ to 2.5 V _p (x) and / or the pulse power of the control light to a value from 3 to 4 P ₀ .

The tuning mode of the frequency of the generated oscillations (see Fig. 13 (I, II) is the generation mode in which, regardless of the shape of the control optical signal, a linear and reversible change in the frequency of the generated oscillations occurs with a linear increase or decrease in the bias voltage and / or power of the control optical signal when they change from threshold values to values at which the generation is interrupted, this generation mode is implemented in a diode structure in two stages. At the first stage, at a constant bias voltage V ₀ = V _p (x) and illuminated by a single rectangular optical pulse of 18 duration τ _{imp. Illumination} »t _prol. , Formed by a control light source 10 at a control optical pulse power of P = P ₀ , a single pulse is generated. depending on the variable external macroparameter, i.e., the change in the magnitude of the applied voltage V ₀ or the power of the control optical pulse 18, as well as when they are changed together, the frequency tuning mode of the generated oscillations (see Fig (I, II) is implemented in various ways. A reversible, linear change in the frequency of the generated oscillations due to a change in the magnitude of the applied voltage V ₀ is realized in the diode structure at P = P ₀ and a linear change in V ₀ from V _p (x) to V _c , where V _c is the voltage at which a breakdown occurs generation, and a reversible change in the frequency of the generated oscillations due to a change in the power P of the control optical pulse 18 is realized in the diode structure with a constant bias voltage V ₀ = V _p (x) and a linear change in the power of the control optical pulse from P ₀ to P _s , where P _s

- the power of the optical pulse at which the generation fails. Also, a reversible change in the frequency of the generated oscillations with a simultaneous change in two external macroparameters V ₀ , P is realized in a similar way with a linear change in V ₀ from V _p (x) to V _c and P from P ₀ to P _s .

Control of the threshold generation voltage (threshold generation field) is a single pulse generation mode in which the threshold generation voltage is controlled by controlling the optical signal power.

This generation mode is realized in a diode structure when a rectangular control light 18 is generated by a control light pulse formed by a control light source 10, with an optical pulse duration τ _{imp. con.} »Τ _prol. moreover, the magnitude of the generation threshold voltage V _p (x) = L ₁ E _p (x) is controlled in a certain range of changes in the external and internal parameters V _p (x), P ₀ , for which a single generation threshold field E _p (x) domain holds the relation E _p (x) = E _P (x) / E _V (x), where E _P (x) is the electric field created by the photogenerated charge of free photocarriers when illuminated with a control pulse of rectangular shape 18, and E _V ( x) is the electric field created by the bias voltage source 7. When this minimum value V _n (x), at which the generation of single domain is provided at a value of V _f (x) = 0,12 E _n (x) and the light pulse of the control light power P _01, and the maximum value V _n (x ) is provided at values of V _p (x) from 0.8 to 0.9 E _p (x) and the power of the control light P ₀₂ , where P ₀₁ , P ₀₂ are the threshold power of the control light pulse corresponding to the minimum and maximum value of the threshold generation voltage V _p (x) at which a single domain is generated.

Aperiodic generation mode (see Fig. 14) - the generation mode in which the power of the control optical signal allows you to control the number and period of the generated pulses - is implemented in the diode structure at a constant V ₀ ~ from 0.1 to 0.15 E _p (x) and changing the power P of the control light pulse of rectangular shape 18 formed by the control light source 10, from P ₀ to P _{with the} duration of the optical pulse τ _{imp. con.} »Τ _prol .

Most of the above modes are similar to the generation modes that are implemented in Gunn diodes by various methods (including during operation

diodes in a custom LCR circuit) [see, for example: Levinstein M.E., Pozhela Yu.K., Shur M.S. Gann effect. - M .: Owls. Radio, 1975. pp. 164-212]. Therefore, the optical microwave pulse generator can be considered as the “optical analogue” of the classic Gunn diode, since it provides the generation of photocurrent pulses, but only when exposed to an optical signal.

Concrete example

Structures M ₁ TD ₁ PTD ₂ M ₂ used as an optical microwave photocurrent pulse generator were fabricated on the basis of intrinsic GaAs (ρ ~ 8 × 10 ⁸ Ohm / cm, concentration of electrically active background impurity N≤10 ¹³ cm ^-3 , n ₀ ~ 7.2 × 10 ⁶ cm ^-3 ).

