WO2015047125A1

WO2015047125A1 - Photoconverter with quantum dots

Info

Publication number: WO2015047125A1
Application number: PCT/RU2013/000841
Authority: WO
Inventors: Алексей Михайлович НАДТОЧИЙ; Михаил Викторович МАКСИМОВ; Алексей Евгеньевич ЖУКОВ; Николай Александрович КАЛЮЖНЫЙ; Владимир Михайлович ЛАНТРАТОВ; Сергей Александрович МИНТАИРОВ
Original assignee: Общество с ограниченной ответственностью "Солар Дотс"
Priority date: 2013-09-26
Filing date: 2013-09-26
Publication date: 2015-04-02
Also published as: RU2016115439A; RU2670362C2

Abstract

The present photoconverter with quantum dots consists of a substrate (1), for example Ge or GaAs, and at least one photoactive p-n junction (2), for example, of GaAs or GaInAs with an indium concentration of 0-2%, said p-n junction comprising a base layer (3), for example, of GaAs or GaInAs with an indium concentration of 0-2%, an undoped layer (4), for example, of GaAs or GaInAs with an indium concentration of 0-2%, containing at least one layer of self-organized quantum dots (5), which are formed by deposition of a layer of lnxGa1-xAs with an indium content of from 20 to 50%, and an emitter layer (6), for example, of GaAs or GaInAs with an indium concentration of 0-2%.

Description

PHOTO TRANSDUCER WITH QUANTUM DOTS

The invention relates to semiconductor photoconverters (solar cells) that convert solar radiation to electricity, and can be used in the semiconductor industry to create systems for generating electrical energy.

The conversion of light energy into electricity using semiconductor photoconverters with an rn junction is based on the creation of electron-hole pairs upon absorption of photons with an energy exceeding the band gap of the material of the photoconverter, and separation of different-pole carriers by a pulling field of the rn junction. A significant increase in the efficiency of such photoconverters is possible only when using structures of multi-junction (cascade) photoconverters, of which the most promising, both in terms of the possibility of achieving the highest values of efficiency and from an economic point of view, are monolithic heterostructure photoconverters based on A ³ B ⁵ solid solutions obtained by epitaxial growth on a semiconductor substrate in a single growth process. Such photoconverters consist of several subcells, including a photoactive pn junction, made of various materials and arranged in decreasing order of the band gap from the photosensitive surface to the substrate. To ensure effective low-impedance isolation of sub-elements of monolithic cascade photoconverters, the use of tunnel diodes is necessary.

Each photoactive rn junction of the cascade structure converts only a part of the solar spectrum, which allows realizing transformation conditions close to optimal and significantly increasing the efficiency. At the same time, the subcells that convert short-wave radiation are characterized by a high open-circuit voltage, since they are made of materials with a larger band gap, and the possibility of using narrow-gap materials can significantly expand the photosensitivity region of cascade photoconverters.

The development of cascade photoconverters that convert concentrated radiation is one of the most promising ways to achieve solar energy parity with traditional sources. Their efficiency exceeds the efficiency of silicon photoconverters by 2-3 times, and the cost can be significantly reduced by using cheap lenses that concentrate solar radiation on small chips. At this, the need for expensive heterostructures of cascade photoconverters, and, consequently, the cost of energy decreases in proportion to the concentration factor, which in modern solar photovoltaic plants reaches 500 - 1000 suns.

The main problem of the most promising cascade photoconverters suitable for industrial production today, based on the GalnP / GalnAs / Ge materials matched by the lattice parameter, is the inconsistency of the currents generated by the subcells of such a structure. This leads to the fact that their efficiency is about 39%, with a theoretical limit of more than 50%.

The inconsistency of the currents is due to the low current of the middle subcell based on GalnAs, which limits the current of the entire structure, therefore, expanding the spectral range of its photosensitivity, which entails an increase in the current generated by it, is the most important task for realizing the efficiency potential of cascade photoconverters.

