WO2015100480A1

WO2015100480A1 - Method for forming thin cigs films for solar cells and device for the implementation thereof

Info

Publication number: WO2015100480A1
Application number: PCT/BY2013/000013
Authority: WO
Inventors: Владимир Яковлевич ШИРИПОВ; Сергей Павлович МАРЫШЕВ; Сергей Михайлович НАСТОЧКИН; Евгений Александрович ХОХЛОВ
Original assignee: Владимир Яковлевич ШИРИПОВ
Priority date: 2013-12-30
Filing date: 2013-12-30
Publication date: 2015-07-09

Abstract

The invention relates to a technique for forming thin semiconducting CIGS films for solar cells on sheet glass substrates, and to vacuum deposition devices for implementing said technique under industrial mass production conditions. The present invention solves the problem of creating a method for forming thin-film CIGS layers for solar cells on large-scale substrates which makes it possible, under mass production conditions, to obtain an optimum integral composition of the material, to ensure the efficient conversion of solar energy into electricity, and to provide a technique which can be reliably reproduced in mass production on large substrates, while at the same time reducing production costs. This problem is solved in that in a known method for forming thin CIGS films, material is applied in successive layers by reactive sputtering in elemental selenium vapour using consecutively arranged magnetron sputtering stations, wherein for the cathodes of the magnetron stations in the positions for the formation of odd-numbered layers, an alloy of Cu-In-Ga is used in which the atomic concentration ratios of Cu/(In+Ga) [Cu/III] and Ga/(In+Ga) [Ga/III] are selected in the ranges of [Cu/III] = 0.47 - 0.51 and [Ga/III] = 0.25 - 0.3, and for the cathodes of the magnetron stations in the positions for the formation of even-numbered layers, an alloy of Cu-Ga is used in which the atomic concentration ratios of Cu/Ga = 2.5 - 2.8. The method and the device also differ in other ways from the prior art.

Description

METHOD FOR FORMING CIGS THIN FILMS FOR SUNNY BATTERIES AND DEVICE FOR ITS

IMPLEMENTATIONS

The invention relates to a technology for the formation of thin semiconductor CIGS films for solar cells on sheet glass substrates and to devices for vacuum deposition that implement such a technology in mass industrial production.

A known method of recrystallization of CIGS films for solar cells [1], including vacuum deposition of ultimately, a slightly copper-depleted layer of Cu (In, Ga) Se ₂ (CIGS) on a substrate, including the formation on the surface of the substrate of the initial layer enriched with copper in the form a mixture of phases Cu (In, Ga) Se ₂ - Cu _x Se and subsequent application of a mixture of Cu (In, Ga) Se ₂ - Cu _x Se to the surface of the initial layer at an excess vapor pressure of Se and (In, Ga) and at the same time increasing the temperature of the substrate.

The specified method does not allow to obtain thin film structures homogeneously stable in thickness, and this, in turn, does not allow to obtain a structure with a high coefficient of transformation of solar energy into electricity. In addition, the method is extremely difficult to use with respect to large substrates.

There is also known a method of vacuum deposition, ultimately, a slightly copper-depleted layer of Cu (In _x Ga _1-x ) Se ₂ [2, 3], which includes three successive stages. At the first stage, a (InGa) ₂ Se ₃ layer is deposited by evaporation from individual sources of In, Ga, and Se on a glass substrate coated with a molybdenum contact layer heated in a temperature range of 250–400 ° C. The second and third stages are carried out at temperatures of about 550 ° C. At the same time, in the second stage, Cu and Se are applied by evaporation from individual sources until the moment when the total composition of the layers the first and second stages will be enriched with copper in comparison with the formula Cu (In _x Gai _-x ) Se ₂ . In the third stage, In, Ga and Se are again applied. The moment of the end of the process is chosen such that when the total composition of all the material obtained in all three stages becomes depleted of copper in comparison with the formula Cu (In _x Ga _1-x ) Se ₂ . For the known spraying method, a method for controlling the moments of the end of the second and third stages, described in [4,5], was created. In the known control method, the change in the heater power necessary to maintain a constant substrate temperature in the second and third stages is monitored. The change in the heater power is caused by the fact that, at the end of the second stage, a free Cu _2-x Se phase appears on the surface of the film layer, the thermal emissivity of which is different from the thermal emissivity of the final Cu (In _x Gai _-x ) Se ₂ compound slightly depleted in copper. Thus, a change in thermal emission leads to a change in power dissipated and consumed to maintain a constant temperature, and this, in turn, allows us to estimate the amount of free phase Cu _2-x Se.

