WO2007056224A2 - Technique and apparatus for depositing layers of semiconductor for solar cell and module fabrication - Google Patents

Abstract

The present invention advantageously provides for, in different embodiments, low-cost deposition techniques to form high-quality, dense, well-adhering Group IBIIIAVIA compound thin films with macro-scale as well as micro-scale compositional uniformities. It also provides methods to monolithically integrate solar cells made on such compound thin films to form modules. In one embodiment, there is provided a method of growing a Group IBIIIAVIA semiconductor layer on a base, and includes the steps of depositing on the base a nucleation and/or a seed layer and electroplating over the nucleation and/or the seed layer a precursor film comprising a Group IB material and at least one Group IIIA material, and reacting the electroplated precursor film with a Group VIA material. Other embodiments are also described.

Description

TECHNIQUE AND APPARATUS FOR DEPOSITING LAYERS OF SEMICONDUCTORS FOR SOLAR CELL AND MODULE FABRICATION

Inventor: Bulent M. Basol

This application, U.S. Application No.: 1 1/266,013 filed November 2, 2006, is a continuation-in-part and claims priority to U.S. Application Serial No. 1 1/081,308 filed March 15, 2005, which claims priority to U.S. Provisional Application Serial No. 60/552,736 filed March 15, 2004.

FIELD OF. THE INVENTION

[0001] The present invention relates to method and apparatus for preparing thin films of semiconductors for radiation detector and photovoltaic applications. BACKGROUND

[0002] Solar cells are photovoltaic devices that convert sunlight directly into electrical power. The most common solar cell material is silicon, which is in the form of single or polycrystalline wafers. However, the cost of electricity generated using silicon-based solar cells is higher than the cost of electricity generated by the more traditional methods. Therefore, since early 1970's there has been an effort to reduce cost of solar cells for terrestrial use. One way of reducing the cost of solar cells is to develop low-cost thin film growth techniques that can deposit solar-cell-quality absorber materials on large area substrates and to fabricate these devices using high-throughput, low-cost methods.

[0003] Group IBIIIA VIA compound semiconductors comprising some of the Group IB (Cu, Ag, Au), Group IIIA (B, Al, Ga, In, Tl) and Group VIA (O, S, Se, Te, Po) materials or elements of the periodic table are excellent absorber materials for thin film solar cell structures. Especially, compounds of Cu, In, Ga, Se and S which are generally referred to as CIGS(S), or Cu(In,Ga)(S,Se)₂ or CuIni_-xGa_x (S_ySe[_-y)_k , where 0<x<l, 0<y<l and k is approximately 2, have already been employed in solar cell structures that yielded conversion efficiencies approaching 20%. Absorbers containing Group IIIA element Al and/or Group VIA element Te also showed promise. Therefore, in summary, compounds containing: i) Cu from Group IB, ii) at least one of In, Ga, and Al from Group IIIA, and iii) at least one of S, Se, and Te from Group VIA, are of great interest for solar cell applications.

[0004] The structure of a conventional Group IBIIIAVIA compound photovoltaic cell such as a Cu(In,Ga,Al)(S,Se,Te)₂ thin film solar cell is shown in Fig 1. The device 10 is fabricated on a substrate 1 1, such as a sheet of glass, a sheet of metal, an insulating foil or web, or a conductive foil or web. The absorber film 12, which comprises a material in the family of Cu(In,Ga,Al)(S,Se,Te)a , is grown over a conductive layer 13, which is previously deposited on the substrate 11 and which acts as the electrical contact to the device. Various conductive layers comprising Mo, Ta, W, Ti, and stainless steel etc. have been used in the solar cell structure of Fig. i . If the substrate itself is a properly selected conductive material, it is possible not to use a conductive layer 13, since the substrate 11 may then be used as the ohmic contact to the device. After the absorber film 12 is grown, a transparent layer 14 such as a CdS, ZnO or CdS/ZnO stack is formed on the absorber film. Radiation 15 enters the device through the transparent layer 14. Metallic grids (not shown) may also be deposited over the transparent layer 14 to reduce the effective series resistance of the device. The preferred electrical type of the absorber film 12 is p- type, and the preferred electrical type of the transparent layer 14 is n-type. However, an n-type absorber and a p-type window layer can also be utilized. The preferred device structure of Fig. 1 is called a "substrate-type" structure. A "superstrate-type" structure can also be constructed by depositing a transparent conductive layer on a transparent superstrate such as glass or transparent polymeric foil, and then depositing the Cu(In,Ga,Al)(S,Se,Te)2 absorber film, and finally forming an ohmic contact to the device by a conductive layer. In this superstrate structure light enters the device from the transparent superstrate side. A variety of materials, deposited by a variety of methods, can be used to provide the various layers of the device shown in Fig 1.

[0005] In a thin film solar cell employing a Group IBIIIAVIA compound absorber, the cell efficiency is a strong function of the molar ratio of IB/IIIA. If there are more than one Group IIIA materials in the composition, the relative amounts or molar ratios of these IIIA elements also affect the properties. For a Cu(In₁Ga)(S₁Se)₂ absorber layer, for example, the efficiency of the device is a function of the molar ratio of Cu/(In+Ga). Furthermore, some of the important parameters of the cell, such as its open circuit voltage, short circuit current and fill factor vary with the molar ratio of the IIIA elements, i.e. the Ga/(Ga+In) molar ratio. In general, for good device performance Cu/(In+Ga) molar ratio is kept at around or below 1.0. As the Ga/(Ga+In) molar ratio increases, on the other hand, the optical bandgap of the absorber layer increases and therefore the open circuit voltage of the solar cell increases while the short circuit current typically may decrease. It is important for a thin film deposition process to have the capability o controlling both the molar ratio of IB/IIIA, and the molar ratios of the Group IIIA components i the composition. It should be noted that although the chemical formula is often written as Cu(In,Ga)(S,Se)₂, a more accurate formula for the compound is Cu(In,Ga)(S,Se)k, where k is typically close to 2 but may not be exactly 2. For simplicity we will continue to use the value of k as 2. It should be further noted that the notation "Cu(X₇Y)" in the chemical formula means all chemical compositions of X and Y from (X=0% and Y= 100%) to (X= 100% and Y=0%). For example, Cu(In₁Ga) means all compositions from CuIn to CuGa. Similarly, Cu(In₅Ga)(S₅Se)₂ means the whole family of compounds with Ga/(Ga+In) molar ratio varying from 0 to 1, and Se/(Se+S) molar ratio varying from 0 to 1.

[0006] The first technique that yielded high-quality Cu(In₅Ga)Se₂ films for solar cell fabrication was co-evaporation of Cu₅ In, Ga and Se onto a heated substrate in a vacuum chamber. However, low materials utilization, high cost of equipment, difficulties faced in large area deposition and relatively low throughput are some of the challenges faced in commercialization of the co-evaporation approach.

[0007] Another technique for growing Cu(In₅Ga)(S₅Se)₂ type compound thin films for solar cell applications is a two-stage process where metallic components of the Cu(In₅Ga)(S₅Se)₂ material are first deposited onto a substrate, and then reacted with S and/or Se in a high temperature annealing process. For example, for CuInSe₂ growth, thin layers of Cu and In are first deposited on a substrate and then this stacked precursor layer is reacted with Se at elevated temperature. If the reaction atmosphere also contains sulfur, then a CuIn(S₅Se)₂ layer can be grown. Addition of Ga in the precursor layer, i.e. use of a Cu/In/Ga stacked film precursor, allows the growth of a Cu(In,Ga)(S,Se)₂ absorber.

[0008] Sputtering and evaporation techniques have been used in prior art approaches to deposit the layers containing the Group IB and Group II1A components of the precursor stacks. In the case of CuInSe₂ growth, for example, Cu and In layers were sequentially sputter-deposited on a substrate and then the stacked film was heated in the presence of gas containing Se at elevated temperature for times typically longer than about 30 minutes, as described in U.S. 4,798,660. More recently U.S. Patent 6,048,442 disclosed a method comprising sputter- depositing a stacked precursor film comprising a Cu-Ga alloy layer and an In layer to form a Cu- Ga/In stack on a metallic back electrode layer and then reacting this precursor stack film with one of Se and S to form the absorber layer. U.S. Patent 6,092,669 described sputtering-based equipment for producing such absorber layers. Such techniques may yield good quality absorber layers and efficient solar cells, however, they suffer from the high cost of capital equipment, and relatively slow rate of production.

[0009] One prior art method described in U.S. Patent 4,581 , 108 utilizes a low cost electrodeposition approach for metallic precursor preparation. In this method a Cu layer is first electrodeposited on a substrate. This is then followed by electrodeposition of an In layer and heating of the deposited Cu/In stack in a reactive atmosphere containing Se. Although low-cost in nature, this technique was found to yield CuInSe₂ films with poor adhesion to the Mo contact layer. In a publication ("Low Cost Thin Film Chalcopyrite Solar Cells", Proceedings of 18^th IEEE Photovoltaic Specialists Conf., 1985, p. 1429) electrodeposition and selenization of Cu/In and Cu/In/Ga layers were demonstrated for CIS and CIGS growth. One problem area was identified as peeling of the compound films during solar cell processing. Later, in another reference ("Low Cost Methods for the Production of Semiconductor Films for CIS/CdS Solar Cells", Solar Cells, vol. 21, p. 65, 1987) researchers studied the cross-section of Mo/CuInSe₂ interface obtained by the above-mentioned method and found the CuInSe₂ to have poor adhesion to the Mo contact layer.

[00010] Irrespective of the specific approach used in a two-stage process, growing for example a Cu(In, Ga)(S₁Se)₂ absorber film, individual thicknesses of the layers forming the metallic stacked structure need to be controlled so that the two molar ratios mentioned before, i.e. the Cu/(In+Ga) ratio and the Ga/(Ga+In) ratio, can be kept under control from run to run and on large area substrates. The molar ratios attained in the metallic stacked structures are generally preserved in macro scale during the reaction step, provided that the reaction temperature is kept below about 600 °C. Therefore, the overall or average molar ratios in the compound film obtained after the reaction step is about the same as the average molar ratios in the metallic stacked structures before the reaction step. In prior art approaches all the Cu, In and/or Ga required for the final desired molar ratios are deposited on the substrate before the reaction step with S and/or Se. In other words, to grow a Cu₀.gIn₀.gGa₀.₂Sex, where x is close to 2, for example, the prior art techniques typically deposit a Cu/In/Ga stack, a In/Cu/Ga stack or a Cu-Ga/In stack, so that the Cu/(In+Ga) molar ratio of the stack is 0.8 and the Ga/(Ga+In) molar ratio of the stack is 0.2. This metallic stack is then selenized at high temperatures to form the compound. One problem associated with such approaches is that these precursors are relatively thick (500-1500 nm) and are typically rich in Group HIB components of In and Ga, which have low melting points of about 156 °C and 30 ⁰C, respectively, and they cause micro-scale non-uniformities, as will be discussed next.

[00011] Figs. 2a-2c demonstrate the problem of micro-scale non-uniformities which may be present in the metallic precursor layers, especially those with IB/IIIA molar ratios of less than or equal to 1. Fig. 2a schematically shows an exemplary Cu/In/Ga metallic stack with the exemplary overall molar ratios of Cu/(In+Ga)=0.8 and Ga/(Ga+In)=0.2 on a substrate. In this approach, a contact film 21 is first deposited onto a substrate 20, forming a base 22. A Cu layer 23 is then deposited over the contact film 21. The thickness of the Cu layer 23 may, for example, be about 200 nm. This Cu thickness and the desired molar ratios cited above require deposition of about 440 nm thick In layer and about 80 nm thick Ga layer. These calculations can be made assuming densities of Cu, In and Ga to be 8.96 g/cc, 7.31 g/cc and 5.91 g/cc, respectively, and the atomic weights to be 63.54 g, 114.76 g and 69.72 g, respectively. Using a density value of 5.75 g/cc and a molar weight of 306.66 g for the selenized compound of Cuo_.8lno.₁sGao.₂Se₁.₉, it is also calculated that the metallic precursor of this example would yield approximately 1880 nm thick CIGS layer, assuming 100% density. The optimum thickness of CIGS layers for thin film solar cell applications is in the range of 500-5000 nm, preferably in the range of 700-2000 nm, smaller thicknesses being preferable because it reduces materials cost.