A known photoconverter with quantum dots (see US patent 6,507,042 B1, issued January 14, 2003), including a semiconductor substrate made of at least 3 elements, and quantum dots that are formed on a semiconductor substrate in such a way that the radiation wavelength can be determined using the lattice parameter of the semiconductor substrate.

The disadvantage of the known photoconverter with quantum dots is the high cost, because the creation of substrates from three or more component solid solutions requires a significant increase in the cost of technology. In addition, low absorption at standard quantum dots can lead to a small increase in photocurrent.

A known photoconverter with quantum dots (see patent CN 202111103U, issued January 11, 2012), comprising at least one layer of quantum dots made by means of elastic stress relaxation, comprising a substrate, a buffer layer, an η-type region, i is an inner region , and a p-type region, with quantum dots made of GaNAs.

The disadvantages of the known photoconverter with quantum dots are the low absorption and complexity of the technology for producing nitrogen-containing layers.

The closest to this technical solution for the combination of essential features is a photoconverter with quantum dots (see US patent 7,863,516, issued January 4, 2011), adopted as a prototype and including a Ge or GaAs substrate, a set of photoactive transitions, one of which includes layers of self-organized InGaAs quantum dots separated by spacer layers of GaAs, AIGaAs or GaPAs, while the effective width The band gap of this transition is 1.16 eV.

Arrays of quantum dots play an important role in the structure of the prototype photoconverter, which expand the photosensitivity of the sub-element to the long-wavelength region.

The disadvantage of the known photoconverter prototype is the low absorption in arrays of standard quantum dots, which makes it impossible to ensure current matching for a cascade photoconverter based on GalnP / GalnAs / Ge.

The objective of the proposed solution is to create quantum dots with increased absorption, formed by the deposition of an InGaAs layer with an indium concentration of 20-50% on the surface of GaAs or InGaAs. This will increase the efficiency of the technical solution by increasing the current generated by the photoactive transition based on Ga (ln) As. An increase in the photogenerated current, as well as the efficiency of the two-stage GalnP / GaAs based photoconverter, is achieved by spreading the spectral sensitivity of the Ga (ln) As transition to the long-wavelength region due to the introduction of quantum dot arrays into it. In addition, the use of the Ga (ln) As junction with such quantum dots opens up the possibility of matching currents in the most promising three-stage structure based on GalnP / GalnAs / Ge, which will significantly increase the efficiency of such a photoconverter.

This object is achieved in that the photoconverter with quantum dots, contains a substrate of Ge or GaAs and at least one photoactive transition made of GaAs or a solid solution of GalnAs with an indium content of 0 - 2%, and includes at least one a layer of self-organized quantum dots formed by deposition of a ln _x Ga _1-x As layer with an indium x content of 20 to 50%.

In a photoconverter, a photoactive transition made of GaAs or a GalnAs solid solution with an indium content of 0 - 2% can include more than 10 layers of quantum dots separated by spacers made, for example, of GaAs or a GalnAs solid solution with an indium content of 0 - 2%. An n-GaAs substrate can be used as a substrate in the photoconverter, onto which a buffer layer can be sequentially deposited, for example, from n-GaAs with a thickness of, for example, 200-300 nm with a doping level of (1-2) 10 ¹⁸ cm ^"3 and a photoactive transition, including a successively deposited layer of the back potential barrier, for example, of n-AIGaAs with an aluminum content of 30%, a thickness of, for example, 100 nm and a doping level of (1-2) 10 ¹⁸ cm ^'3 , a base layer, for example, of n-GaAs with a thickness of, for example, 3-3.5 μm and a doping level of (5-9) 10 ¹⁶ cm ^"3 , undoped layer, for example measures, from GaAs, which can include more than 10 layers of quantum dots separated by spacers made, for example, of GaAs, an emitter layer, for example, of p-GaAs with a thickness of, for example, 300-500 nm and a doping level (2-5 ) 0 ¹⁸ cm ^"3 , a wide-gap window layer, for example, of p-AIGaAs with an aluminum content of 85%, a thickness of, for example, 30 nm and a doping level of (1-2) 10 ¹⁸ cm ^{" 3} , as well as a contact sublayer, for example , of p-GaAs with a thickness of 300 nm and a doping level of (5–9) 10 ¹⁸ cm ^"3 .