The well-known three-stage method of forming a layer of Cu (In, Ga) Se2 slightly depleted in copper is characterized by the following disadvantages, if used in mass industrial production:

Firstly, there are restrictions on the maximum size of a flat glass substrate due to the uniform exposure of the surface of a large glass substrate to all the components of the film — Se, Cu, In, and Ga;

Secondly, the production line for the implementation of this method is limited by the design, in which the film is applied from bottom to top, and the substrate of sheet glass is horizontally receiving surface down. With this design, under conditions of a temperature of about 550 ° C, which is almost equal to the softening temperature of the glass, deformation of the substrate under the influence of gravitational forces or the sum of internal stresses in the deposited layers is inevitable; Thirdly, in the first stage, a (InGa) ₂ Se ₃ layer is applied, the crystal lattice of which is not a chalcopyrite structure, which is necessary for the Cu (In _x Gai _-x ) Se ₂ layer. As a result of the transformation of this type of crystal lattice into the structure of chalcopyrite, which occurs at the second stage of the deposition process, the interface between Mo and Cu (In _x Ga _1-x ) Se ₂ is a source of strong mechanical stresses, which ultimately affects the reliability and longevity of the solar cell.

Fourth, the transformation of the (InGa) ₂ Se ₃ layer formed in the first stage into a Cu (In _x Gai _-x ) Se ₂ layer in the second stage is due to the diffusion of copper condensing on the surface of this layer. As a result, in order to achieve acceptable uniformity of the composition of Cu (In _x Ga _1-x ) Se ₂ , when the ratio of atomic concentrations Cu / (In + Ga) lies in the range of 0.88 - 0.92 at any point in the cross section of the layer, either adjust the diffusion rate of copper to the deposition rate, which is practically not feasible for the industrial method, or significantly reduce the deposition rate of copper in the second stage, which will significantly limit the performance of the process. Since the thickness of the layer deposited in the first stage is close to half of the required thickness of the final layer, then at a rate of vacuum deposition acceptable for industrial production, inside the resulting Cu (In _x Gai _-x ) Se ₂ layer, the gradient of the ratio Cu / (In + Ga), with a minimum at the boundary with the Mo layer. At the same time, the reproducibility of the technology in accordance with the existing method will be extremely low, due to the inability to control the diffusion rate of copper over time during the second stage.

Fifth, in the proposed control methods for the moments of the end of the second and third stages there are no clear criteria. As a result, the end of the second stage can be characterized by the presence of the Cu ₂ phase on the film surface. _x Se of various thicknesses. Moreover, depending on the deposition rate of In, Ga and Se in the third stage, when the rate the transformation of this phase into chalcopyrite Cu (In _x Ga _{) -x} ) Se _{2 is} controlled by the ratio of the deposition rate and the rate of mutual diffusion of the components, results are observed when a conducting phase Cu _2-x Se is present inside the finished layer (the deposition rate exceeded the diffusion rate of the components). If the diffusion and deposition rates are close to equilibrium, and the thickness of the Cu _2-x Se layer at the end of the second stage is too large, then the surface roughness at the end of the whole process is unacceptably large.

Sixth, the maximum conversion efficiency in solar panels of this type is fixed in those cases when, on the surface of the Cu (In _x Gai _-x ) Se ₂ layer, a layer is finally formed in which the ratio Cu / (In + Ga) <0.4 , and its thickness does not exceed 20 - 50 nm. In such a layer, the conductivity type undergoes a transformation from p to n, and, as a result, upon completion of the formation of the Cu (In _x Ga] _iX ) Se ₂ layer, a very thin “built-in” homogeneous pn junction is automatically formed at its surface. This transition is stabilized by a subsequent operation — the deposition of the thinnest layer (~ 50 nm) of CdS from solutions.

The closest technical solution to the claimed method of forming thin copper-depleted films of Cu (In _x Gai _-x ) Se ₂ for solar cells and a device for its implementation in mass production is selected patent US 7,544,884 [6].