[00012] Referring back to Fig. 2a, deposition of about 200 nm thick Cu layer 23 is followed by the deposition of a nominally 440 nm thick In layer 24 and a nominally 80 nm thick Ga layer 25. The Cu layer 23 in the resulting metallic precursor stack 26 is shown as smooth and uniform, whereas, the In and Ga layer surfaces are depicted as non-planar. Although the surface morphologies of these layers depend strongly on the deposition techniques used, it is generally true that low-melting-temperature metals such as In and Ga tend to "ball" when deposited in thin film form, especially if they are deposited on top of each other. It should be noted that melting temperature of (Ga+In) mix is lower than the melting temperature of In which is around 156 ⁰C.

[00013] The metallic precursor layer 26 of Fig. 2a may have the desired Cu/(In+Ga) and Ga/(Ga+In) molar ratios in a global sense or in macro scale. However, in micro scale the situation is quite different as can be seen in Fig.2b which shows a magnified view of the region 27 of Fig. 2a. Since the In layer thickness "tl" at and around point "A" is much larger than the In layer thickness "t2" at point B, the local Cu/(In+Ga) ratio is much smaller at and around point "A" than at and around point "B". Furthermore Ga/( Ga+In) ratio is also different at these two points, i.e. higher at point B compared to point A. It should be appreciated that after the reaction step with Se these micro-scale non-uniformities in the molar ratios of the metallic components are mostly transferred to the compound, yielding a CIGS layer that has in-plane compositional variations due to varying Cu/(In+Ga) and Ga/(Ga+In) ratios. This situation is schematically shown in Fig. 2c, which depicts a compound layer 29 obtained by reacting the precursor stack of Fig. 2B with Se. The region Rl in Fig. 2c corresponds approximately to the area around point A in Fig. 2b and the region R2 corresponds approximately to the area around point B in Fig. 2b. Accordingly, region Rl is an In-rich region and region R2 is a Cu-rich region. It should be noted that boundaries between these regions may not be as defined as suggested in Fig. 2c. Boundaries are shown as such just to demonstrate the point. In real films even the crystalline structures of these regions may be different. Cu-rich regions, after selenization, may contain large faceted grains of Cu-selenides, whereas, In or Ga-rich regions would be smoother with smaller grains. When a solar cell is fabricated on the compound layer 29 of Fig. 2c, the copper rich region R2 containing highly conductive Cu-selenide phases would increase the leakage current across the device and reduce its voltage output, whereas the In-rich region Rl would increase its series resistance. Both effects would deteriorate the solar cell efficiency if they are extreme. Since the non-uniform surface morphologies such as the one depicted in Figs. 2a-2c vary from run to run and from substrate to substrate, repeatability of the manufacturing process of this solar cell would also be poor and the yield for high efficiency large area device fabrication would be low. For best solar cell efficiencies and high manufacturing yields compound layers with macro-scale as well as micro-scale compositional uniformity are needed.

[00014] It should be noted that the example given above explained the micro-scale non- uniformity problem for the case of non-uniform or rough Group HIA layer deposited on a smooth Group IB layer. However, similar problems are observed even if the morphologies of In and Ga layers were smooth in their as-deposited state. The reason is that even if the starting morphologies of In and Ga were smooth, during the reaction step with the Group VIA material, such as Se, the metallic precursor is heated up, typically to temperatures above 350 °C. As the heating step is performed, In and Ga start to melt at temperatures above about 30 ⁰C before reacting with the Group VIA material and they de-wet the substrate they are deposited on such as the Cu surface. This de-wetting phenomenon forms "balls" giving rise to rough morphologies similar to the one shown in Fig. 2b. Furthermore, balling becomes more extensive as the amount of the low-melting phase (In and/or Ga) in the stack is increased or the thickness of In and/or Ga layers are increased.

[00015] Fig. 3a shows an exemplary metallic precursor stack 36 deposited over a base 22, wherein the base 22 comprises a substrate 20 and a contact film 21 as in Fig. 2a, and the metallic precursor stack 36 comprises a substantially smooth Cu layer 33, a substantially smooth In layer 34 and a substantially smooth Ga layer 35, which may be deposited onto the base 22 by various thin film deposition techniques such as evaporation, sputtering or electrodeposition by taking certain measures such as keeping the base at below room temperature by forced cooling during deposition. In this example, the individual Cu, In and Ga layers within the stack 36 have flat surface morphologies. Fig. 3b shows the morphology of this precursor stack after it is heated up to a temperature of for example 160 °C, which is higher than the In melting temperature as well as the melting temperature of the (Ga+In) composition lying over the Cu layer 33. Since the (Ga+In) composition of our example is 20% Ga and 80% In, the melting temperature is in the range of 100-120 "C according to the In-Ga binary phase diagram. The surface morphology of the (In+Ga) layer of Fig. 3b is very non-uniform and would give rise to the micro-scale compositional non-uniformities in the compound layers fabricated using such precursors as described before with reference to Figs. 2a, 2b and 2c. It should be noted that although the interface between the Cu layer 33 and the (In+Ga) layer 36a of Fig. 3b is shown to be sharp, this interface in practice may be diffused depending upon the temperature of the heat treatment step.

[00016] One approach to address the micro-scale non-uniformity problem is described in US Patent 5,567,469 awarded to Wada et al. In this approach a portion of the low-melting phase or component, e.g. indium, is introduced into the precursor layer in the form of a compound selected from the group consisting of oxides, selenides and sulfides. These compounds of In have very high melting points. Therefore, when the precursor is heated up to perform the reaction step with Group VIA components), melting and balling of In is reduced since at least some of the In is in the form of a high melting point compound.

[00017] The discussion above concentrated on the problem of micro-scale compositional non-uniformities in metallic precursor layers used in prior art two-stage process approaches. An additional important problem, namely adhesion, was also identified for metallic precursor layers obtained by the low cost electrodeposition technique. Although electrodeposition is attractive to use in terms of its lower cost, there are other limiting factors in using the prior art electroplating approaches for the preparation of metallic Group IB and Group IIIA elemental stacks for the fabrication of Group IBIIIAVIA compound films as will be described below.

[00018] Cu, In and Ga have very different plating potentials. The molar standard electrode potentials of Cu/Cu²⁺, In/In³⁺ and Ga/Ga³⁺ metal/ion couples in aqueous solutions are about +0.337 V, -0.342 V, and -0.52 V, respectively. This means that Cu can be plated out at low negative voltages. For In deposition, on the other hand, larger negative voltages are needed. For Ga deposition even larger negative voltages are required. Therefore, to form a stack containing Cu, In and Ga, Cu is typically electroplated first. This is then followed by deposition of In and then Ga. Otherwise, while electroplating one species, the other species on which deposition is carried out may partially dissolve into the electrolyte. For example, if a stack of Cu/Ga/In is electrodeposited, while depositing In over Ga, some Ga may be dissolving into the In deposition solution. This then would lead to poor control over Cu/(Ga+In) and Ga/(Ga+In) molar ratios in the precursor and the absorber layer after its formation. Similarly, deposition of a Cu layer over an In layer may result in loss of In from the In layer into the Cu plating electrolyte during processing. Therefore, prior-art methods have employed Cu/In/Ga stacks electroplated in that order. However, after selenization such stacks yielded compound layers with poor adhesion to the base or the substrate. Furthermore, compositional micro-scale non-uniformities such as those described with reference to Figs. 3a and 3b did not allow formation of high quality Group IBIIIAVIA layers suitable for the fabrication of high efficiency solar cells. It should be appreciated that thin film deposition techniques with adhesion problems cannot be reliably scaled up for manufacturing of electronic devices, especially solar cells which are expected to have a lifetime of over 20 years.

[00019] As the brief review above demonstrates, there is still a need to develop low- cost deposition techniques to form high-quality, dense, well-adhering Group IBIIIAVIA compound thin films with macro-scale as well as micro-scale compositional uniformities.

SUMMARY OF THE INVENTION

[00020] The present invention advantageously provides for, in different embodiments, low-cost deposition techniques to form high-quality, dense, well-adhering Group IBIIIAVIA compound semiconductor thin films with macro-scale as well as micro-scale compositional uniformities.

[00021] In one embodiment, there is provided a method of growing a Group IBIIIAVIA semiconductor layer on a base. The method includes depositing on the base a nucleation layer and a precursor film comprising a Group IB material and at least one Group MA material.

[00022] In a preferred embodiment, the method further includes the step of reacting the precursor film with a Group VIA material to grow the desired semiconductor layer.

[00023] In another embodiment there is provided a method of growing a Group IBIIIAVIA semiconductor layer on a base. The method includes depositing on the base a nucleation layer and electroplating over the nucleation layer a precursor film comprising a Group IB material and at least one Group IIIA material, and reacting the electroplated precursor film with a Group VIA material.

[00024] In yet another embodiment a seed layer is first deposited on the base before the precursor film is deposited and reacted with a Group VIA material. [00025] In various other embodiments, different approaches are described to monolithically integrate Group IBIIIAVIA compound solar cells on a single substrate to fabricate modules.

BRIEF DESCRIPTION OF THE DRAWINGS

[00026] These and other aspects and features of the present invention will become apparent to those of ordinary skill in the art upon review of the following description of specific embodiments of the invention in conjunction with the accompanying figures, wherein:

[00027] FIG. 1 is a cross-sectional view of a solar cell employing a Group IBIIIAVIA absorber layer.

[00028] FIG. 2a shows a prior art Cu/In/Ga metallic precursor stack deposited on a base exemplifying the non-uniform surface morphology of In and Ga layers due to "balling".

[00029] FIG. 2b shows a larger view of section 27 in FIG. 2a.

[00030] FIG. 2c shows a prior art Group IBIIIAVIA compound film with micro-scale compositional non-uniformity that is obtained by reacting the metallic precursor layer shown in FIG. 2b with a Group VIA material.

[00031] FIG. 3a shows a Cu/In/Ga metallic precursor stack deposited on a base exemplifying a uniform surface morphology in its as-deposited form. FIG. 3B shows the metallic precursor of FIG. 3a after it is heated to elevated temperatures exemplifying formation of a non-uniform surface morphology due to melting of Group IDA phases.

[00032] FIG. 4 shows a Cu-In binary phase diagram.

[00033] FIG. 5 shows a Cu-Ga binary phase diagram.

[00034] FIG. 6 shows a preferred process sequence of the present invention.

[00035] FIG. 7 shows a preferred precursor stack comprising Cu-Ga alloy and In layers.

[00036] FIG.8a shows a first metallic precursor sub-layer with non-flat surface topography.

[00037] FIG.8b shows a second metallic sub-layer deposited on the non-flat surface of the first metallic sub-layer, forming a flat surface topography.

[00038] FIG. 9 shows an overall precursor layer with flat surface comprising a porous and rough precursor film and an electroplated layer deposited over it in a leveling fashion.

[00039] FIG. 10 shows an apparatus for electroplating, cleaning and annealing thin layers. 43099

[00040] FIG. 11 a shows a Cu layer deposited on a base.

[00041] FIG. 1 Ib shows a Cu/(Ga-bearing film) stack deposited on a base.

[00042] FIG. 1 Ic shows the alloy layer formed on the base by reacting the stack of FIG.

HB.

[00043] FIGS. 12 a- I2e show steps of a monolithic integration approach.

[00044] FIGS. 13d-13d show process steps to obtain high resistivity between contact layer pads.

[00045] FIGS.14a- 14c show some steps of an alternative monolithic integration technique.

[00046] FIG. 15a shows a seed layer or nucleation layer deposited over contact layer pads.

[00047] FIG. 15b shows a precursor layer deposited on the structure in FIG. 15 A.

[00048] FIG. 16 shows a structure with high resistance material formed between contact layer pads.

[00049] FIGS.17a and 17b show precursor stack structures.

[00050] FIG. 17c shows the compound layer formed over contact layer pads using precursor structures of Figs. 17a and 17b.

[00051] FIGS. 18a- 18e show processing steps for monolithic integration.