In the photoconverter, a p-GaAs substrate can be used as a substrate on which a buffer layer can be sequentially deposited, for example, from p-GaAs with a thickness of, for example, 200-300 nm with a doping level of (1-2) 10 ¹⁸ cm ^"3 , and a photoactive transition, including a successively deposited layer of the back potential barrier, for example, of p-AIGaAs with an aluminum content of 30%, a thickness of, for example, 100 nm and a doping level of (1-2) 10 ¹⁸ cm ^"3 , the base layer, for example, of p-GaAs with a thickness of, for example, 3-3.5 μm and a doping level of (1-2) 10 ¹⁷ cm ^"3 , unalloyed layer, for example p, from GaAs, which can include more than 10 layers of quantum dots separated by spacers made, for example, of GaAs, an emitter layer, for example, of n-GaAs with a thickness of, for example, 100-200 nm and a doping level (2-5 ) Yu ¹⁸ cm ^"3 , a wide-gap window layer, for example, of n-AIGaAs with an aluminum content of 85%, a thickness of, for example, 30 nm and a doping level of (1-2) 10 ¹⁸ cm ^{" 3} , as well as a contact sublayer, for example , of n-GaAs with a thickness of, for example, 300 nm and a doping level of (2-5) 10 ¹⁸ cm ^"3 .

In the photoconverter, a p-GaAs substrate can be used as a substrate, onto which a buffer layer, for example, from p-GaAs, a lower photoactive transition, including a successively deposited layer of the back potential barrier, for example, from p-AIGaAs or GalnP, can be sequentially deposited, base layer, e.g. from p-GaAs, unalloyed a layer, for example, of GaAs, which may include more than 10 layers of quantum dots separated by spacers made, for example, of GaAs, an emitter layer, for example of l-GaAs, and a layer of a wide-gap window, for example, of n-AIGaAs or a two-layer a wide-gap window including a layer, for example, of GalnP and a layer, for example, of AllnP, a tunnel diode, which may contain successively deposited layers / i ⁺⁺ -GaAs or n ⁺⁺ -GalnP and a p ⁺⁺ -AIGaAs layer, the upper photoactive transition comprising a sequentially deposited layer of the back potential barrier, for example, from p-AIGalnP, b zovy layer, for example, p-GalnP, an emitter layer, e.g., of p-GalnP and layer of a wide window, for example, n-AIpR as well as a contact sublayer, e.g., of n-GaAs.

In the photoconverter, a p-Ge substrate can be used as a substrate, onto which a nucleation layer can be sequentially deposited, for example, from GalnP, which creates a lower photoactive transition in the germanium substrate due to diffusion of phosphorus atoms, a buffer layer, for example, from GalnAs, with the indium content of 0-2%, the lower tunnel diode, which may include sequentially osazhennye wide-barrier layers, n ⁺⁺ -layer p ⁺⁺ -layer, average photoactive transition layer comprising a rear potential barrier nap an example, from p-AIGaAs or GalnP, a base layer, for example, from p-GalnAs with an indium content of 0-2%, an undoped layer, for example, from GalnAs with an indium content of 0-2%, which may include more than 10 layers of quantum dots, separated by spacers made, for example, of GalnAs with an indium content of 0-2%, an emitter layer, for example of l-GalnAs with an indium content of 0-2%, and a layer of a wide-gap window, for example, of l-AIGaAs or a two-layer wide-gap a window, which may include a layer, for example, of GalnP and a layer, for example, of AllnP, the upper tunnel diode, which may contain after ovatelno deposited layers n ⁺⁺ -GaAs or n ⁺⁺ -GalnP and a layer of p ⁺⁺ -AIGaAs, upper photoactive transition comprising sequentially osazhennye rear potential barrier layer, for example, p-AIGalnP, a base layer, e.g., of the p-GalnP , an emitter layer, for example, of l-GalnP and a wide-gap window layer, for example, of l-ΑΙΙηΡ, as well as a contact sublayer, for example, of l-GalnAs with an indium content of 0-2%.