According to this technical solution, in conditions of mass production, continuous lines are used in which paired magnetron targets are sprayed with direct or medium frequency current. In this case, the potential on each of the paired targets is set by an independent power supply and, thus, the necessary concentration of components in the film layers is selected. In the technical solution adopted for the prototype, two implementation options are presented.

In the first case, paired magnetron targets are, respectively, copper selenide (CuSe ₂ ) and a mixture of indium and gallium selenides. The working gas in this case is argon.

In the second case, paired magnetron targets are made of copper and indium gallium alloy, respectively. The working gas in this case is a mixture of argon - hydrogen selenide (H ₂ Se).

A device for the formation of a final layer of Cu (In _x Ga _1-x ) Se ₂ with a small copper deficiency is a sequence of several pairs of these magnetron targets in a vacuum corridor of a continuous line. In this case, paired targets can be made in the form of planar cathodes or in the form of rotating cylindrical cathodes. The substrate, in this case, can be made both in the form of a sheet of glass, or in the form of a continuous tape of metal. When the layer is formed, the substrate sequentially passes the positions of paired magnetron targets, assuming the deposited material in successive portions.

The required final composition of the Cu (In _x Gai _-x ) Se ₂ layer is achieved by selecting the ratio of power dissipated in the target materials from copper selenide (pure copper) and indium / gallium selenides (indium / gallium alloy).

Significant advantages of this method over the three-stage method of forming a film of Cu (In _x Gai _-x ) Se ₂ described in [2,3] are:

- removal of restrictions on the size of the substrate, by using linear magnetrons of planar or cylindrical type;

- dividing the deposited film Cu (In _x Gai _-x ) Se ₂ during its growth by the number of layers greater than three. In this case, the distribution of the concentration ratio Cu / (In + Ga) becomes more uniform over the cross section of the final layer (the final layer is divided by the number of layers in accordance with the number of deposition stations, which may be more than three).

However, the disadvantages of the third to fifth of the previous method remain relevant. Moreover, due to the extremely low temperature melting in the In-Ga alloy, the design of the installation, which allows working with a vertical arrangement of the glass substrate, is practically excluded. And with the horizontal location of the target-substrate system, there remain difficulties associated with softening and deformation of the glass. This means that the second drawback from the above list is not eliminated.

In addition, the use of complex composite targets of copper selenides and indium-gallium mixture is practically unacceptable for processes of real mass industrial production in view of the very high cost of their manufacture and the practical unsuitability for regeneration.

It should also be added that controlling the composition of the final layer of Cu (In _x Gai _-x ) Se ₂ by changing the power supplied to different targets in one pair is too crude a method taking into account the drift of parameters, which is caused by a change in the composition of targets containing volatile selenium, and the development of targets in time.

Thus, the known method and device options for its implementation have significant disadvantages, which greatly complicate their use in mass industrial production.

The problem solved by this invention is the creation of a method of forming thin-film layers of CIGS for solar cells on large substrates, which allows you to:

- in conditions of mass production, to obtain the optimal integral composition, characterized by the ratio Cu / (In + Ga) = 0.74 - 0.88, Ga / (In + Ga) = 0.25 - 0.3 and the atomic concentration of selenium at fifty%,

- provide a gradient of the ratio Ga / (In + Ga) with a minimum at the surface and a maximum at the boundary with molybdenum,

- to ensure the ratio Cu / (In + Ga) <0.4 in a thin surface layer (40-60 nm thick)

- ensure the surface roughness of the layers is not more than ± 5% of the thickness LAYER

- to ensure the presence of predominant crystallographic orientation in the directions <112>, <101/103> and <220/204>,

- to ensure the reproducibility of the above parameters in mass production using vacuum flow lines of the vertical type,

- to exclude the use of highly toxic compounds of type H ₂ Se in production processes,

- guarantee the possibility of regeneration of spent targets to reduce production costs.