Electron drift velocity in an inhomogeneous electric field calculated directly on the basis of measuring the magnitude and distribution of the electric field in the volume of the diode structure using the electro-optical technique and in the tested structures was cm / s [see Kasherininov P.G., Kichaev A.V., Perepelitsyn Yu.N., Yaroshetsky I.D. Determination of the distribution of electric field strength in the volume of semiconductor structures based on electro-optical crystals. // Abstract. doc. XII All. conf. in physics of semiconductors. - Kiev, Naukova Dumka, 1990, v.2, p. 36-37]. After cutting the ingot (cutting plane (110)), perfect GaAs single crystals were punctured from the washers, from which subsequently using generally accepted technological methods (grinding and chemical-mechanical polishing) sets of experimental samples made in the form of rectangular parallelepipeds were made. Each set of samples contained a batch of single crystals, which for the same linear dimensions l × h had different thicknesses d (l - length, h - height, d - thickness). The step of changing l, h is 1 mm (from 10 to 1 mm), d is 500 μm (from 2.5 mm to 0.5 mm). After processing on the surface of two opposite faces l × h, homogeneous monolayers of a nanoscale dielectric with a thickness d from 5 to 40 nm were created by etching the samples in an acid mixture (65% HCl + 35% HNO ₃ ) for 35-60 seconds, followed by oxidation of these faces in dry O ₂ atmosphere for 1 hour. Then, after oxidation, by chemical deposition of Au on the surface of the oxidized faces, optically transparent metal electrodes 10–20 nm thick were formed on which an In layer with an area of S ~ 300 μm and a thickness of 2-3 μm was formed at one of the corners of the oxidized face, to which they were soldered

wire metal electrodes. Upon completion of these operations, a layer of Si ₃ H ₄ was applied to each of the faces of the single crystal, which is both an antireflection and a sealing coating. To reduce the surface leakage currents and eliminate short circuits between the electrodes along the perimeter of the face containing the antireflection coating, a SiO ₂ layer with a thickness of 2 μm and a width of 300 μm was applied over the formed layers.

The characteristics of the generated pulses, as well as the operation of the diode under various modes, were recorded on a stand, which is a complex of radio measuring and optical equipment. Two identical optoelectronic modules manufactured by the author were used as a source of optical pulses of control light of a rectangular and complex shape and a second source of pulses of control light. Structurally, the optoelectronic module was a functionally complete unit, including an electronic power supply circuit for a semiconductor laser of the LM3 800-1000 type with a wavelength of λ = 0.8-0.86 μm, the optical radiation from which was output through a standard multimode fiber with a fiber optic connector. The pulses of control light supplied to the generating structure were focused by a microlens optical system, mechanically combined with an optical connector. In addition, the radio-measuring equipment included an AKTAKOM ANR-1003 electric signal generator of complex shape and 2 G5-56 rectangular electric pulse generators, from which electric pulses of the required frequency, duration and shape were fed to electronic power circuits of semiconductor lasers, which ensured the formation of a control optical signal with the required parameters. In addition, the formation of short electrical pulses with a duration of less than 10 ns and with a repetition rate of up to 15 GHz was also provided by electronic power circuits of a semiconductor laser during their operation in a special mode.

The characteristics of the generated photocurrent pulses were recorded at room temperature (T = 20–25 ° C) and the diode operated at a resistive load, i.e. In a coaxial circuit containing a bias voltage source, a precision load resistance R _H = 50 Ohm with a power of 2 W was connected in series with the diode, from which a voltage proportional to the current in the circuit was removed. To record the characteristics of the generated pulses, we used a four-channel oscilloscope of the Tektronix series (minimum time scan 1 ns, sensitivity 1 mV-

100 V, digital memory in the mode of recording single pulses) and an S8-13 oscilloscope with a sensitivity range of 10 ^-4 V - 100 V, to the input of which a coaxial head with a sample could be connected directly. Depending on the operating mode, generating structures having different thickness d and the area of the illuminated face l × h were used. The study of all lasing regimes occurred at the same polarity of the bias voltage, i.e. with a negative potential supplied to the electrode of the face illuminated by the pulses of the control light from the control light source 8, and, accordingly, when applying a positive potential to the electrode located on the opposite side. The leading and trailing edges of the generated pulses were estimated at a level of 0.9, and the signal amplitude — by the magnitude of the voltage taken from the load resistance R _n = 50 Ohms and recorded by the oscilloscope.

The static domain mode (see FIG. 3) was implemented in a structure with linear dimensions l × h × d = 10 × 10 × 1.0 mm. When the static domain mode was realized, a constant voltage V ₀ = 550 V was applied to the diode structure, and then a rectangular light pulse 16 of duration τ _imp was applied to the sample from the control light source 10 _{. con.} = 2 ns, P ₀ = 100 mW with a repetition rate of control pulses of 10 ⁶ Hz. Studies of this mode showed that a single static domain with a peak pulse power P _{imp. Is} generated in the diode structure when illuminated _. ~ 6.48 watts.