An important feature of the present invention is the ability to provide increased absorption in quantum dot arrays, which allows a significant increase in the photocurrent of the subcell from Ga (ln) As.

This technical solution is illustrated in the drawing, where: in FIG. 1 schematically shows a real photoconverter with quantum dots;

in FIG. Figure 2 shows typical spectral characteristics of the external quantum efficiency Ga (ln) As of the GalnP / GalnAs / Ge subcell of a cascade photoconverter without quantum-well heterostructures (curve 1), with quantum wells (curve 2), and with quantum dots (curve 3).

in FIG. Figure 3 shows the photocurrent generated by 10 rows of quantum dots obtained upon relaxation of an InGaAs quantum well with various indium concentrations from 20 to 100%).

The present quantum dot photoconverter is shown in FIG. 1. It consists of a substrate 1, for example, Ge or GaAs and at least one photoactive pn junction 2, for example, of GaAs or GalnAs with an indium concentration of 0-2%, containing a base layer 3, for example, of GaAs or GalnAs with an indium concentration of 0-2%, unalloyed layer 4, for example, of GaAs or GalnAs with an indium concentration of 0-2%, containing at least one layer of self-organized quantum dots 5 made by deposition of the ln _x Gai layer. _x As with an indium content of x from 20 to 50%, the emitter layer 6, for example, from GaAs or GalnAs with an indium concentration of 0-2%.

To obtain semiconductor heterostructures of cascade photoconverters with high quality, which is necessary to create efficient pn junctions, it is necessary to rely on materials existing in nature to ensure the matching of the lattice parameters of all layers that make up the heterostructure. That is why the most promising today are three-junction photoconverters based on strictly isoperiodic semiconductor materials Gao.51 lno.49 Gao.99lno.01As / Ge. However, the materials are Ga ₀ . ₅ iln ₀ .49 (Eg = 1.9 eV),

(Eg = 1.4 eV) and Ge (Eg = 0.66 eV) do not allow the implementation of an optimal three-junction photoconverter design from the point of view of the spectral density of photons per each subcell. But due to the high stability of the parameters, even with prolonged operation and reproducibility of industrial technology, this structure is currently the main industrial trend, and has an efficiency in the production of about 39%.

The main disadvantage of the combination of materials Gao.51 lno.49 Ga ₀ gglno.oiAs - Ge is the large band gap of the middle Gao.99lno.01As subelement. For both the cosmic and the terrestrial solar spectra in the case of absorption by each subelement of all photons with with the energy of the greater band gap of their material and the separation of all photogenerated carriers, the photocurrents of subcells will be:

- 22.43 mA / cm ² (AMO) and 18.1 1 mA / cm ² (AM1.5D) for the Ga ₀ .5iln ₀ .49P transition (all photons from 0 to 670 nm);

- 16.58 mA / cm ² (AMO) and 15.62 mA / cm ² (AM1.5D) for Ga ₀ , 99ln _0, oiAs transition

(all photons from 670 to 900 nm);

- 37.08 mA / cm ² (AMO) and 29.21 mA / cm ² (AM1.5D) for the Ge transition (all photons from 900 to 1900 nm).

This leads to the fact that, for both the AMO spectrum and the AM1.5D spectrum, the minimum current (in the case of absorption of all photons with energies in the range 1, 9 -1, 4 eV) will generate an average subelement, and the current generated lower will significantly exceed the currents of the upper and middle subelements.