The problem is solved in that in the known method of forming thin CIGS films for large solar cells by layer-by-layer deposition in a vacuum of a CIGS film on a sheet glass substrate with a layer of conductive molybdenum previously applied to it, in a vacuum corridor of a continuous line, according to the invention, the material is applied in successive layers by reactive spraying in elemental selenium vapor using sequentially arranged magnetron sputtering stations current or twin-magnetron mid-frequency sputtering of flat or cylindrical targets, while the cathodes of the magnetron stations at the positions forming the odd layers use the Cu-In-Ga alloy, in which the ratio of atomic concentration Cu / (In + Ga) [Cu / III ] and Ga / (In + Ga) [Ga / III] are selected in the range [Cu / III] = 0.47 - 0.51 a [Ga / III] = 0.25 - 0.3, and as magnetron cathodes stations at the positions forming even layers use the Cu-Ga alloy, in which the ratio of atomic concentrations Cu / Ga = 2.5 - 2.8, while, at the position forming the first layer, the temperature under ozhki maintained constant in the range of 340 - 400 ° C, and the second and subsequent positions ensure constant substrate temperature in the range of 530 - 80 ° C, a vacuum before removing from the preform is cooled to a temperature of 90 - 110 ° C.

The problem is also solved by the fact that the total number of layers is chosen odd.

The problem is also solved by the fact that the number of layers is selected in the range from 5 to 1 1.

The problem is also solved by the fact that the temperature of the substrate is controlled by a non-contact pyrometer on the glass side.

The problem is also solved by the fact that a constant temperature of the substrate is supported by infrared emitters on the glass side, and as a parameter for the end of the deposition of a particular layer, the magnitude of the change in thermal radiation is used while maintaining a constant temperature of the substrate.

The problem is also solved by the fact that at a constant speed of movement of the workpiece in a vacuum corridor in the zone of application of the layer, the magnitude of the change in thermal radiation to maintain a constant substrate temperature is used as a parameter of the deposition rate of the layer.

The problem is also solved by the fact that at the positions forming even layers, the end of the process is determined after reaching the minimum radiation temperature and subsequently raising it by 0.8 - 6 ° C.

The problem is also solved by the fact that at the positions forming the odd layers, the end of the process is defined as the achievement of the maximum radiation temperature and its subsequent decrease by 1 ° - 5 ° C.

The problem is also solved by the fact that the power of the magnetron spray station is controlled depending on the magnitude of the change in heat flux from the heaters.

The problem is solved in that for the known device for the vacuum deposition of thin semiconductor films in the vacuum corridor of the production line with sequentially arranged spray stations equipped with planar or cylindrical DC magnetrons or twin magnetrons, and the control system according to the invention, in front of the first spray station and after the first spray station, heating chambers are placed, infrared heating elements and temperature measuring devices made in the form of optical pyrometers are installed on the side of the glass substrate on the spray station, and the control system contains heat flow sensors from heaters.

The problem is also solved by the fact that the pyrometers are multi-point.

The problem is also solved by the fact that the extended vacuum chamber is made in the form of two parallel rectilinear branches of the working and return, located in the same vacuum corridor, while the return branch serves to return products to the place of unloading, loading and cooling of products in vacuum. At the same time, a lock chamber for loading and unloading products is located on one side of the vacuum corridor.

The problem is also solved by the fact that the target of the first and all odd spray stations is made of Cu-In-Ga alloy, in which the ratio of atomic concentrations is Cu / (In + Ga) [Cu / III] and Ga (In + Ga) [ Ga / III] is chosen in the range of [Cu / III] = 0.47 - 0.51 a [Ga / III] = 0.25 - 0.3, and the target of the second and subsequent even spray stations is made of Cu-Ga alloy, in which, the ratio of atomic concentrations Cu / Ga = 2.5 - 2.8, with the total number of spray stations selected odd.

The invention is illustrated by drawings.

In FIG. 1 shows a diagram of a vacuum installation for applying a CIGS film in a vacuum to a sheet glass substrate in the manufacture of solar cells in vacuum. In FIG. 2 shows a deposition station with a magnetron. In FIG. 3 to 11 show a graph of the variation in heat flux at the spray stations 1 to 9. FIG. 12 and 13 are integrated plots of the variation in heat flow temperature for sputtering a nine- and eleven-layer film. On Fig presents relationship profiles [Cu / III] and [Ga / III] obtained from numerical data. FIG. 15 is a typical SEM image of a cleaved CIGS film. FIG. 16 is an X-ray diffraction pattern of a two-layer CIGS-Mo structure on a glass substrate