The domain suppression mode (see Fig. 4) was implemented in a diode structure with linear dimensions l × h × d = 5 × 5 × 0.5 mm, to which a constant voltage V ₀ = 610 V was applied, and then from the control light source 10, a rectangular light pulse 16 of duration τ _{imp. con.} = 1 ns, P ₀ = 30 mW and a rectangular pulse of light generated by the second control light source 17 having the same duration and power. The delay time between the control pulses of light 16 and 17, as well as their repetition rate (in the range of 100 MHz to 10 GHz) were controlled using synchronized power circuits of semiconductor lasers, forming control pulses 16 and 17. The study of the domain suppression mode showed that the maximum the frequency of generation of a single pulse f = 1.1 GHz is achieved with a temporary shift between the pulses of the control light 16 and 17, comparable with the duration of the generated pulse, which is close to the maximum

the achieved frequency for the structure with the distance between the contacts L ₁ = 0.5 mm The durations of the leading and trailing edges of the photocurrent pulse were estimated at a level of 0.9 and at the maximum generation frequency, respectively, amounted to: τ _{before. fr.} from 50 to 60 rs, τ _{back fr} . ~ 300 rs.

The trigger generation mode (see Fig. 5) was implemented on a diode structure with linear dimensions l × h × d = 7 × 7 × 1.0 at a bias voltage of V ₀ 291 V (E _p (x) ~ 2.6 kV / cm) and illumination with a single pulse of control light 18 of duration τ _{imp. con.} = 1.1 μs with a threshold power of P ₀ = 15 mW. Studies of this mode showed that this mode is implemented at threshold bias voltages and control pulse power, and the generation in this mode is stable, but the generation frequency in this mode is minimal.

The span generation mode (see Fig. 6) was implemented on the same diode structure with a constant bias voltage V ₀ = 600 V and illumination with a control light pulse of 18 constant power P = 35 mW, the duration of which varied from 5 ns to 1.1 μs (ie, at V ₀ V V _p (x) and P ₀ P ₀ , which respectively amounted to: V _p (x) = 291 V and P ₀ = 15 mW). Testing of this mode showed that in the structure with the distance between the contacts L ₁ = 1 mm, the frequency f of the generated pulses is close to 10 ⁸ Hz, and the period of the generated pulses T is constant and close to the time of flight of the carriers.

The mode of generating pulses of a photocurrent of complex shape when illuminated by control pulses of complex shape (see, respectively, Figs. 7, 8, 9, 10) was implemented on a structure with linear dimensions l × h × d = 8 × 8 × 0.5 mm at a constant value applied bias V ₀ = V _p (x) = 130 V and illumination by light pulses 19, 20, 21, 22 with a power of P = 40 mW (P ₀ = 20 mW). To this end, control light pulses of various complex shapes 19, 20, 21, 22 (see FIG. 1) having a close repetition rate and duration were successively supplied from the control light source 10 to the diode structure. Testing of this mode showed that irrespective of the shape of the control light pulse, the envelope of the generated pulses is one-to-one repeats the shape of the corresponding control optical pulse of complex shape.

The multi-domain flight mode (see Figs. 11, 12) was implemented on a structure with linear dimensions l × h × d = 7 × 7 × 1.5 mm. When implementing this mode of magnitude

the threshold voltage V _p (x), the threshold power P ₀ and the duration of the rectangular pulse of control light 18 were selected experimentally and in the diode structure under test, respectively, were: V _p (x) = 590 V, P ₀ = 27 mW with the duration of the control optical pulse τ _{imp . con.} ~ 8.5 μs. Testing this generation mode showed that, at threshold bias voltage and control pulse power, a single-domain generation mode is implemented in the diode structure. The two-domain generation mode, in which two photocurrent pulses are generated simultaneously, having different amplitudes and periods of oscillation, at a maximum shift time between the generated pulses, comparable with the period of their generation, is realized by increasing the bias voltage V ₀ to a value of 1.5 V _p (x ) and the pulse power P of the control light 18 to a value of P≥2P ₀ . After the occurrence of two-domain generation, an increase in the bias voltage in the range from 1.65 to 1.8 V _p (x) or in the power of the control light pulse P in the range from 2.4 to 2.6 P ₀ leads to a reduction in the shift time between the generating pulses and the occurrence of beats (see Fig.11). A further change in the magnitude of the applied bias V ₀ to a value of 2.5 V _p (x) and / or the pulse power of the control light 18 to a value from 3 to 4 P ₀ . leads to the emergence of a stable three- or four-frequency generation of photocurrent pulses (see Fig. 12).