DETAILED DESCRIPTION

[00052] Present invention overcomes the shortcomings of prior art techniques by addressing the important manufacturability and yield issues such as micro-scale compositional control and adhesion of the semiconductor absorber films to their substrates. It also enables low cost manufacturing of thin film solar cells.

[00053] In one embodiment, the metallic precursor preparation step is divided into at least two sub-steps to enhance more intimate mixing and reaction between the Group IB materials and Group IIIA materials, to reduce de-wetting of the substrate by the low melting point Group IIIA material phases and to reduce micro-scale compositional non-uniformity. As shown in Figs. 6a-6d the first step of this metallic precursor preparation approach is the deposition or application of a first layer 60 and a first film 61 over a base 22. The terms layer and film are used interchangeably herein, though at times different terminology is used solely for convenience of differentiating words, and should be construed as such. Similarly, depositing of layers on a substrate is also referred to as growing of layers on the substrate or application of layers to the substrate. It should be noted that the base 22 may be similar to the base shown in Fig. 3a. Layers within the base 22 are not shown to simplify the drawing so that the invention could be described more clearly.

[00054] Referring back to Fig. 6a, the first layer 60 comprises at least one Group IB material and the first film 61 comprises at least one Group IIIA material. Although what is shown in Fig. 6a is the preferred deposition order of the first layer and the first film, it is possible to change this order, i.e. to deposit the first film 61 first on the base 22 and to deposit the first layer 60 over the first film 61. The preferred thickness for the first layer 60 is in the 10-150 ran range, more preferably in the 50-100 nm range. The preferred thickness for the first film 61 is in the 20- 250 nm range, more preferably in the 100-200 nm range. The structure shown in Fig. 6a is treated at a first treatment step to promote intermixing between the first layer 60 and the first film 61. The first treatment step may comprise heating, microwave treating, laser treating etc. of the structure. Treatment may be carried out in air, in vacuum, in a reducing atmosphere comprising for example hydrogen or carbon monoxide or in a substantially inert atmosphere at a temperature within the range of 50-350 °C, preferably within the range of 80-200 °C, for a time period varying from one second for the case of laser treating to 30 minutes for the case of furnace annealing. Preferably treatment is carried out for a period of 5-600 seconds, more preferably for a period of 5-300 seconds. The first treatment step results in a first intermixed layer 62 on the base 22 as shown in Fig. 6b. The intermixed layer 62 comprises Group IBIIIA solid solutions and/or alloys and has a much smoother surface morphology compared to the layer depicted in Fig. 3B because the first film 61 comprising low temperature melting Group IIIA materials is thin. With thinner Group IIIA material layer, the surface tension is lower and therefore the balling phenomenon is less severe.

[00055] The next step in the process is deposition of a second layer 63 and a second film 64 on the first intermixed layer 62 as shown in Fig. 6c. The second layer 63 comprises at least one Group IB material and the second film 64 comprises at least one Group IIIA material. Although what is shown in Fig. 6c is the preferred deposition or application order of the second layer and the second film, it is possible to change this order, i.e. to deposit the second film 64 on the first intermixed layer 62 and to deposit the second layer 64 over the second film 64. The preferred thickness for the second layer 63 is in the 10-150 nm range, more preferably in the 50- 100 nm range. The preferred thickness for the second film 64 is in the 20-250 nm range, more preferably in the 100-200 nm range. The structure shown in Fig. 6c is treated at a second treatment step to promote intermixing between the first intermixed layer 62, the second layer 63 2006/043099

and the second film 64. The second treatment step may comprise heating, microwave treating, laser treating etc. of the structure. Treatment may be carried out in air, in vacuum, in a reducing atmosphere or in a substantially inert atmosphere at a temperature within the range of 50-350 °C, preferably within the range of 80-200 ⁰C, for a time period varying from one second for the case of laser treating to 30 minutes for the case of furnace annealing. Preferably treatment is carried out for a period of 5-600 seconds, more preferably for 3-300 seconds. The second treatment step results in a second intermixed layer 65 on the base 22 as shown in Fig. 6d. The second intermixed layer 65 comprises Group IBIIIA solid solutions and/or alloys and has substantially flat surface morphology and uniform micro-scale composition.

[00056] It should be noted that the deposition and treatment steps described above may be repeated several times, preferably 2-5 times, to obtain a metallic precursor of desired thickness and composition with micro-scale compositional uniformity. If more steps are used, the individual thicknesses of the deposited or applied layers may be reduced, improving the surface morphology. After the desired thickness of the Group IBIIIA metallic precursor is obtained, such as shown in Fig. 6d, this precursor may be reacted with at least one Group VIA material to form a Group IBIIIAVIA compound layer with high density and good micro-scale compositional uniformity. Referring back to Fig. 6c, optionally, once the structure in this figure is obtained the structure may be reacted with a Group VIA material without carrying out the second treatment step. This way a Group IBIIIAVIA compound layer with good micro-scale compositional uniformity may also be formed. Referring back to Figs. 6a-6c, this preferred embodiment of the present invention improves the surface morphology and micro scale compositional uniformity of the metallic precursor by dividing the precursor deposition process into multiple steps and introducing intermixing or annealing steps. This way, non-uniformities in individual intermixed layers are minimized since they are thin. Even if there were non-uniformities in the first intermixed layer formed on the base, the next intermixed layer formed on the first one would reduce these non-uniformities.

[00057] The invention will now be described using the example of forming or growing a Cuo_.glno.sGao.oSei_.g layer.

[00058] Example 1: A Mo coated glass sheet may be used as the base. A 100 nm thick Cu layer may be deposited over the Mo layer. This is then followed by the deposition of a 220 nm thick In film and a 40 nm thick Ga layer. The stack is annealed at a temperature of 80-200 "C for 5-600 seconds to enhance alloying between Cu, In and Ga. Over the alloyed layer, 100 nm of Cu, 220 nm of In and 40 nm of Ga are then deposited or applied. The precursor is selenized by well known approaches such as in hydrogen selenide gas or selenium vapor to form the Cuo.₈Ino.₈Gao_.2Sei.₉ compound. It should be noted that selenization may be carried out by various other means such as depositing Se over the metallic precursor and heating up the stacked layer, heating the substrate in a Se-containing gaseous or liquid atmosphere etc. for times ranging from 5 minutes to 60 minutes.

[00059] Example 2: A Mo coated glass sheet may be used as the base. A 100 nm thick Cu layer may be deposited over the Mo layer. This is then followed by the deposition of a 220 nm thick In film and a 40 nm thick Ga layer. The stack is annealed at a temperature of 80-200 °C for 5-600 seconds to enhance alloying between Cu, In and Ga. Over the alloyed layer, a 100 nm of Cu, a 220 nm of In and 40 nm of Ga are then deposited. A second anneal step is applied at 80- 200 ⁰C for 5-600 seconds to promote further alloying between the layers of the metallic precursor. The precursor thus obtained is then selenized by well known approaches such as in hydrogen selenide or selenium vapor to form the Cuo.₈Ino.₈Gao.2Sei.₉ compound. It should be noted that selenization may be carried out by various other means such as depositing Se over the metallic precursor and heating up the stacked layer, heating the substrate in a Se-containing gaseous or liquid atmosphere etc. for times ranging from 5 minutes to 60 minutes.

[00060] Example 3: The approaches in Example 1 or Example 2 are used except that Cu, In and Ga layers may be deposited in four steps instead of two steps. Accordingly, thickness of Cu, In and Ga for each deposition step may be reduced to 50 nm, 110 nm and 20 nm, respectively. By heat treating the layers after each deposition step for reduced times of preferably 2-300 seconds (except for the last one for the case of Example 1), a smooth and compositionally uniform metallic precursor may be obtained. Selenization of this precursor yields compositionally uniform, high quality Cuo.sIno.βGao.aSei_.g compound layer.

[00061] In another embodiment of the present invention the morphology of the deposited films are further improved and the micro-scale compositional uniformity is further enhanced by dividing the metallic precursor preparation step into at least two sub-steps and selecting the composition of the sub-layers deposited by the sub-steps such that the sub-layers deposited early on the base do not contain substantial amount of segregated Group IIIA materials after the treatment steps promoting mixing and alloying between the metallic components. This approach will now be explained by using the phase diagrams shown in Figs. 4 and 5.

[00062] Fig. 4 shows the binary phase diagram for Cu-In (reference: P.R. Subramanian and D.E. Laughlin, Bulletin of Alloy Phase Diagrams, vol. 10, No.5, p.554, 1989), including the position of the stable alloy phase of Cunlng. It can be seen from this diagram that if a film with Cu/In ratio of less than 1 1/9 (1.22) is heated above about 156 °C, a liquid phase of In-rich solution would segregate from a Cu-rich phase of Cui iln₉ and/or other Cu-rich Cu-In alloy phases with around 30-37% In. If the Cu/In ratio is more than and equal to 1.22 but less than about 1.7 (region A in Fig. 4) and the temperature is in the 156-310 °C range, only solid phases of Cu_lling and those corresponding to about 37% In would be present in the film under equilibrium conditions. If the Cu/In ratio is more than about 1.7, then only Cu-rich solid phases would be present under equilibrium conditions up to a temperature of about 550 °C.

[00063] Similarly, it can be seen from the Cu-Ga binary phase diagram (reference: M. Hansen, Constitution of Binary Alloys, Mc. Graw Hill, 1958, p.583) of Fig. 5 that compositions with a Ga/Cu ratio of larger than 2 would yield a Ga-rich liquid phase if heated to above 30 ⁰C. In region B of the phase diagram (about 40-67 % Ga, and temperatures up 254 ⁰C) there would be only solid phases. For a Ga content of about less than 40%, only solid phases would be present even at high temperatures of above 550 ⁰C.

[00064] Present embodiment of the invention minimizes liquid phase segregation during the mixing/alloying step of the process by carefully selecting the Group IB/IIIA molar ratios i.e. Cu/In, Cu/Ga or Cu/(In+Ga) molar ratios, of the sub-layers deposited on the base. This way, de-wetting and balling by the low-melting phases is minimized along with micro-scale compositional non-uniformities resulting from them as will be described in the example given below.

[00065] Example 4: A CuInSe₂ layer may be formed on a Mo coated glass base by carrying out the following steps; a) deposit a 200 nm of Cu on the base, b) deposit a 360 ran of In over the Cu so that the Cu/In molar ratio is 1.22, c) treat the stack at a temperature range of 156 - 310 °C, preferably for a time period of 5-600 seconds,to form an alloyed layer comprising substantially the Cuπln9 solid phase, d) deposit 80 nm of In over the alloyed layer, and e) selenize thus obtained metallic precursor as described before. It should be noted that after the last In deposition a low temperature anneal step, such as at 100-200 ⁰C for about 2-300 seconds, may also be carried out before the selenization step to enhance intermixing between the alloyed layer and the last In layer.

[00066] Example 5: The process of Example 4 may be varied in a way that even more Cu (alternately less In) may be deposited in steps a) and b) so that the Cu/In ratio is more than 1.22 but is within the region A of Fig. 4. In this case after step c) the alloyed layer comprises Cuπlng solid phase as well as other solid phases shown in Fig. 4, with In content of 37% or less, to the left of region A. In this case more In needs to be deposited in step d) to compensate for the higher Cu/In ratio of the alloyed layer. Rest of the steps would be similar to those described in Example 4. It should be noted that the Cu/In ratio in the alloyed layer may be made even higher by selecting the thicknesses of the Cu and In layers corresponding to less than about 37% In (to the left of region A). In this case a much higher temperature range (up to about 600 ⁰C) may be used during the treating step c) without forming a liquid phase to cause de-wetting and balling.

[00067] Example 6: A CuInSe₂ layer may be formed on a Mo coated glass base by carrying out the following steps; a) deposit 100 nm of Cu on the base, b) deposit 180 nm of In so that Cu/In molar ratio is 1.22, c) treat the stack at a temperature range of 156-310 ^°C, preferably for a period of 2-300 seconds, to form an alloyed layer comprising substantially the Cuulng solid phase, d) repeat steps a), b) and c), and then e) deposit 80 nm of In and, f) selenize thus obtained metallic precursor as described before. It should be noted that after the last In deposition step a low temperature anneal step, such as at 100-200 °C for about 2-300 seconds, may also be carried out before the selenization step to enhance intermixing between the alloyed layer and the last In layer.