The installation consists of two vacuum corridors 1 and 2 connected by a transport system 3. On the side of loading / unloading products in the lock chamber 4 and on the opposite side, rotary devices 5 of the transport system 3 are installed. The transport system 3 is made by known solutions and is, for example, a chain conveyor on which the snap-in 6 is mounted for fixing the substrate 7. In the working part of the vacuum corridor 1, heating elements 8, pyrometers 9 for measuring temperature and heat sensors are installed on the substrate side about flow 10. Moreover, the heat flux sensors 10 are installed only at the spraying positions where the magnetrons are installed 11. The heat flux sensors 10 are connected to the control system (not shown) by magnetrons 1 through control units 12 1. Heating elements 8, for example, heating elements, connected to the energy source 13 through the control unit 14 of the heating elements. The control unit 14 receives and analyzes the signals from the pyrometers 9 and controls the power sent to the heating elements 8 to ensure a constant temperature of the substrate 7. After turning the conveyor in the cooling vacuum corridor 2, the workpieces, moving forward, cool to the temperature necessary to extract them from the vacuum chamber. The extraction is carried out in the lock chamber 4, combined with a device for turning the conveyor. A layer of molybdenum 15 is preliminarily applied to the substrate 7, and a layer of a thin film (not shown) is applied during the passage through the spraying positions.

The technological process of vacuum deposition of a CIGS film consisting of 9 sequentially sprayed layers is as follows. The glass substrate 7 with a molybdenum layer 15 previously sprayed onto its surface is placed in the lock chamber 4 at room temperature and pumped out with a vacuum pump to pressure of the order of 1-10 Pa. Then the substrate is transferred to the first heating chamber, in which the temperature of the substrate is brought to 370 ° C.

Then the substrate is transferred to the first spray station, where a copper depleted layer is deposited on its surface by sputtering a target from a Cu-In-Ga alloy with an atomic concentration ratio [Cu / N] = 0.49, [Ga / III] = 0.28 at linear specific power on the target at 500 watts / cm and an operating voltage of 800V. In FIG. Figure 3 shows the temperature profile of the heat flux sensor, which records the change in the heat flux of the heaters, and therefore the power consumed by them during the deposition process of the first layer while maintaining a constant substrate temperature of 370 ° C. The moment of completion of the spraying process is determined after reaching the maximum heat flux temperature and lowering the temperature by 1 ° C. Then the substrate is moved to a second heating chamber, in which the temperature of the substrate is raised to 550 ° C.

After reaching and stabilizing the set temperature, the substrate is transferred to a second spray station, on which a copper-enriched layer is deposited on its surface by sputtering a target from a Cu-Ga alloy with a ratio of atomic concentrations Cu / Ga = 2.7 with a linear specific power on the target of 20 W / cm and an operating voltage of 500 ° V. In FIG. 4 shows the temperature profile of the heat flux sensor during the deposition process of the second layer while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the deposition process is determined after reaching a radiation temperature of a minimum and a subsequent rise in temperature by 3 ° C.

Then, the substrate is transferred to a third spray station, on which the next copper-depleted layer is deposited onto its surface by spraying a target whose composition is identical to the target composition of the first spray station. In FIG. 5 shows the temperature profile of the heat flux sensor during the deposition of the third layer while maintaining a constant substrate temperature of 550 ° C. End time the spraying process is determined after reaching the maximum radiation temperature and lowering the temperature by 5 ° C.

Then, the substrate is transferred to the fourth spray station, on which a copper-enriched layer is deposited on its surface by spraying a target whose composition is identical to the target composition of the second spray station. Figure 6 shows the temperature profile of the heat flux sensor during the deposition process of the fourth layer while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the spraying process is determined after reaching a radiation temperature of a minimum and a subsequent rise in temperature by 6 ° C.

Then the substrate is transferred to the fifth spray station, on which a copper-depleted layer is deposited on its surface by spraying a target whose composition is identical to the target composition of the first and third spray stations. Figure 7 shows the temperature profile of the heat flux sensor during the fifth layer deposition process while maintaining a constant substrate temperature of 550 ° C. The end of the deposition process is determined after reaching the radiation maximum temperature and lowering the temperature by 5 ° C.

Then the substrate is transferred to the sixth spray station, on which a copper-enriched layer is deposited onto its surface by spraying a target whose composition is identical to the target composition of the second and fourth spray stations. On Fig shows the temperature profile of the heat flow sensor during the deposition process of the sixth layer while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the spraying process is determined after reaching a radiation temperature of a minimum and a subsequent rise in temperature by 2 ° C.