Thus, in a GalnP / GalnAs / Ge photoconverter, in the first GalnP-based sub cell, -25% of the light flux is absorbed, in the second (GalnAs) ~ 15%, and in the third (Ge) ~ 40% (the structure is transparent for 20% of photons). In other words, the wide spectral characteristic of the Ge subelement causes the excessive generation of nonequilibrium carriers. Since the total working photocurrent in cascade photoconverters is limited by the smallest of the generated subcells, it is advantageous to expand the spectral sensitivity range of such a sub cell to increase the overall efficiency.

It is important to note that in real cascade photoconverters based on GalnP / GalnAs / Ge, the photocurrents of subcells are of the order of:

- 20 mA / cm ² (AMO) and 18 mA / cm ² (AM1.5D) for Ga ₀ . ₅ ln ₀ .49P transition;

- 17 mA / cm ² (AMO) and 14 mA / cm ² (AM1.5D) for Ga ₀ , 99lno _, oiAs transition;

- 25 mA / cm ² (AMO) and 22 mA / cm ² (AM1.5D) for the Ge junction.

One of the ways to achieve the current balance and, accordingly, increase the efficiency of the traditional structure of cascade photoconverters is to use metamorphic (i.e., strongly mismatched in lattice parameter) structures of the wide-gap GalnP / GalnAs tandem with the indium content in solid solutions 10-20% higher than the matched by the lattice parameter of the structure. In the case of metamorphic photoconverters, the wide-band tandem approaches the optimal distribution of spectral density transformed by each sub-element (current matching mode), but it requires the III-V structure to be grown on Ge through buffer layers with variable composition and significant a change in the lattice parameter, which entails the presence of a large number of defects and misfit dislocations and levels the increase in efficiency.

Another way to increase the current of the middle subcell of the GalnP / GalnAs / Ge photoconverter is to use quantum-well heterostructures with quantum wells or quantum dots. However, both of these approaches have limitations that do not allow matching the currents of the GalnP / GalnAs / Ge subcells of the photoconverter.

In the case of quantum wells, this limitation is elastic stresses. This is due to the fact that the creation of a material with a band gap of less than GaAs is possible only with a change in the lattice parameter. The most typical is the use of InGaAs quantum wells, which have a lower absorption energy in comparison with Ga (ln) As. In view of the fact that quantum wells are stressed structures, i.e. the elastic stresses of the crystal lattice remain after their growth; there is a limitation on the shift of the absorption edge of the Ga (ln) As-based subcell (not more than 950–980 nm). An increase in the indium concentration or the thickness of the quantum wells, which leads to a decrease in the energy of the absorption edge and, accordingly, to a long-wavelength shift of the photosensitivity edge of the Ga (ln) As subelement, will increase the elastic stresses, which will ultimately lead to degradation of the subelement parameters due to the formation of misfit dislocations. In addition, very high stresses in structures with quantum wells do not allow storing a large number of quantum wells, i.e. increase absorption by increasing their number. In the case of quantum wells emitting in the wavelength range of 980 nm, storage of no more than three wells is possible without the formation of dislocations. Storing four or more quantum wells requires the use of voltage compensating layers or very wide spacers. However, the formation of stress compensating layers is a complex technique and leads to the introduction of additional interfaces into the structure that adversely affect the characteristics of the device, while the use of too wide spacers would reduce the efficiency of carrier collection.

However, the advantages of quantum wells in comparison with quantum dots are that they have a relatively large absorption coefficient, which can provide a level of external quantum efficiency of the photoconverter in the absorption range of more than 30%. Thus, when using quantum wells, it is possible to obtain a high level of quantum efficiency, however, the shift of the absorption edge during their use remains small. Due to the fact that the photocurrent is defined as the integral under the curve of the external quantum efficiency of the photoconverter (Fig. 2), the total photocurrent gain when using quantum wells (integral in the region> 900 nm) will be small, which will not allow matching the currents of the GalnP / GalnAs / Ge subcells of the photoconverter .