The frequency tuning mode of the generated oscillations (see Fig. 13) was implemented in a diode structure with linear dimensions l × h × d = 6 × 6 × 2.5 mm. When testing this mode, the value of the threshold voltage V _p (x) and the threshold power P _{0 of the} control light 18, at which a single domain is generated, was selected experimentally. For the test structure with a control pulse duration of 20 μs and a pulse illumination frequency of F = 10 ³ Hz, the threshold voltage and pulse power of control light 18 were: V _p (x) = 545 V, P ₀ = 6.5 mW. All characteristics of the photocurrent signal recorded at threshold values of P ₀ and V _p (x) were taken as the reference point both when studying the dependence of the frequency of the generated oscillations on the bias voltage, and when studying the dependence of the frequency of the generated oscillations on the power of the control pulse.

After that, at constant threshold power P ₀ and pulse duration of the control light 18, the magnitude of the applied voltage sequentially increased stepwise, and the magnitude of the voltage change V ₀ at each step was determined

by a change in the frequency (period) of the generated oscillations per octave (a change in frequency by one octave corresponds to its change by 2 times). The final value of V _{0 was} determined by the generation stall voltage V _c , reaching which the bias voltage also decreased stepwise to its threshold value V _p (x). In this case, for each voltage value at which the lasing frequency changed by an octave, all the characteristics of the recorded photocurrent signal were measured, which subsequently calculated the magnitude of the NDC and, accordingly, the dependence of the frequency of the generated oscillations on the bias voltage. On Fig (I) is presented in a double logarithmic scale a graph of the average time-time current-voltage characteristics of the diode structure under test, containing section AB, which shows the change in the frequency of the generated oscillations when a rectangular light 18 of constant power illuminates with a pulse and the bias voltage V ₀ from V _p (x) to V _c . Each value of the magnitude of the NDC in the plot AB is fixed when the frequency of the generating oscillations is changed by one octave when the value of the applied voltage V ₀ changes in the range from 545 V to 1500 V, when the diode structure is illuminated by a 20 μs light pulse with a power of P ₀ ≈6 5 mW. The test showed that for this structure, the stall voltage was V _c ≥1500 V, and a change in the bias voltage from 545 V to 1500 V and vice versa provided a linear and reversible change in the NDC and the frequency of the generated oscillations in the range of 11 octaves.

Testing the dependence of the frequency of the generated oscillations on the pulse power of the control light 18 was carried out similarly. Those. at a constant value of V ₀ = V _p , (x) the power of the control optical signal increased step by step from its threshold value P ₀ to the power at which the generation of P _s failed. After that, the power of the control pulse decreased stepwise to its threshold value. Moreover, as in the case of a change in the magnitude of the applied voltage, the magnitude of a step-wise change in the power of the control optical pulse 18 was determined by changing the frequency of the generated pulses per octave. In this case, for each value of the power of the optical pulse, which provides an increase or decrease in the frequency of the generated oscillations per octave, all the characteristics of the recorded photocurrent signal were measured, which subsequently also calculated the magnitude of the NDC and, accordingly, the dependence of the frequency of the generated oscillations on the power of the control optical signal. On Fig (II) is presented in

double logarithmic scale graph of the average time-time current-voltage characteristics of the same diode structure containing the AB section, which shows the change in the frequency of the generated oscillations at a fixed value of V ₀ = V _p , (x) and the change in the power of the control optical signal 18 from its threshold value of P ₀ to the power at which there was a breakdown of the generation of P _c . Each value of the magnitude of the NDC in the plot AB is fixed when the frequency of the generating oscillations is changed by one octave at a fixed value of V ₀ = V _p , (x) and the power of the control optical pulse P changes in the range from 6.5 mW to 42 mW. Testing of this mode showed that for a given diode structure, a change in the power P of the control optical pulse 18 in the range from 6.5 mW to 42 mW provided a linear and reversible change in the NDC and the frequency of the generated oscillations in the range of 13 octaves.

Similarly, we tested the dependence of the NDC and the frequency of the generated oscillations with jointly changing values of the bias voltage and the pulse power of the control light 15. In this case, the generation failure occurred when the bias voltage V _c = 890 V and the pulse power of the control light R _c = 19 mW. Tests of this mode showed that a joint linear change in the magnitude of the applied voltage and pulse power of the control light from their threshold values to the values at which the generation fails, also leads to a linear and reversible change in the magnitude of the NDC and the frequency (period) of the generated oscillations in the range of 11 octaves .