Then the substrate is transferred to the seventh spray station, on which a copper-depleted layer is deposited on its surface by spraying a target whose composition is identical to the composition of the target of the first, third and fifth spray stations. Figure 9 shows the temperature heat flow sensor profile during the seventh layer deposition process while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the deposition process is determined after reaching the radiation maximum temperature and lowering the temperature by 2 ° C.

Then the substrate is transferred to the eighth spray station, on which a copper-enriched layer is deposited onto its surface by spraying a target whose composition is identical to the composition of the second, fourth and sixth spray stations. Figure 10 shows the temperature profile of the heat flux sensor during the eighth layer deposition process while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the spraying process is determined after reaching a radiation temperature of a minimum and a subsequent rise in temperature by 1 ° C.

Then, the substrate is transferred to the ninth spray station, on which a copper-depleted layer is deposited on its surface by spraying a target whose composition is identical to that of the first, third, fifth and seventh spray stations. Figure 1 1 shows the temperature profile of the heat flux sensor during the deposition process of the ninth layer while maintaining a constant substrate temperature of 550 ° C. The moment of completion of the deposition process is determined after reaching the radiation maximum temperature and lowering the temperature by 2 ° C.

On Fig presents the integral temperature profile of the deposition process of a nine-layer CIGS film.

On Fig presents the integral temperature profile of the deposition process 1 1-layer film CIGS.

Then, the substrate using the transport system 3 is moved to the rotary chamber and then to the cooling vacuum corridor 2, in which the substrate 7 with the film 15 is moved to the discharge position 4 and cooled to a temperature of 90 - 105 ° C. Then the substrate is moved to the lock chamber 4, let in air and removed from the vacuum installation.

In the process of vacuum deposition of a CIGS film, two types of targets are used with different percentages of Cu, In, Ga. The composition of the targets is determined based on the following considerations. Known [2-5] is the optimal integral composition of CIGS films with high efficiency up to 19%, characterized by the ratio Cu / (In + Ga) = 0.74 - 0.88, Ga / (In + Ga) = 0.25 - 0 , 3 and atomic concentration of selenium at the level of 50%. In the indicated concentration range, the α-phase of the CIGS chalcopyrite structure is formed, which is most effective as an absorber of sunlight. When these values are exceeded, the Cu2 – _x Se phase precipitates in the film volume, which, due to the metallic nature of the conductivity, shunts the absorber material and the region of the - - - transition. When the lower limit of the indicated values is exceeded, the CIGS film is a mixture of the a and β phases and is characterized by an increased resistivity, which leads to a decrease in the efficiency of CIGS as an absorber. To achieve the above optimal concentration ratios of elements in the film, Cu-In-Ga and Cu-Ga targets are used. The composition of the Cu-In-Ga target ([Cu / III] = 0.47 - 0.51 a [Ga / III] - 0.25 - 0.3) is determined by the metallurgical features of the manufacture of the target and the triple phase diagram of the solid solution of these elements, as a result, the odd layers of the CIGS film formed from this target have a reduced concentration of gallium and copper relative to the chemical composition of chalcopyrite. Therefore, to form the stoichiometric composition of the Cu (InGa) Se ₂ film after deposition of copper-depleted Cu-Ga layers, copper-enriched Cu-Ga layers are deposited from the Cu-Ga (Cu / Ga) = 2.5 - 2.8 target), the composition of which is selected based on from a double phase diagram of a solid solution of these elements and the technology of its manufacture.

The use of the Cu-In-Ga target for sputtering the first layer is due to the need to form a seed layer on the molybdenum contact layer in composition and structure close to chalcopyrite. The temperature of the substrate at which this layer is sprayed is slightly lower than on the other layers and is 340 - 400 ° C. This is due to the fact that, at elevated temperatures above 400 ° C, large internal stresses arise in the sprayed layer and film peeling occurs at the boundary with molybdenum. In addition, at such temperatures, film growth in the form of pillars (columnar growth) is observed, which leads to the appearance of leakage currents along the grain boundaries in the final structure of the solar cell and a decrease in its efficiency. At temperatures below 340 ° C, the process of the germinal layer slows down sharply and does not provide the required grain structure.