The use of quantum dots solves the problem of elastic stresses. Quantum dots, as a rule, are created by the method of self-organization during relaxation of a highly stressed InAs or InGaAs layer with an indium concentration of 50% or more (Stransky-Krastanov growth mode and its modifications). The relaxation of elastic stresses leads to the appearance of pyramidal islands, called quantum dots, located on a thin wetting layer that completely covers the surface. Thus, the use of quantum dots allows one to obtain a relaxed defect-free medium that absorbs in the wavelength range up to 1100 nm and beyond.

However, the disadvantage of quantum dots is the low absorption of photons in them. This is due to the fact that quantum dots cover only 10-20% of the surface. In addition, the density of states of quantum dots is a set of delta functions, which also reduces the absorption in them compared to quantum wells, and limits the level of external quantum efficiency of the photoconverter in the absorption region of a medium with quantum dots at a level of less than 10%

Thus, when using quantum dots grown by the Stransky-Krastanov method, there is the possibility of a significant shift of the absorption edge, however, it is impossible to obtain a high level of quantum efficiency (Fig. 2). In this case, the total increase in the photocurrent when using quantum dots will also be small, which will not allow matching the currents of the GalnP / GalnAs / Ge subcells of the photoconverter.

The essence of the invention is an original method of forming quantum dots during the deposition of InGaAs with a relatively small composition of 20-50% on the GaAs surface.

In the present invention, it is proposed to incorporate into the Ga (ln) As sub-element an absorbing medium based on quantum dots formed by the deposition of 20-50% InGaAs onto the GaAs surface and providing effective stress relaxation inside the InGaAs quantum well.

The application of this technology made it possible to create GaAs-based photoelectric converters with an absorption edge up to 1100 nm and a quantum efficiency level of more than 30%, i.e. high absorption (as in the case of quantum wells) was demonstrated along with the long-wavelength edge of the absorption spectrum (as in quantum dots). The total increase in the photocurrent due to the introduction of quantum dots amounted to more than 3 mA / cm ² , which makes it possible to match the subcell currents of the cascade GalnP / GalnAs / Ge photoconverter for the AM1.5D ground-based spectrum.

GaAs based photoconverters with quantum dots were obtained by relaxing the InGaAs quantum well layer with an indium concentration of 20 to 100%. The maximum increase in photocurrent was observed at an indium concentration of 20 to 50% (Fig. 3).

In the case of a smaller composition according to India, the well relaxation will occur at a very large thickness due to the small difference in the lattice parameters of the quantum well and GaAs, which will not allow one to obtain dislocationless quantum dots. In this case, the shift of the absorption edge will be small due to the small InGaAs band gap at an indium concentration of less than 20%.

An increase in the indium concentration of more than 50% led to a decrease in the absorption coefficient of the medium with quantum dots and a decrease in the photocurrent from them (Fig. 3).

YU

Claims

CLAIM

1. Photoconverter with quantum dots, containing a substrate of Ge or GaAs and at least one photoactive transition made of GaAs or a solid solution of GalnAs with an indium content of 0 - 2%, and includes at least one layer of self-organized quantum dots formed by depositing a ln _x Ga _1-x As layer with an indium x content of 20 to 50%.

2. The photoconverter according to claim 1, characterized in that the photoactive transition made of GaAs or GalnAs solid solution with indium content of 0 - 2% includes more than 10 layers of quantum dots separated by spacers made of GaAs or GalnAs solid solution with the indium content is 0 - 2%.

3. A photoconverter according to claim 1, characterized in that a buffer layer of n-GaAs 200-300 nm thick with a doping level of (1-2) 10 ¹⁸ cm ^"3 , and a photoactive transition, consistently deposited on a p-GaAs substrate are sequentially deposited a deposited back-potential barrier layer of p-AIGaAs with an aluminum content of 30%, a thickness of 100 nm and a doping level of (1-2) 10 ¹⁸ cm ^"3 , a base layer of n-GaAs with a thickness of 3-3.5 μm and a doping level (5 -9) 10 ¹⁶ cm ^"3 , unalloyed GaAs layer, including more than 10 layers of quantum dots separated by spacers made of GaAs, 300-500 nm thick p-GaAs emitter layer with a doping level of (2-5) 10 ¹⁸ cm ^"3 , p-AIGaAs wide-gap window layer with 85% aluminum content, 30 nm thickness and a doping level of (1-2) 10 ¹⁸ cm ^"3 , as well as a contact sublayer of p-GaAs 300 nm thick and a doping level of (5-9) 10 ¹⁸ cm ^{" 3} .