[00068] Example 7: A CuInSe₂ layer may be formed on a Mo coated glass base by carrying out the steps of Example 6 except that the thicknesses of the Cu and In films are adjusted in steps a), b) and d) to yield Cu/In molar ratios of more than 1.22 and the thickness of the In layer in step e) is adjusted to yield an overall Cu/In ratio of 1. In this case higher temperatures of up to about 600 °C may be used in the treatment step c) for even shorter time periods of 2-10 seconds, especially if the overall In content in alloyed layers is less than about 37%.

[00069] Example 8: A CuGaSe₂ layer may be formed on a Mo coated glass base by carrying out the following steps; a) deposit 200 nm of Cu on the base, b) deposit 264 nm of Ga so that Cu/Ga molar ratio is about 1.5, c) treat the stack at a temperature range of 30-600 ⁰C, preferably for a period of 5-600 seconds, to form an alloyed layer comprising substantially the Cu-rich solid phases with compositions to the left of region B in Fig. 5, d) deposit about 66 nm of Ga over the alloyed layer to make the overall Cu/Ga ratio about 1, and e) selenize thus obtained metallic precursor as described before. It should be noted that after the last Ga deposition step a low temperature (preferably < 254 °C) anneal may also be carried out for a brief period such as for 2-300 seconds before the selenization step to enhance intermixing between the alloyed layer and the last Ga layer.

[00070] Examples 4-8 above described the present embodiment in terms of CuInSe₂ and CuGaSe₂ film growth. Those skilled in the art would recognize that same approaches may be utilized for the growth of Cu(In₁Ga)Se₂ or Cu(In₁Ga)(SSe)₂ layers of varying compositions and in general for the growth of many different Group IBIIIAVIA compound layers. One specific embodiment which will be described in the example below comprises depositing a Cu and In containing layer which yields an alloy composition without low melting phases after a first treatment step, and a Cu and Ga containing layer which yields an alloy composition without low melting phases after a second treatment step. When formed on top of each other these two layers form an overall metallic precursor film with desired composition in macro as well as micro scale.

[00071] Example 9. A Cu(Ino_.ό9Gao.₃i)Se_k layer with k close to 2 may be formed on a Mo coated substrate by carrying out the following steps; a) deposit a Cu layer on the Mo surface, b) deposit an In layer on the Cu layer so that the Cu/In molar ratio is about 1.22, c) treat the stack at an elevated temperature, preferably at a temperature in the range of 156-310 °C for a period of 5-600 seconds to form a first alloyed layer comprising substantially the Cuπlrtø solid phase, d) deposit a Cu layer on the alloyed layer, e) deposit a Ga layer over the Cu layer such that the Cu/Ga ratio is about 0.5, f) treat the stack at an elevated temperature preferably at a temperature in the range of 30- 254 ⁰C for a period of 5-600 seconds to form a second alloyed layer comprising substantially the CuGa₂ solid phase, g) selenize thus obtained metallic precursor. It should be noted that the thicknesses of Cu, In and Ga deposited in steps a), b), d) and e) may be adjusted to adjust the overall stoichiometry or composition. If the Cu thickness of step a), In thickness of step b), Cu thickness of step d) and Ga thickness of step e) are selected to yield one mole of Cunlitø and two moles of CuGa₂, the overall composition of the metallic precursor would be CUuIHgCu₂Ga₄, which is equivalent to

or CuIno.e9Gao.₃1. When selenized, this provides a compound layer with Cu/(In+Ga) ratio of 1 and Ga/(Ga+In) ratio of 0.31. This is a desirable composition for high efficiency solar cell fabrication. The treatment step, f), may be skipped in the process sequence. The deposition sequence may also be altered. For example, steps d), e) and f) may be carried out first. This may than be followed by steps a), b) and optionally c). The overall precursor may then be selenized as in step g). Deposition sequences may further be altered by depositing In first Cu later and/or Ga first Cu later etc. The approach described in this example is unique in terms of utilizing two alloy compositions, one Cu-In alloy (Cui jln₉) and one Cu-Ga alloy (CuGa₂) which are stable solid phases that do not melt at temperatures of up to about 254 ⁰C, therefore yielding morphologically and compos itionally uniform metallic precursors which can be reacted with Group VIA materials to form uniform Group IBIIIAVIA compound layers. In that respect a CuGaa/CuuIng stack or a CuuW CuGa₂ stack with exemplary molar ratios mentioned above may be formed by any technique such as 006/043099

sputtering, evaporation, electroplating etc., and then exposed to Se and/or S to form a good quality compound layer. It should be noted that the interface between the Cuπlng and CuGa₂ layers of the stack is not expected to be very sharp, i.e. some degree of reaction and intermixing is expected at this interface between the two alloy phases during processing.

[00072] The Group IB and Group MA materials of the present invention may be deposited by various thin film deposition techniques such as sputtering, evaporation or wet electroless deposition. One preferred method for practicing the teachings of the present invention is the low cost electrodeposition technique, which when used in accordance with the teachings of the present invention, provides technical benefits beyond the already described economic benefits.

[00073] For the case of Cu(In₅Ga)Se₂ film growth Cu, In and Ga layers are electrodeposited at controlled thicknesses and alloying or intermixing steps of the process are used employing furnace annealing, laser, microwave or RTP (rapid thermal processing). Examples will be given below to demonstrate the use of electrodeposition for growing a Cu(In₅Ga)Se₂ film with the examplary composition of

.

[00074] Example 10: A Mo coated substrate may be used as the base. An approximately 100 nm thick Cu layer may be electrodeposited over the Mo layer. This is then followed by the electrodeposition of an about 220 nm thick In film and a nominally 40 nm thick Ga layer. The stack is annealed at a temperature of 80-200 °C preferably for 5-600 seconds to enhance alloying between Cu, In and Ga. Over the alloyed layer, 100 nm of Cu, 220 nm of In and 40 nm of Ga is then electrodeposited. The precursor is selenized by well known approaches such as in hydrogen selenide gas or selenium vapor for a time period of 5 minutes to 60 minutes to form the Cuo.₈lno.₈Gao.₂Se₁.9 compound. It should be noted that selenization may be carried out by various other means such as depositing Se over the metallic precursor and heating up the stacked layer, heating the substrate in a Se-containing gaseous or liquid atmosphere etc.

[00075] Example 11 : A Mo coated substrate may be used as the substrate. An approximately 100 nm thick Cu layer may be electrodeposited over the Mo layer. This is then followed by the electrodeposition of an about 220 nm thick In film and a nominally 40 nm thick Ga layer. The stack is annealed at a temperature of 80-200 ⁰C preferably for 2-300 seconds to enhance alloying between Cu, In and Ga. Over the alloyed layer, a nominally 100 nm of Cu, a nominally 220 nm of In and about 40 nm of Ga are then electrodeposited. A second anneal step is applied at 80-200 °C preferably for 2-300 seconds to promote further alloying between the layers of the metallic precursor. The precursor thus obtained is then selenized by well known approaches such as in hydrogen selenide or selenium vapor for a time period of 5 minutes to 60 minutes to form the Cuo.₈lno.sGaα₂Sei.₉ compound. It should be noted that selenization may be carried out by various other means such as depositing Se over the metallic precursor and heating up the stacked layer, heating the substrate in a Se-containing gaseous or liquid atmosphere etc.

[00076] The present invention has unique capabilities to allow efficient use of the low cost electrodeposition technique for the deposition of metallic components Cu, In and Ga. In general Cu, In and Ga have very different plating potentials. Therefore, to form a metallic precursor stack containing Cu, In and Ga, the Cu layer is typically electroplated first. This is then followed by deposition of all of the necessary In and then all of the necessary Ga. Otherwise, while electroplating one species, the other species on which deposition is carried out may partially dissolve into the electrolyte making compositional control very poor.

[00077] In an embodiment of the present invention Cu and at least one Group MA component are electrodeposited first on a base. A treatment step is then used to promote alloying and formation of at least one Group IBIIIA alloy and/or solid solution. Plating potentials of Group IIIA materials within a Group IBIIIA alloy and/or solid solution are shifted and the amount of Group IIIA material at the surface of the Group IBIIIA alloy and/or solid solution film is also reduced. Both of these factors allow electrodeposition of the next stack efficiently without loosing substantial Group IIIA material into the plating solution of the next step. For example, a precursor layer may be formed on a base as follows. A Cu layer is first electrodeposited on the base. This is followed by the electrodeposition of a Ga layer. The treatment step such as heat treatment forms an alloyed Cu-Ga layer. Over the Cu-Ga layer an In layer is electrodeposited. By diffusing Ga into the Cu layer and chemically tying Ga to Cu in a Cu-Ga alloy composition or solid solution, the electrode potential of Ga in the alloy is shifted compared to the electrode potential of pure Ga, and therefore electrodeposition of In over Cu-Ga alloy is achieved without much loss of Ga into the In plating solution. Furthermore, the Ga content on the surface of the Cu-Ga alloy or solid solution layer is much less than the Ga content on the surface of the Cu/Ga stack initially electroplated on the base. Therefore, Ga content available at this surface for possible removal during the In deposition step is drastically reduced by the use of the alloying step of the present invention. Another benefit of the process of the present invention is the fact that during alloying Ga diffuses into Cu and/or Cu diffuses into Ga. The result is that Ga is brought close to the interface of the metallic precursor film and the base without compromising compositional control. Bringing Ga close to the base improves the adhesion of the Group IBIIIAVIA compound layer to the base after the selenization and/or sulfidation reaction steps. This argument may be made for all alloyed layers obtained through the steps of the present invention including all Cu-In alloys and Cu-Ga alloys.

[00078] Improved adhesion to the base may also be obtained by directly plating a Cu- Ga mixed or alloy layer onto the base. In a preferred embodiment of the present invention the Cu-Ga alloy layer 25 is electrodeposited from a suitable electrolyte onto the conductor 22 as shown in Fig. 7. The substrate 23 may be a rigid or flexible conductor or insulator material such as a metal or glass sheet or a metallic or insulating foil. Metallic foils include Ti, stainless steel or Mo foils. Insulating foils include those made from high-temperature materials and polymers such as polyimides and mica. The conductor 22 comprises materials that make good ohmic contact to the absorber layer after it is fully formed. Such ohmic contact materials include Ti, Mo, W, Ta and their nitrides. The Cu-Ga alloy layer 25 may contain 5-50 atomic % Ga, preferably 10-30 atomic % Ga. The thickness of the Cu-Ga alloy layer 25 may be in the 100-500 nm range. Electrodeposition may be carried out using glycerine based electrolytes containing Ga and Cu ions. Gallium and Cu ion sources may be metal salts such as copper and gallium chlorides, which are dissolved into the glycerine solution with the help of weak acids such as citric acid or tartaric acid. Electrodeposition may be carried out at room temperature or in chilled electrolytes kept at 5-15 °C. Plating current densities may be in the range of 0.5-40 mA/cm², preferably in the range of 1-20 mA/cm². GaUium-to-Cu ratio in the electrolyte may be changed in the range of 0.5-5 to control the Ga/Cu ratio in the deposited film. Deposited layers may contain at least one of the alloy species such as CUgGa₄, Cu₃Ga₂ and CuGa₂.. In one preferred embodiment the Cu-Ga layer comprises a solid solution that can be represented by the chemical formula of Cu₍i._x)Ga_x where x is less than or equal to about 0.20. Gallium is tightly bound to Cu in this solid solution composition, which improves stability and repeatability of the process disclosed in this invention. It is important that the Cu-Ga alloy layer does not contain a substantial amount of free Ga phase. To avoid such a situation, after deposition of a Cu-Ga layer on the base 24 (Fig. 7), the structure may be annealed for a suitable period of time to assure complete alloying of Cu and Ga. For example, after electrodeposition of the Cu-Ga layer on the base, the "base/Cu-Ga layer" structure may be annealed at a temperature range of 50-500 °C, preferably at a temperature range of 100- 200 °C for a period of 1 second to 15 minutes, preferably 5 seconds to 1 minute. Annealing may be carried out in air, vacuum, inert gas or reducing atmosphere, using a furnace, oven or rapid thermal annealing system. Laser annealing, or microwave annealing may also be employed. In laser annealing the Cu-Ga layer is exposed to a large area beam of a laser such as a CO₂ laser, YAG laser or Ar laser for a short period of time, such as a few seconds, to promote alloying. [00079] After annealing, the Cu-Ga alloy layer 25 is fully formed and an In layer 26 is electroplated over the Cu-Ga alloy layer 25. Indium plating may be carried out using well- established electrolytes such as the In-sulfamate electrolyte, which is commercially available from Indium Corporation of America. The plating current density may be in the range of 10-100 mA/cm², preferably in the range of 20-50 iuA/cra" for this process. The thickness of the In layer may be in the 200-1000 ran range depending upon the thickness of the Cu-Ga alloy layer and the desired Cu/(In+Ga) and Ga/(Ga+In) molar ratios. Once the structure of Fig. 7 is obtained it may be reacted with Group VIA materials to form the compound layer as described before.