The control of the threshold generation voltage (threshold generation field) was implemented on a structure with linear dimensions l × h × d = 7 × 7 × 2.5 mm.

During the tests, the threshold voltage value, power and duration of a rectangular pulse of control light 18, at which a single pulse is generated, were chosen experimentally (when testing the diode structure, these values were: V _p (x) = 75 V, P ₀₁ = 41 mW, τ _{imp. ill.} ~ 22 μs). Tests of this generation mode showed that a stepwise increase in the threshold generation voltage (step 75 V) leads to the generation of a single domain at a lower power of the control optical pulse (the step of decreasing the power P is 5.5 mW). In the test sample, generation of a single domain with a maximum threshold voltage of generation and, accordingly, a minimum power of the control pulse P amounted to V _p (x) = 550 V, P ₀₂ = 8 mW, i.e. between the value

the threshold generation voltage and the power of the control optical pulse, an inversely proportional dependence is realized.

The aperiodic generation mode (see Fig. 14) was implemented on the same diode structure, i.e. a structure with linear dimensions l × h × d = 7 × 7 × 2.5 mm. During the tests, the value of the threshold voltage, power, and duration of the rectangular pulse of the control light 18 at which a single pulse was generated were the same, i.e. V _p (x) = 75 V, P ₀ = 41 mW, τ _{imp. con.} ~ 22 μs. After that, at a fixed value of V _p (x) and τ _{imp. The} control optical signal 18 was linearly amplified, which led to a sequential increase in the number of generated pulses per photocurrent pulse. In this case, the amplitude and period of oscillation of each subsequent pulse sequentially increased until, as a result of the increase in the power of the control light pulse, the generation of the test diode structure did not stall, upon occurrence of which the pulse power decreased to its threshold value.

Testing of this mode showed that in a diode structure with a minimum constant value of the threshold bias voltage V _p (x) = 75 V, a change in the power P of the control pulse in the range from 41 mW to 71 mW leads to an aperiodic generation mode in which the period of generated oscillations T is linear decreases (increases), and the number of generated oscillations is limited by the output stall power P _s . In addition, in this generation mode, the maximum power of the generated signal was achieved, the peak electric power of which was P _imp. ≈27.4 watts

Thus, in comparison with the prototype of the claimed utility model for the first time provides microwave generation of photocurrent pulses only due to the impact of lighting. In addition, the inventive device allows you to optically and electrically control the parameters of the generated oscillations at various generation modes, increase the frequency of the generated oscillations by five orders of magnitude in comparison with the prototype, and also provide microwave generation of photocurrent pulses when illuminated by complex light pulses, and optical pulses are generated occurs when the device is operating at room temperature. This became possible due to the fact that conditions were created under which only due to the influence of illumination in the optical generator the formation of negative differential conductivity of variable magnitude and the formation of a volume photocurrent domain

instabilities of the drift type, as well as the formation by illumination of a nonuniform profile of the distribution of the electric field, which remains inhomogeneous in the ν region of the structure outside the domain, when a domain arises and is transferred through the structure.

It should be noted that a similar diode structure can be made on the basis of other high-resistance semiconductors, for example, CdZnTe, CdTe, InP, etc. The basis for this statement is the generality of their electrophysical properties, in particular, single crystals of such semiconductors have a similar band structure. In addition, single crystals of such semiconductors can be grown in high resistance, containing a minimum concentration of electrically active impurities, which allows them to provide the maximum ratio σ _f / σ _t necessary for the formation of volumetric photocurrent domain instability of a drift type.

Claims (2)

1. An optical generator containing a diode structure comprising an ν layer and electrodes made on opposite faces of the ν layer, as well as a bias voltage source electrically connected to the electrodes of the diode structure and a control light source, is optically coupled to one of its faces, according to utility model, said structure additionally contains two nanoscale monolayer dielectrics located symmetrically with respect to the ν-layer between electrodes made in the form of optically transparent electrodes on the surface which are coated with antireflection and contain two dielectric guard rings with respect to the ν layer, each of which encloses a nanoscale dielectric monolayer, an optically transparent electrode and an antireflection coating along the perimeter, and the control light source is optically coupled to one of the antireflection coatings, while the ν layer is made based on a high-resistance direct-gap semiconductor crystal having intrinsic conductivity, for which the product of the equilibrium carrier concentration by the length of the sample satisfies the condition: n ₀ L <10 ¹² cm ^-2 . 2. The optical generator according to claim 1, characterized in that it further comprises a second control light source parallel to the first and similar to it, the radiation axis of which is perpendicular to the plane of the second antireflection coating.