4. The photoconverter according to claim 1, characterized in that a buffer layer of p-GaAs 200-300 nm thick with a doping level of (1-2) 10 ¹⁸ cm ^"3 , and a photoactive transition including deposited p-AIGaAs back potential barrier layer with an aluminum content of 30%, a thickness of 100 nm and a doping level of (1 -2) 10 ¹⁸ cm ^"3 , a base layer of p-GaAs with a thickness of 3-3.5 μm and a doping level (1 -2) 10 ¹⁷ cm ^"3 , unalloyed GaAs layer, including more than 10 layers of quantum dots separated by spacers made of GaAs , an emitter layer of n-GaAs with a thickness of 100-200 nm and a doping level of (2-5) 10 ¹⁸ cm ^"3 , a wide-gap window layer of n-AIGaAs with an aluminum content of 85%, a thickness of 30 nm and a doping level of (1-2) 10 ¹⁸ cm ^"3 , as well as a contact sublayer of n-GaAs with a thickness of 300 nm and a doping level of (2-5) 10 ¹⁸ cm ^{" 3} .

5. The photoconverter according to claim 1, characterized in that a p-GaAs buffer layer, a lower photoactive transition including a sequentially deposited back-potential barrier layer of p-AIGaAs or GalnP, and a base layer of p-GaAs are successively deposited on the p-GaAs substrate an undoped GaAs layer comprising more than 10 layers of quantum dots separated by spacers made of GaAs, an emitter layer of l-GaAs and a layer of a wide-gap window of l-AIGaAs, or a two-layer wide-gap window, comprising a layer of GalnP and an A&R layer, tunnel diode containing been consistent deposited layers n ⁺⁺ -GaAs or n ⁺⁺ -GalnP and a layer of p ⁺⁺ -AIGaAs, upper photoactive transition layer comprising sequentially osazhennye rear of the potential barrier of the p- AIGalnP, the base layer of the p-GalnP, an emitter layer of n GalnP and a wide-gap window layer of n-лηΡ as well as a contact sublayer of n-GaAs.

6. The photoconverter according to claim 1, characterized in that a GalnP nucleation layer is successively deposited onto the p-Ge substrate, creating a lower photoactive transition in the germanium substrate due to diffusion of phosphorus atoms, a buffer layer of l-GalnAs, with an indium content of 0-2 %, buffer layer, lower tunneling diode, including successively deposited wide-gap barrier layers, L ⁺⁺ p ⁺ * layer, middle photoactive transition, including back potential barrier layer from p-AIGaAs or GalnP, base layer from p-GalnAs with content of indium 0-2%, unalloyed layer of GalnAs with an indium content of 0-2%, including more than 10 layers of quantum dots separated by spacers made of GalnAs with an indium content of 0-2%, an emitter layer of p-GalnAs with an indium content of 0-2%, and a wide-gap layer window of l-AIGaAs or wideband bilayer window, comprising a layer and a layer of GalnP ANpR upper tunnel diode comprising sequentially deposited layers of n ⁺⁺ -GaAs or n ⁺⁺ -GalnP and a layer of p ⁺⁺ -AIGaAs, upper photoactive junction, including sequentially deposited layer of the back potential barrier of p-AIGalnP, the base layer of p-GalnP, um tterny layer of p-GalnP and layer of a wide window of L-ΑΙΙηΡ as well as a contact sublayer of a p-GalnAs indium content of 0-2%.