[00080] Referring back to the issue of micro-scale non-uniformities, forming a first metallic precursor sub-layer comprising at least one Group IB material and then depositing a second metallic Group IIIA-rich precursor sub-layer over the first sub-layer in a manner that provides a substantially flat surface topography for the overall metallic precursor has certain benefits. Fig. 8A shows an exemplary first metallic precursor sub-layer 83 formed on the base 82 comprising a substrate 80 and a contact film 81. The first metallic precursor sub-layer has a surface topography with thick regions 84 and thin regions 85. As described before in reference to Fig. 2b, such non-uniform surface may result from the de-wetting or balling phenomena when, especially Group IIIA-rich films are deposited over films comprising at least one Group IB material, or as seen in Fig. 3b, such surfaces may form upon heat treatment of metallic precursors comprising Group IB and Group IIIA materials. The sub-layer 83 may result from annealing of the Cu-Ga layer of Fig. 7 as described earlier. The average thickness of the sub-layer may be in the range of 200-2000 nm and the local variation in the thickness may be as large as +/- 70% of the average thickness. For example, an exemplary sub-layer with an average thickness of 600 nm may have thin areas as thin as 180 nm and thick regions as thick as 1020 nm. What ever is the origin, such non-uniformities are detrimental to micro-scale compositional uniformity of the precursor films and the compound layers obtained using them. Furthermore, as discussed before the thick regions 84 are usually richer in low melting Group IIIA materials compared to the thin regions 85. To overcome this problem Fig. 8b demonstrates a process step that deposits a second metallic Group IIIA-rich precursor sub-layer 86 over the first sub-layer 83 in a manner that levels the surface topography. This way a thicker Group IIIA-rich layer is deposited over the thin region 85 of the first sub-layer 83, and a thinner Group IIIA-rich layer is deposited over the thick region 84 of the first sub-layer 83. If the thin and thick regions were poor and rich in Group IIIA material, respectively, it can be seen that the relatively flat Group IIIA-rich layer 86 compensates for the micro-scale compositional non-uniformities and yields a more compositionally and structurally uniform precursor stack 87 that can be reacted with Group VIA materials to form a high quality Group IBIIIAVIA compound layer.

[00081] Electrodeposition and electroless deposition techniques are wet processing approaches that have unique qualities to practice the invention disclosed above. Solutions or electrolytes used for the above techniques may be formulated to obtain "leveled" deposits on rough surfaces. Various organic and inorganic additives may be utilized in such electrolytes to enhance deposition in the valleys or low surfaces on the substrate while suppressing deposition on the peaks or high points. Therefore, if a Group IIIA-rich sub-layer was electroplated on the first metallic sub-layer 83 of Fig. 8a for example, plating would be enhanced over thin regions 85 and suppressed over thick regions 84, yielding a metallic Group IIIA-rich sub-layer 86 with surface profile very similar to that shown in Fig. 8b. Many different types of additives may be used in the electrolytes or solutions to achieve the above described leveling effect. These include accelerators, inhibitors, levelers, surfactants etc. that comprise chemicals such as thiourea, polyethylene glycol, polyether sulfides, mercapto compounds, coumarin, aromatic sulfonamides, saccarin, bis-sodium sulfopropyl disulfides, high molecular weight polymers with amine or amide functional groups etc. Valleys with widths as small as 10-100 nm or even smaller may be preferentially filled and leveled with an electroplated material, such as a Group IIIA-rich sublayer, using such additives in the electroplating solution.

[00082] It should be noted that electroplating or electroless plating of a Group IIIA-rich surface film over a previously prepared substantially metallic precursor layer has the benefits of; i) covering up and adjusting the composition of any regions that may contain excess Group IB materials, which after formation of the Group IBIIIAVIA layer and solar cell devices would cause shunting and device performance reduction, ii) flattening the surface morphology of the overall precursor layer and the compound film after reaction with Group VIA materials so that better quality junctions may be formed by the deposited CdS layer or other junction forming material, iii) densification of the surface layer of the overall precursor layer so that after compound formation better junctions may be fabricated using such compound layers. These benefits are applicable to fixing problems with precursor layers that are deposited by other techniques also. In this case the electrodeposition step which applies a Group IIIA-rich layer on a previously formed precursor layer may be considered to be a surface treatment step that improves surface morphology, increases density and improves the micro-scale compositional uniformity.

[00083] For example, Fig. 9 shows a metallic precursor layer 93 that may be formed on a base 92 by a nano-particle deposition approach such as spraying or doctor-blading of a Cu, or Cu-In, or Cu-Ga, or Cu-In-Ga containing nano particle ink. Nano particles in this case may be <200 nm in size and they may be dispersed in a solvent such as water, alcohol or ethylene glycol with the aid of well known organic surfactants and dispersing agents to form an ink. The granular precursor layer 93 may be in its as-deposited form or it may have been subjected to a treatment step such as an annealing step at an elevated temperature, such as at 100-400 "C, for the purpose of at least partially fusing the nano particles to each other and to the contact film 91 of the base 92. Alternately the granular precursor layer 93, in general, may be any substantially metallic precursor layer with poor surface morphology and/or poor micro-scale compositional uniformity prepared by any method. For example the precursor layer may be obtained by first depositing on the base an ink of oxide particles comprising Cu, In and Ga to form an oxide film and then reducing the oxide film to obtain a substantially metallic film, for example containing more than 90 molar percent metallic components comprising Cu, In and Ga. As can be seen from this figure, the top surface of the granular precursor layer 93 is rough due to the granular nature of the nano particles or the heat treatment or reduction steps used to form the precursor layer. There are many gaps, which may be micron or sub-micron in size at or near the surface. As an example, the precursor layer may have a thickness of 200-2000 nm and its surface roughness maybe in the order of 50-500 nm. A Group IIIA-rich layer 94 is electrodeposited over the rough surface of the precursor layer 93 to form an overall precursor layer 96 with a surface 95 which is substantially free of roughness and porosity. For example, the surface 95 may have a roughness in the order of 5-10 nm. This is achieved because the plating solution has the capability to get into the smallest cavities or pores and then fill them up as described in relation with Fig. 8b. The precursor layer 93 may be Cu-rich, i.e. Cu/In, or Cu/Ga or Cu/(In+Ga) ratio is >1. In this case the Group IIIA- rich layer 94 may comprise In and/or Ga and may be plated out of electrolytes containing leveling or gap-filling additives. The Group IIIA-rich layer may also comprise a plurality of layers such as In and Ga layers. The thickness of the Group IIIA-rich layer is adjusted to yield an overall desired stoichiometry or composition for the overall precursor layer 96. Alternately, the precursor layer 93 may have a Cu/In or Cu/Ga or Cu/(In+Ga) ratio of 1 or <1. In this case a thinner Group IIIA-rich layer would be needed. The precursor layer 93 may also be substantially made of Cu particles. In this case In and/or Ga is electrodeposited on this rough and porous Cu layer to adjust the overall composition to a desired Cu/In, Cu/Ga or Cu/(In+Ga) ratio and at the same time fill in the pores and level the surface of the overall precursor layer for high quality compound film formation after reaction with at least one Group VIA material. [00084] Schematic of an apparatus 40 that may be used for the electrodeposition steps of the present invention above-described processing steps is shown in Fig. 10. The apparatus will be described by taking as an example electrodeposition of Cu-Ga and In. It would be obvious to those skilled in the art that the general design of the apparatus of Fig. 10 may be used to practice all electroplating related embodiments of this invention. The apparatus 40 in Fig. 10 is an in-line system comprising multiple stations. The apparatus 40 processes a substrate 40a, which may be in the form of a flexible foil or a rigid sheet. Depositions are carried out on the conductor 41, which is previously coated on one face of the substrate 40a. First a Cu-Ga layer is deposited on the conductor 41 in the Cu-Ga electroplating station 42a. Copper-Ga electroplating station 42a comprises an electroplating cell 45, with an enclosure 52 that receives the electroplating solution through inlet 50 and delivers it onto the surface of the conductor 41 through opening 51. Electroplating solution flows in the direction shown by arrows 48. An anode 47 is placed into the enclosure 52, which is made of an insulating material such as polypropylene. The anode 47 may be made of an inert material such as Pt or Pt-coated Ti, or it may be a Cu or Cu-Ga alloy anode. The anode 47 may have pores or holes in it to allow the plating solution to flow through. Electrical contacts 46 are provided to softly touch the surface of the conductor 41. It should be noted that if the process is carried out in the in-line mode, the substrate 40a continuously moves in the direction "P" as the Cu-Ga layer is deposited. Therefore, while the first contact 46a touches the surface of the conductor 41 before plating, the second contact 46b touches the surface of the Cu-Ga layer that has just been deposited through the opening 51 onto the conductor 41. In that respect, scratching of the Cu-Ga surface by the second contact 46b should be avoided by using lightly touching, spring loaded or roller type contacts. It is also possible to use only one set of contacts (first contact 46a) eliminating the second contact 46b. Although two contacts are schematically shown in Fig. 4, it should be understood that any number of contact may be used as long as they are outside the area defined by the opening 51. If the substrate 40a is conductive, then electrical contacts may be made on the back surface 40b, eliminating any concerns about scratching the deposited layers.

[00085] During electrodeposition, a voltage is applied between the contacts 46 and the anode 47 to render the contacts and therefore the conductive surface of the substrate more cathodic. This causes deposition to take place on the conductive surface of the substrate. Plating solution is directed through the opening 51 onto the surface of the conductor 41 and then flows towards outside of the enclosure 52 to be collected, refurbished and re-circulated. The opening 51 may be in the form of a near-rectangular slit. The plating current density employed, the width of the slit in the direction P, and the speed of the substrate motion utilized determine the thickness of the Cu-Ga layer obtained on a portion of the substrate traveling above the opening 51. The length of the slit (in a direction perpendicular to P) determines how large a substrate is processed and the throughput of the apparatus 40. The width of the opening 51 or slit may be in the 1-10 cm range, whereas, its length may be in the range of 30-120 cm.