Since the first CIGS layer has a copper and gallium depleted composition, the total number of layers should be odd, since when spraying even layers on the surface of the growing film, there is a Cu _2-x Se phase with a metallic conductivity. It must be neutralized with a layer with a reduced copper content. In addition, to achieve high efficiency of the solar cell, when spraying the last layer, it is required to create an increased concentration of indium selenides in the surface layer of CIGS ~ 40-60 nm thick (ratio [Cu / N] <0.4). This makes it possible to invert the natural p- chalcopyrite conductivity into i-type conductivity. In this case, a hidden homogeneous β-transition is formed under the CIGS surface, which in the finished device, together with the subsequently deposited heterogeneous layer of sulfur cadmium α-type conductivity, will determine the parameters of the solar cell diode. When spraying an odd number of layers, this is quite simple to do, for example, by lowering the power on the Cu-In-Ga target of the last layer. In addition, when applying such a process scheme, a gradient of the [Ga / III] ratio is provided with a minimum at the surface and a maximum towards the boundary with molybdenum. In this case, the internal electric field of the CIGS film is formed, which contributes to the efficient transfer of charges to external contacts. On Fig.14 shows the profiles of the ratios [Cu / III] and [Ga / III] obtained from the numerical data of the Auger analysis during deep etching of the formed structure by the primary ion beam.

The number of layers and, accordingly, spray stations depends on the thickness of the CIGS film and for a film thickness of 1.3 - 2.0 μm is from 5 to 11. The thickness of the first layer cannot be selected more than 0.4 - 0.5 μm, since more thick films have weak adhesion to the molybdenum layer due to internal stresses, and further thickness growth is possible only with an increase in the total number of layers.

On Fig presents a typical SEM image of a chip CIGS film sprayed on a molybdenum layer, which serves as the back contact electrode in the structure of the solar cell. CIGS film sprayed by the proposed method has a dense coarse-grained structure in the form of equidistant grains with dimensions of the order of the thickness of the deposited layer. Grain boundaries are very rare, extending from the CIGS surface to the molybdenum layer, which is important from the point of view of reducing leakage currents along the grain boundaries in the structure of the solar cell.

On Fig presents x-ray two-layer structure of CIGS-Mo on a glass substrate. The ratio and high intensity of lines 204/220 and 312 of sample 422-SM297-290 are close to the values characteristic of polycrystalline chalcopyrite. Moreover, a significant amount of the phase corresponds to chalcopyrite strongly textured along the axis 112. From the point of view of the quality of the> -and transition in the device, the structure of the obtained layer is close to optimal.

Using the developed technology and equipment allows us to solve all six problems stated in the present description, to obtain the optimal integral composition of the material, to ensure its homogeneity, as well as to collect material with high efficiency of converting solar energy into electricity. In addition, the inventive method allows for reliable reproducibility and stability of the technology in mass production on large substrates while reducing production costs.

Sources of information taken into account during the examination:

1. US patent 5436204, M.CL. H01L 21/302, publ. 06/25/1995

2. J. Kessler, C. Chityuttakan, J. Lu, J. SchSldstrom, and L. Stolt, "Cu (In, Ga) Se2 thin films grown with a Cu-poor / rich / poor sequence: growth model and structural considerations , "Prog. Photovolt: Res. Appl., Vol. 1 1, pp. 319-331, 2003

3. A. M. Gabor, J. R. Turtle, D. S. Albin, M. A. Contreras, R. Noufi, and A. M. Hermann, "High-Efficiency CuInxGai-xSe2 Solar-Cells Made from (Iax, Gai-x) 2Se3 Precursor Films," Appl. Phys. Lett., Vol. 65, pp. 198-200, 1994.

4. N. Kohara, T. Negami, M. Nishitani, and T. Wada, "Preparation of Direct Quality Cu (In, Ga) Se2 Thin Films Deposited by Coevaporation with Composition Monitor," Jpn. J. Appl. Phys., Vol. 34, pp. LI 141-L1 144, 1995.

5. M. Nishitani, T. Negami, and T. Wada, "Composition monitoring method in CuInSe2 thin film preparation," Thin Solid Films, vol. 258, pp. 313-316, 1995.