[00086] As the surface of the conductor 41 is coated with the Cu-Ga layer, it moves to a rinse/dry station 43 where the surface of the Cu-Ga layer is rinsed and cleaned off any chemical residues. After rinsing, the surface may be dried by blowing air or nitrogen over it. After rinsing and drying, the portion of the surface already coated with Cu-Ga moves into the anneal station 44. As described before, use of an anneal station 44 is optional but it is preferable to assure good compositional control. If the electrodeposition step carried out in the Cu-Ga electroplating station 42a yields a Cu-Ga layer that is fully alloyed, there may not be a need for an anneal station 44. In the anneal station 44 the freshly deposited Cu-Ga layer is exposed to heat originating from a heat source 55. The heat source 55 may be a resistive heating element, a bank of heating lamps, a laser beam etc. as described before. The Cu-Ga layer is annealed in the anneal station 44 to assure substantial alloying between Cu and Ga and formation of a Cu-Ga alloy layer. Once the Cu-Ga alloy layer is formed on the surface of the conductor 41, it moves into the In electroplating station 42b for the deposition of an In layer to obtain the precursor structure such as the one shown in Figs. 7 or 8b. The In electroplating station may be very similar to the Cu-Ga electroplating station and therefore its details have not been shown in Fig. 10. The anode in this case may be an inert anode or an In anode. The width of the opening, and the plating current density are selected to yield the desired In layer thickness and the desired overall Cu/(In+Ga) and Ga/(Ga+In) ratios. Since the preferred mode of operation for the apparatus 40 in Fig. 10 is "inline", the speed of the substrate 40a in direction P is the same for all steps that are carried out in a serial mode. Therefore, thickness of various layers electroplated may be controlled by the plating current densities employed at each process station. It should be understood that various stages of the present invention may also be carried out in separate pieces of equipment. For example, Cu- Ga electroplating and annealing may be carried out in one apparatus and In electroplating may be carried out in a separate apparatus. Although four process stations are shown in Fig. 10, many numbers of process stations may be added to the apparatus of Fig. 10. For example, multiple Cu- Ga electroplating stations and In electroplating stations, and anneal stations may be employed to increase throughput. To practice the inventions described through Figs. 6a-6d, two or more units shown in Fig. 10 may be added in series. Even selenization/sulfidation stations may be added at the end to react the freshly deposited Cu-Ga/In precursor stacks with Group VIA materials to form the compound layers as will be described later.

[00087] In another prefeiτed embodiment of the present invention a Cu-Ga alloy layer is formed by reacting a Cu layer with a Ga-bearing layer. Fig. I Ia shows a base 240 coated with a Cu layer 300. The Cu layer 300 is preferably deposited on the conductor 220 through electrodeposition, although other well-known film deposition techniques may also be used. After deposition of the Cu layer 300, a Ga-bearing layer 310 is deposited over the Cu layer 300 as shown in Fig. 1 Ib. The Ga-bearing layer is preferably a Ga layer but may also comprise a Ga-In alloy. The Ga-bearing layer is preferably deposited using electroplating, although other thin film deposition techniques may also be used. Since Ga and Ga-In alloys melt at low temperatures of <156 °C, techniques such as melt spraying or dipping may also be utilized for deposition of these materials. In dipping techniques the substrate is dipped into and withdrawn from a melt which may be a melt of Ga or Ga and In. In this case a small amount of Cu (1-10%) may also be included in the melt to avoid leaching of the Cu from the electrodeposited Cu layer 300 into the melt. Once the stack of Cu layer 300 and the Ga-bearing layer 310 are formed as shown in Fig. 1 Ib, the stack is annealed as described before to form a Cu-Ga alloy layer 320 as shown in Fig. 1 Ic. The Cu-Ga alloy layer may also contain some In in the form of elemental In, or Cu-In alloys or In-Ga alloys, if the Ga-bearing layer 310 contains In. Atomic % of In in the Ga-bearing layer is preferably in the range of 0-20%. Thus the melting point of the Ga-bearing layer is preferably below 30 ⁰C. After the annealing step and formation of the Cu-Ga alloy layer 320, an In layer is deposited, preferably by electrodeposition, over the Cu-Ga layer to obtain a structure similar to the one in Fig. 7. A preferred process flow of this invention which uses low-cost, large area deposition approaches is; a) electrodeposition of a Cu layer on a base, b) electrodeposition of a Ga layer over the Cu layer, c) annealing of the Cu/Ga stack to form a Cu-Ga alloy layer, and d) electrodeposition of an In layer over the Cu-Ga alloy layer to form a precursor layer as shown in Fig. 7. The apparatus of Fig. 10 may be easily configured to carry out these process steps. Electrodeposition of In from electrolytes with leveling ability addresses any roughness and micro-scale compositional non-uniformity issues that may be present in the Cu-Ga alloy layer as described before in relation with Figs. 8a and 8b.

[00088] Reaction of metallic precursors with Group VIA materials may be achieved various ways. In one embodiment the precursor layer is exposed to Group VIA vapors at elevated temperatures. These techniques are well known in the field and they involve heating the precursor layer to a temperature range of 350-600 °C in the presence of at least one of Se vapors, S vapors, and Te vapors provided by sources such as solid Se, solid S, solid Te, H₂Se gas, H₂S gas etc. for periods ranging from 5 minutes to 1 hour. In another embodiment a layer or multi layers of Group VIA materials are deposited on the precursor layer and the stacked layers are then heated up in a furnace or in a rapid thermal annealing furnace and like. Group VIA materials may be evaporated on, sputtered on or plated on the precursor layer. Alternately inks comprising Group VIA nano particles may be prepared and these inks may be deposited on the precursor layers to form a Group VIA material layer comprising Group VIA nano particles. Dipping, spraying, doctor-blading or ink writing techniques may be employed to deposit such layers. Reaction may be carried out at elevated temperatures for times ranging from I minute to 30 minutes depending upon the temperature. As a result of reaction, the Group IBIIIAVIA compound is formed from the precursor. It should be noted that reaction chambers may be added to the apparatus of Fig.10 to carry out the whole process in-line.

[00089] All the embodiments of the present invention involving the electrodeposition and/or electroless deposition technique may be carried out of solutions with leveling additives as mentioned before. There are other features that may be used to effectively level the deposits obtained by the electroless or electrodeposition of layers comprising Cu and/or In and/or Ga . One of these approaches is the use of viscous electrolytes. In this case the viscosity of the electrolyte is increased using additives such as thickening agents available from companies such as Rohm and Haas. Such solutions or electrolytes have viscosities at least 1.10 times that of water. The viscosity of such electrolytes may further be adjusted by lowering the temperature to below 20 C. Lowering the temperature of the plating solutions also minimizes the chance of Ga melting and causing micro scale non-uniformities. Another approach to effectively level the depositing layers comprising Cu, and/or In, and/or Ga of the present invention is the use of a brush that brushes the surface of the depositing film. The brush may be a soft cloth, a napped polyester or polypro piece etc. that can be mounted at the nozzles of the apparatus shown in Fig. 10 or any other apparatus that may be used. In brush plating the diffusion layer on the surface of the depositing film is minimized due to the sweeping action of the brush and thus the rate of deposition increases. Additionally, depending on the hardness of the selected brush the surface of the depositing film may be buffed and highly leveled and shiny films may be obtained. Depositing Ga or Ga-In containing layers by brush plating it is preferable to reduce the electrolyte temperature to below 20 C, preferably below 15 C to avoid any melting and smearing of the low melting compositions such as the In-Ga eutectic composition of about 15% In, which has a melting point of about 15 C. Especially for the deposition of Cu-Ga, Cu-In, and In-Ga alloy or mixture films, where deposition potentials of the depositing species are different from each other, use of brush plating and/or viscous or gel electrolytes, and/or low temperature processing improve the compositional control of the depositing material. Acoustic or megasonic agitation in electrolytes may also be used for efficient mass transfer during electroless deposition or electrodeposition.

[00090] In previous paragraphs we described various approaches to address the micro- scale non-uniformities that may form in Cu(In,Ga)(S,Se)₂ thin films grown by the two-step or multi-step processes. One other factor that plays an important role in the formation of such non- uniformities is the poor adhesion between the ohmic contact material and the precursor layer deposited over it. For example, the most commonly used ohmic contact materials in the solar cell structure of Fig. 1 is Mo. When a metallic (such as Cu and/or In and/or Ga) or mostly metallic precursor layer is deposited over the Mo film the adhesion is generally poor because of the presence of an oxide layer on the Mo and/or inertness of the Mo. Poor adhesion aggravates the problems of de- wetting and balling that have been described before. This problem may be addressed by depositing over the contact layer (i.e. conductive layer 13 in Fig. 1) a nucleation film that allows better nucleation and consequently better adhesion. Nucleation layers include Rhodium, Ruthenium Palladium and Osmium. Copper, for example, nucleates much better on a Ru layer than on a Mo, Ti or Ta layer. Specifically for wet deposited, i.e. electroplated or electroless plated copper adhesion to the base may be greatly improved through the use of nucleation layers. Therefore, in processing the device structure first a contact layer such as Mo is deposited on the substrate. Then a nucleation layer such as Ru is deposited. This is then followed by the deposition of a precursor layer comprising Cu and/or In and/or Ga and optionally Se. Reaction is then carried out for the formation of the Cu(In,Ga)(S,Se)₂ compound layer. The nucleation layer may have a thickness of l-20nm, preferably 2-6 nm and it may be deposited by techniques such as electroless deposition, electroplating, atomic layer deposition, CVD, MOCVD, and PVD among others. Thin nucleation layers are preferred for cost reduction.

[00091] For the case of depositing a precursor film through a wet process such as electroplating or electroless plating, a seed layer may be used in place of or on top of the nucleation layer. For example, if a Cu/In/Ga stack or a Cu-Ga/In stack is electroplated or electroless plated over the contact layer such as a Mo layer, first a seed layer may be deposited over the Mo layer or Mo/nucleation layer stack. The seed layer then acts as an adhesion/nucleation layer on which the electroplated metal bonds well. The seed layer maybe a Cu layer or a layer that comprises Cu. For example, a sputtered, CVD deposited, or ALD deposited Cu film of 2-100 ran thickness acts as an efficient seed layer upon which the precursors comprising at least one of Cu, In and Ga may be deposited with good adhesion and morphology. The seed layers described above also improve adhesion and uniformity of Cu(In,Ga)(S,Se)₂ layr deposited by techniques other that electrodeposition.

[00092] One unique type of seed layer that is especially suited for improving adhesion and reducing or eliminating non-uniformities due to de-wetting or balling effects is a seed layer comprising polymeric material. In this approach the conductive seed layer material such as Cu is deposited over the contact material such as Mo along with a small percentage of polymeric material. The percentage of the polymer may be less than 1 volume %, preferably less than 0.1 volume %. One method to obtain such films or coatings is electrografting which yields uniform conductive coatings with a thickness of 2-20 nm. The conductive material in electrografted layers may be selected from a wide variety of metals. However, for the present application the conductive material may be selected from a group that is compatible with the Cu(In, Ga)(S₁Se)₂ compound, for example Cu, In, Ga, Te, Se, Co, Mo, Ni, C, Ru, Pd etc. The preferred conductive material is Cu. Electrografting is commonly used to deposit polymeric materials or electrically conductive particles along with polymeric materials on conductive substrates (see e.g. U.S. Patent 6,375,821). Continuous copper layers with <0.1% polymeric content may be obtained by including copper ions into the electrografting solution. Using cathodic deposition, copper is then deposited along with the polymer on the contact layer forming a thin seed layer with extremely good adhesion to the contact layer such as a Mo layer. When a precursor comprising Cu is electrodeposited or electroless deposited over this seed layer the adhesion and wetting is also good. Polymeric material may desorb and leave the material when high temperature annealing or reaction carried out, or it may stay at the contact layer interface. This does not interfere with the efficiency of the solar cells because the contact interface is far from the active junction of the device which is near the Cu(In,Ga)(S,Se)₂/transparent layer interface.

[00093] Solar cells may be fabricated on the compound layers of the present invention using materials and methods well known in the field. For example a thin (<0.1 microns) CdS layer may be deposited on the surface of the compound layer using the chemical dip method. A transparent window of ZnO may be deposited over the CdS layer using MOCVD or sputtering techniques. A metallic finger pattern is optionally deposited over the ZnO to complete the solar cell.

[00094] Once the solar cells are fabricated they are interconnected to fabricate modules with the desired voltage and current ratings. Interconnection may be done externally using wires or ribbons connecting the negative terminal of one solar cell to the positive terminal of the next to increase the voltage output (series connection) and/or connecting the like terminals to each other to increase the current output (parallel connection). Alternately, in thin film solar cells monolithic integration approach may be used to fabricate module structures. Figs. 12a- 12E show a monolithic integration technique. Fig. 12a shows a base comprising a substrate 400 and a contact layer 401 or conductive layer. The substrate 400 is insulating or it comprises an insulating portion such as an insulating layer at its top surface where the contact layer is deposited. As described before the contact layer forms the ohmic contact to the finished device. The contact layer 401 is then scribed using etching, laser scribing, mechanical scribing etc. forming isolation gaps or scribe lines 402 which isolate the contact layer pads 401 a (see Fig. 12b). The Cu(In,Ga)(S,Se)₂ compound layer 403 is deposited over the contact layer pads 401a (Fig.12c) and then scribed again forming openings 404 in the compound layer 403 forming compound layer islands 403a as shown in Fig. 12d. A transparent layer 405 such as a ZnO layer or a CdS/ZnO stack is deposited over the compound layer islands 403a and establishes physical and electrical contact with the contact layer pads 401a through the openings 404. The last step is shown in Fig. 12e and it involves forming cuts 406 in the transparent layer 405 forming transparent layer strips 405a. The cuts 406 may extend into and even go through the compound layer islands 403a. The integration scheme of Fig. 12 interconnects the positive terminal (contact layer pads 401a) of each cell to the negative terminal (transparent layer strips 405a) of the adj ascent cell. This way series connection and voltage addition is achieved.

[00095] The integrated structure of Fig. I2e has a leakage path through the scribe lines 402 between the contact layer pads 401a. This is because the gap defined by the scribe lines 402 is filled with the Cu(In₅Ga)(S₅Se)₂ compound which may have high or low resistivity depending upon its composition, especially the Cu/(In+Ga) ratio. The consequence of this is the fact that as the Cu/(In+Ga) ratio increases the resistivity of the compound decreases and the leakage between the contact layer pads increases reducing device efficiency.

[00096] To address this problem a reactive material layer 500 may be deposited in between the contact layer pads 401a as shown in Fig. 13a. Then the Cu(In₁Ga)(S₁Se)₂ compound layer is formed over this structure. After the compound formation a structure 501 (see Fig. 13b) is obtained where sections 503 of the compound layer 403 has higher resistivity compared to sections 502 of the compound layer due to the reaction/intermixing of the reactive material layer 500 with the compound at sections 503. This way the resistivity of the active solar cell sections 502 is adjusted for best solar cell efficiency and the resistivity of the sections 503 are adjusted to be high, for example higher than about 1000 ohm-cm, preferably higher than 10000 ohm-cm, to minimize leakage currents. The reactive material layer 500 when reacted and intermixed with Cu, In, Ga, Se and/or S, forms a higher resistivity compound layer. Examples of the reactive material layer are layers comprising In and/or Ga and/or Al. The reactive material layer may comprise Group HIA materials in their elemental or alloy forms or it may comprise Group IHA material oxides, sulfides, tellurides or selenides. For example, the ractive material layer may comprise In metal, Ga metal, Ga-In alloy, In-oxide, Ga-oxide, Ga-In-oxide, In-selenide, Ga-selenide, Group IIIA-sulfides etc. It should be noted that the reactive material layer 500 may be highly conductive in its as deposited form. However, it must be reactive, e.g. it must be selenizable or sulfidizable, and after the reaction it must form a high resistivity region between the contact layer pads 401a. If a Group IHA material layer such as a Ga layer, an In layer or a Ga-In layer is used as the reactive material layer 500 and then a Cu(In, Ga)(S₅Se)₂ layer is formed over it with a predetermined Cu/(In+Ga) ratio value, then after the reaction and compound film formation, the Cu/(Ga+In) ratio in sections 502 would be substantially equal to the predetermined value whereas this ratio would be lower in sections 503 due to the additional amount of Group HIA material present in these sections. It should be noted that the reactive material layer 500 may be deposited even over the contact layer pads 401a as shown in Fig. 13c. The important point, however, is that the thickness of the reactive material layer is smaller over the contact layer pads 401a then within the scribe lines 402 in between the contact layer pads 401a. Therefore, when a compound layer is formed, resistivity is increased more in regions (within scribe lines 402) where the reactive material layer is thicker. It is also possible that the reactive material just fills the scribe lines without extending over the contact layer pads as shown by dotted line 510 in Fig. 13c.

[00097] Reactive material layers may be deposited by various means such as electroless plating, ink writing, physical vapor deposition, chemical vapor deposition etc. Masking methods may be used to deposit them selectively into the scribe lines. Alternately, selective seeding techniques may be utilized within the scribe lines to initiate reactive material layer deposition selectively in these regions. If the substrate is transparent, radiation may be shone from the substrate side and radiation passes through the scribe lines but is not allowed to pass through the contact layer pads. If a deposition approach that is enhanced or initiated by radiation is utilized for the deposition of the reactive material layer, then deposition initiates and takes place substantially only within scribe lines where radiation is present. It should be noted that the reactive material layers do not need to be continuous layers. They may be in the form of islands. 006/043099

They may have a thickness in the range of 1-1000 nm, preferably 5-500 nm. One way of depositing a reactive material layer comprising low melting metals such as Ga and/or In is to uniformly coat a melt of these metals over the structure shown in Fig. 12b. By spinning or physically buffing the surface of the contact layer pads 401a and then cooling down the substrate, a reactive material layer structure such as the ones shown in Fig. 13a or Fig. 13c may be obtained. Ga-In alloys with melting points lower than 30 °C is well suited for this approach.

[00098] ^" Another method to introduce a reactive material into the scribe lines 402 is shown in Fig. 13d. In this case a layer of reactive material layer 500 is deposited first over the substrate 400 and then the contact layer pads 401a are formed such that the reactive material layer 500 is exposed through the scribe lines 402. When a compound layer is formed over this structure, reactive material exposed through the scribe lines react with the compound layer increasing its resistivity as described before. It is also possible that, the structure of Fig. 13d is put through a pretreatment step before the compound layer is formed over it. For example, if the reactive material layer is In-oxide or In-Tin-oxide, the structure of Fig. 13d may first be annealed in a reducing atmosphere or exposed to a reducing chemical solution with or without an applied voltage, to convert the portions of the reactive material layer exposed through the scribe lines into a layer comprising In metal. Then during compound layer formation (not shown) In reacts efficienctly with the froming compound layer. Alternately, if the pretreatment step is skipped, resistivity increase is obtained by reaction of In-oxide with the forming compound layer. The decision of whether or not a pre-treatment step is needed should be made taking into consideration details of the compound formation process. For example if compound formation involves Cu, In and/or Ga deposition followed by hydride selenization involving H₂Se and/or H₂S gases, then there may not be a need for a pre-treatment step because these gases efficiently reduce In-oxide to In during reaction. If the process involves selenization and/or sulfidation using Se and/or S vapors, a pre-treatment step may be used to activate the In in the In-oxide or In-tin- oxide layer portions within the scribe lines.

[00099] Formation of higher resistivity compound layer within isolation gaps or scribe lines may also be achieved through an approach depicted in Figs. 17a- 17c. This approach is especially suited for two-step or multi-step processing approaches. For example, in Fig. 17a, a first precursor layer 700 is deposited substantially only on the contact layer pads 401a. A second precursor layer 701 is then deposited over whole surface. Group IB/IIIA molar ratio in the first precursor layer 700 is larger than that in the second precursor layer 701. Consequently, after the precursors are annealed/reacted to promote intermixing a compound layer 403 as shown in Fig. 17c, wherein the Group IB/IIIA molar ratio in regions 706 is smaller than in regions 705. For example, the first precursor layer 700 may be a Cu-Ga alloy layer and the second precursor layer 701 may be an In layer. If the relative thicknesses of the first and second precursor layers are selected properly, after the compound formation step involving reaction of the precursor stack with Se, the compound layer in region 705 may be a layer of Cu(In₁Ga)Se₂ with Cu/(In+Ga) ratio of less than 1.0, wherein the compound in regions 706 would be substantially In-selenide with a Cu/(In+Ga) ratio of zero. In-selenide is a very high resistivity material. It should be appreciated that the In layer of this example may be replaced with Ga-In or even Cu-Ga-In and the invention may still be practiced as long as the Cu/(In+Ga) ratio in the first precursor layer 700 is larger than the Cu?(In+Ga) ratio in the second precursor layer 701. It is also possible that at least one of the first precursor layer 700 and second precursor layer 701 comprise at least one Group VIA material such as Se, S and Te. The order of deposition for the first precursor layer 700 and the second precursor layer 701 may also be reversed as shown in Fig. 17b. After the annealing/reaction step, the structure in Fig. 17b may also be converted into the structure shown in Fig. 17c.

[000100] The monolithic integration method shown in Figs. 12a-12e is not applicable to techniques involving electrodeposition for the formation of the compound layers. As shown in Fig. 12a the substrate 400 is insulating. Therefore, when scribe lines 402 are formed (see Fig. 12b), an insulating surface is exposed through the scribe lines 402. If the formation process for the compound layer 403 involves electrodeposition (such as electrodeposition of a precursor layer followed by a reaction step to form the compound) then no deposition takes place on the insulating surfaces since electrodeposition selectively goes on the contact layer pads 401a. This yields a structure such as the one shown in Fig. 14a where the compound layer 403 is in the form of separated pads. To integrate this structure, two approaches may be taken. If at least one of the left side wall 401c or right sidewall 401b of the contact layer pads 401a are well covered by the compound layer 403 (see Fig. 14b), then the openings 404 are formed, a transparent layer 405 such as a ZnO layer or a CdS/ZnO stack is deposited, and cuts 406 in the transparent layer 405 are opened. It should be noted that, although both left sidewall 401c and right sidewall 401b of the contact layer pads 401a are shown to be covered by the compound layer 403 in Fig. 14b, this integration approach would work as well if only the right side walls 40 Ib of the contact layer pads 401a were covered by the compound layer 403. In this case the openings 404 would be shifted closer to even over the left sidewalls 401c of the contact layer pads, minimizing loss of device area to the formation of openings 404. [000101] If the coverage of the side walls of the contact layer pads by the compound layer is not good, then insulating plugs 600 may be formed within the scribe lines 402 as shown in Fig. 14c. Then the structure in Fig. 14c is used just like the structure in Fig. 12c, to carry out the integration steps shown in Figs. 12d and 12e. The insulating plugs 600 may be organic or inorganic and may be deposited by techniques such as ink writing, physical vapor deposition, screen printing etc. One unique way of obtaining insulating plugs 600 within scribe lines 402 is depositing a photosensitive material such as photoresist over the structure of Fig. 14 and then exposing this photosensitive material to light so that photosensitive material within the scribe lnes 402 cures forming the plugs and un-exposed portions of the photosensitive material is washed away by chemicals. The exposure of the photosensitive material may be done by laser beam(s) aimed at the scribe lines, or illumination (especially UV illumination) through a mask etc. For substrates that are transparent illumination may be done through the substrate. This way the contact layer pads 401a and compound layer 403 act as a natural mask for illumination. Light passes only through the scribe lines 402 and cures the photosensitive material there. After chemical washing, a structure is formed similar to that shown in Fig. 14c where the insulating plugs 600 comprises the photosensitive material.

[000102] The monolithic integration approach of Figss 12a-12e may also be used for compound layers formed using an electrodeposition approach provided at least one process step is added to the process flow. For example, benefits of using a nucleation or seed layer for adhesion improvement and uniformity improvement have already been discussed. Use of a seed layer also enables the employment of the integration approach of Figs. 12a-12e. In this case a seed layer 605 is deposited over the structure with contact layer pads 40la as shown in Fig. 15a. The seed layer 605 covers the surface of the contact layer pads 401a as well as the surface of the substrate 400 exposed through the scribe lines 402. The seed layer is conductive, therefore, enabling electrodeposition even within the scribe lines. After the structure of Fig. 15a is obtained a precursor layer 606 comprising Cu and/or In and/or Ga and or Se and/or S may be electrodeposited over the seed layer (Fig. 15b). The precursor layer 606 may then be annealed and/or reacted to form a Cu(In₅Ga)(S₁Se)₂ compound layer, in effect forming the structure shown in Fig. 12c. The integration steps shown in Figs. 12d and 12e may then be carried out. It should be noted that various precursor deposition approaches and various precursor compositions described throughout this manuscript may be used in the process step shown in Fig. 15b. Even if the precursor layer 606 is deposited by a technique other than electrodeposition (such as sputtering, evaporation, nano-particle deposition etc), use of a seed layer may still be desired T/US2006/043099

because it improves the crystallinity, morphology, adhesion and uniformity of the compound layer grown, especially within the scribe lines. Compound layers grown on insulating substrate surfaces such as glass surfaces often lack adhesion and uniformity and they may give rise leakage between the contact layer pads as described before.