6. US patent 6974976, M.CL. H01L 31/109, publ. 12/13/2005

7. US Patent 7544884, M.C. H01L 31/04, publ. 06/06/2009 - a prototype for the method and device.

8. A.S. USSR 1427881, M.cl. C23C 14/56, publ. 06/19/1995

Eurasian Patent Attorney,

Per. JY 0240/7) /? E.A. Swider

Claims

CLAIM

1. A method of forming thin CIGS films for large-sized solar cells by layer-by-layer deposition of a CIGS film in a vacuum on a sheet glass substrate with a layer of conductive molybdenum previously applied to it, in a vacuum corridor of a continuous line, heating the fact that the material is applied in successive layers by reactive spraying in elemental selenium vapor using successively arranged direct current magnetron sputtering stations or a twin-magnetron average hourly total sputtering of flat or cylindrical targets, while using Cu-In-Ga alloy in the ratio of atomic concentration Cu / (In + Ga) [Cu / III] and Ga as cathodes of magnetron stations at positions forming odd layers / (In + Ga) [Ga / III] is chosen within the limits of [Cu / III] = 0.47 - 0.51 a [Ga / III] = 0.25 - 0.3, and as the cathodes of magnetron stations at the positions forming even layers using a Cu-Ga alloy, in which the ratio of atomic concentrations Cu / Ga = 2.5 - 2.8, while at the position forming the first layer, the temperature of the substrate is kept constant at in the range 340–400 ° С, and at the second and subsequent positions, the temperature of the substrate is constant in the range of 530–580 ° С, and before being removed from the vacuum, the workpiece is cooled to a temperature of 90–110 ° С.

2. The method according to claim 1, in that the total number of layers is chosen odd.

3. The method according to claim 1 or claim 2, with the fact that the number of layers is selected in the range from 5 to 11.

4. The method according to any one of paragraphs c to Z o tl and h is that the fact that the temperature of the substrate is controlled by a non-contact pyrometer on the glass side.

5. The method according to any one of cl clause that the constant temperature of the substrate is supported by infrared emitters from the side of the substrate, and the magnitude of the change in thermal radiation while maintaining a constant temperature of the substrate is used as a parameter for the end of the deposition of the layer.

6. The method according to any one of cl clauses of 5 degrees, in that, at a constant speed of movement of the workpiece in the vacuum corridor in the deposition zone, the amount of change in thermal radiation to maintain a constant temperature of the substrate is used as a parameter for the deposition rate layer.

7. The method according to claim 5 or with the fact that at the positions forming even layers, the end of the process is determined after reaching the value of the radiation temperature of the minimum and its consequent rise by 0.8 - 6 ° C .

8. The method according to claim 5 or, moreover, the fact that at the positions forming the odd layers, the end of the process is defined as the achievement of the maximum radiation temperature and its subsequent decrease by 1 - 5 ° С.

9. The method according to claim 5 or with the fact that the power of the magnetron of the spray station is controlled depending on the magnitude of the change in heat flux from the heaters.

10. A device for vacuum deposition of thin semiconductor films in a vacuum corridor of a flow line with sequentially located spray stations equipped with planar or cylindrical magnetrons with direct current or twin magnetrons, and a control system, which is ensured by the fact that before the first spray station and after the first spray station the heating chambers are placed, infrared heating elements and means are installed on the side of the glass substrate on the spray station and temperature measurements made in the form of optical pyrometers, the control system comprises sensors measuring the heat flow from the heaters, target odd spray The stations are made of a Cu-In-Ga alloy, in which the ratio of atomic concentrations Cu / (In + Ga) [Cu / III] and Ga / (In + Ga) [Ga / III] is chosen within the limits [Cu / W] = 0.47 - 0.51 and [Ga / III] = 0.25 - 0.3, and the targets of even spray stations are made of Cu-Ga alloy, in which the ratio of atomic concentrations Cu / Ga = 2.5 - 2, 8, with the total number of spray stations selected odd.

11. The device according to claim 10, which includes the fact that the pyrometers are multi-point.

12. The device according to claim 0, wherein the extended vacuum chamber is made in the form of two parallel rectilinear branches of the working and return branches located in the same vacuum corridor, while the return branch serves to return products to the place unloading - loading and cooling products in a vacuum. At the same time, a lock chamber for loading and unloading products is located on one side of the vacuum corridor.

Eurasian Patent Attorney,

Per. JYo 0240 fX Ys E.A. Swider