[000103] The nucleation or seed layer 605 shown in Fig. 15a may be a layer that reacts with the compound layer as the compound layer is formed. A thin Cu layer, for example, may be used to render the scribe lines conductive (Fig. 15a), but then, after the formation of the compound layer this thin Cu layer becomes a part of the compound layer itself. This way no un- reacted conductive leakage path is left between the contact layer pads after the formation of the compound layer. In this respect, materials comprising the compound layer components [i.e. Cu, In, Ga, S, Se, Te, Al, for the case of Cu(In, Ga,Al)(S,Se,Te)₂ compound growth] are good candidates for seed layer because they ready react and be part of the compound layer.

Electrodeposition of precursor layers into the scribe lines may also be achieved if a conductive material is formed selectively within the scribe lines. Structures shown in Figs. 13a, 13c and 13d are suitable for this purpose. For the case of Fig. 13d for example, if the reactive material layer 500 is conductive, then precursor electrodeposition may be readily carried out over the contact layer pads 401a as well as on the surface of the reactive material layer 500 exposed through the scribe lines 402. After the formation of the compound layer a structure similar to that shown in Fig. 16 may be obtained. As can be seen from Fig. 16 the reactive material layer 500 may be substantially unreacted in regions underneath the contact layer pads 401a because of protection by the contact pads 401a, but it has reacted and intermixed with the compound layer 403 within and over the scribe lines 402. If the material of the reactive material layer is selected for increasing resistivity of the compound layer then a structure such as the one shown in Fig. 13b. It should be noted that unreacted portions of the reactive material layer under the contact layer pads are not shown in Fig. 13b.

[000104] The structure in Fig. 13d may be obtained by various means such as using masks during deposition of the contact layer pads (i.e. depositing the contact layer in the form of pads rather than depositing a blanket contact layer and then masking and etching or scribing to form the pads), depositing a blanket contact layer and then scribing, masking and etching etc. to form the pads. One unique way of achieving the structure of Fig. 13d is to use a reactive material layer 500 comprising a substance that is substantially transparent to a wavelength. A blanket contact layer may then be deposited over the reactive material layer and laser scribing may be carried out either from the side of the contact layer or the side of the substrate employing the wavelength. This way, the contact layer is scribed but the reactive material layer stays substantially intact.

[000105] Another approach to form contact layer pads such as those shown in Figs. 12b and 13d is forming anti-nucleation regions on the surface that the contact layer is deposited, the anti-nucleation regions corresponding to the scribe line locations. This way, when a blanket contact layer is being deposited, the contact layer does not nucleate and form at scribe line locations, i.e. it deposits in the form of pads only on regions where there is no anti-nucleation agents. Anti-nucleation agents may comprise organic and inorganic materials such as various waxes, and liquid or solid lubricants etc.

[000106] Fig. 18 shows a sequence of process steps that may be employed to achieve monolithic integration for compound layer growth techniques involving reaction of precursors. In Fig. 12c a compound layer 403 is first formed and then scribing and deposition sequences shown in Figs. 12d and I2e are carried out. In Fig. 18a a precursor layer 800 is first deposited over the contact layer pads 401a. The precursor layer 800 is then removed or partially removed from locations 801 by mechanical scribing, laser scribing, etching, electroetching etc. This is followed by the annealing/reaction step resulting in the structure shown in Fig. 18b where the compound layer 802 grew in the form of separated pads. The structure of the compound layer 802 in Fig. 18b is equivalent to the structure of the compound layer 403a shown in Fig. 12d and may be used in an integration process flow discussed in reference with Figs. 12d and 12e. The precursor layer 800 may comprise Cu, In and/or Ga and optionally at least one of Se, S and Te. If the deposition of the precursor layer 800 comprises more than one precursor deposition steps, the scribing or removing from locations 801 may be carried out any time before the last precursor deposition step. For example, in Fig. 18c, a first precursor layer 803 is deposited in the form of isolated pads or it is first deposited and then scribed forming isolation scribes 804. Then a second precursor layer 805 is deposited as shown in Fig. ISd. After reaction/annealing a compound layer 806 is formed with regions 807. The composition of regions 807 is different than the composition of the compound layer 806 and it may be controlled by selecting the nature of the first precursor layer 803 and the second precursor layer 805. For example, if the first precursor layer 803 is rich in Group HIA material(s) and the second precursor layer is rich in Group IB material, then the Group IB/IIIA molar ratio at regions 807 is larger that the Group IB/IIIA molar ratio anywhere else in the compound layer 806. High Group IB/IIIA molar ratio reduces resistivity at regions 807. Therefore, a transparent conductive stack 808 may be deposited over the compound layer 806 and the regions 807. After forming isolation gaps 809 in transparent conductive stack 808, monolithic integration is achieved. It should be noted that this approach may be used in conjunction with the previously discussed approaches to form high resistivity regions at the scribe lines 402, resulting in a structure where the electrical current leakage between contact layer pads is minimized and the isolated sections of the transparent conductive stack are shorted to the contact layer pads through the low resistivity at regions 807.

[000107] We have discussed monolithic integration approaches where shaping of the compound layers (through use of scribing, etching, electroetching etc.) are done after the compound layers are formed. We also discussed the cases where the shaping may done after a precursor deposition step and then the shaped precursor is converted into a shaped compound layer. Approaches in between these cases may also be used where a precursor may be deposited, partially reacted, shaped, and then fully reacted to form a shaped compound layer. Partial reaction may, for example, involve a low temperature (<400 C) reaction with or without presence of a Group VIA material. Full reaction step may involve a high temperature reaction at a temperature range of 400-600 C.

[000108] Although the present invention is described with respect to certain preferred embodiments, modifications thereto will be apparent to those skilled in the art.

Claims

1. A method of fabricating a Cu(In₅Ga)(S₁Se)₂ compound layer on a contact layer of a base comprising, depositing a nucleation layer on the contact layer, and forming through a wet processing technique a precursor film over the nucleation layer.

2. The method of claim 1 wherein the wet processing technique is one of electroplating and electroless plating.

3. The method of claim 2 wherein the precursor film comprises copper and at least one of indium and gallium.

4. The method of claim 3 wherein the nucleation layer comprises at least one of ruthenium, rhodium, palladium and osmium.

5. The method of claim 4 further comprising the step of reacting the precursor film with at least one of Se and S.

6. The method of claim 3 further comprising the step of reacting the precursor film with at least one of Se and S.

7. The method of claim wherein the step of reacting comprises annealing the precursor film in presence of gaseous species comprising at least one of S an Se.

8. The method of claim 1 wherein the nucleation layer comprises at least one of ruthenium, rhodium, palladium and osmium.

9. The method of claim 1 further comprising the step of reacting the precursor film with at least one of Se and S.

10. The method of claim 9 herein said annealing is carried out at a temperature of higher than or equal to 350 ⁰C.

1 1. A method of fabricating a Cu(In₅Ga)(S, Se)₂ compound layer on a contact layer of a base comprising: depositing a nucleation layer on the contact layer, electroplating a precursor film over the nucleation layer, and forming a Group VIA layer over the precursor film.

12. The method of claim 11 wherein the precursor film comprises copper and at least one of indium and gallium and the Group VIA layer comprises at least one of S and Se.

13. The method of claim 12 wherein the nucleation layer comprises at least one of ruthenium, rhodium, palladium and osmium.

14. The method of claim 12 further comprising the step of reacting the precursor film with the Group VIA layer through annealing at a temperature higher than or equal to 350⁰C.

15. The method of claim 14 wherein reacting is carried out in an atmosphere comprising gaseous Group VIA material species.

16. The method of claim 13 further comprising the step of reacting the precursor film with the Group VIA layer through annealing at a temperature higher than or equal to 35O°C.

17. The method of claim 16 wherein reacting is carried out in an atmosphere comprising gaseous Group VIA material species.

18. A method of forming a thin film solar cell having a Group IBIIIAVIA absorber layer with microscale uniformities comprising the steps of: applying, over a sheet-shaped base, a thin Ruthenium film; depositing a precursor in at least one layer, the precursor including a Group IB material and at least one Group IIIA material, the precursor as-deposited having a Group IB to Group IIIA molar ratio of less than or equal to 1.0, and wherein the thin Ruthenium film provides for microscale uniformities of the precursor; reacting the precursor with at least one Group VIA material to form the Group IBIIIAVIA absorber layer; and depositing a transparent layer over the Group IB IHAVIA absorber layer.

19. The method according to claim 18 wherein the step of applying applies the thin Ruthenium film over a base having a Mo surface.

20. The method according to claim 18 wherein the step of depositing comprises a wet process.

21. The method according to claim 20 wherein the wet process is electroplating.

22. The method according to claim 18 wherein the Group IB material is Cu.

23. The method according to claim 22 wherein the Group IIIA material is one of In and Ga.

24. The method according to claim 22 wherein the Group IIIA material includes both In and Ga

25. The method according to claim 23, wherein Group VIA material is one of Se and S.

26. The method according to claim 23, wherein the Group VIA material includes both Se and S.

27. The method according to claim 24, wherein the Group VIA material is one of Se and S.

28. The method according to claim 24, wherein the Group VIA material includes both Se and S.

29. The method according to claim 21 wherein the Group IB material is Cu, Group IIIA material is at least one of In and Ga, and Group VIA material is Se.

30. The method according to claim 18 wherein the step of applying is at least one of electroplating, electroless plating, atomic layer deposition, chemical vapor deposition, metal- organic chemical vapor deposition and physical vapor deposition.

31. The method according to claim 30 wherein the step of applying is electroplating.

32. The method according to claim 29 wherein the step of applying is at least one of electroplating, electroless plating, atomic layer deposition, chemical vapor deposition, metal- organic chemical vapor deposition and physical vapor deposition.

33. The method according to claim 32 wherein the step of applying is electroplating.

34. The method according to claim 18 wherein the step of applying applies the thin Ruthenium film over a base that includes a conductive foil.

35. The method according to claim 18 wherein the step of applying applies the thin Ruthenium film over a base that includes a flat insulating material.

36. The method according to claim 18 wherein the thickness of the thin Ruthenium film is in the range of 1-20 nm.

37. The method according to claim 29 wherein the thickness of the thin Ruthenium film is in the range of 1-20 nm.

38. The method according to claim 33 wherein the thickness of the thin Ruthenium film is in the range of 1-20 nm.

39. The method according to claim 18 wherein the step of reacting comprises heating the precursor in a gas containing at least one Group VIA material.

40. The method according to claim 29 wherein the step of reacting comprises heating the precursor in a gas containing Se.

41. The method according to claim 21 wherein the step of reacting comprises heating the precursor in a gas containing at least one Group VIA material.

42. The method according to claim 18 wherein the step of reacting comprises depositing a Group VIA material layer on the precursor thus forming a precursor/Group VIA material stack, and heating the stack.

43. The method according to claim 29 wherein the step of reacting comprises depositing a Se layer on the precursor thus forming a precursor/Se stack, and heating the stack.

44. The method according to claim 21 wherein the step of reacting comprises depositing a Group VIA material layer on the precursor thus forming a precursor/Group VIA material stack, and heating the stack.

45. The method of claim 18 further comprising the step of forming electrical contacts on the transparent layer.

46. the method of claim 18 wherein the transparent layer comprises at least one of CdS and ZnO.

47. A method of forming a thin film solar cell having a Group IBIIIAVIA absorber layer with microscale uniformities comprising the steps of: applying, over a sheet-shaped base, a thin Ruthenium film; depositing a precursor in at least one layer, the precursor including a Group IB material, at least one Group MA material, and at least one Group VIA material, the precursor as-deposited having a Group IB to Group IIIA molar ratio of less than or equal to 1.0, and wherein the thin Ruthenium film provides for microscale uniformities of the precursor; and reacting the Group IB material, at least one Group IIIA material, and at least one Group VIA material to form the Group IBIIIAVIA absorber layer; and depositing a transparent layer over the Group IB IHA VIA absorber layer.

48. The method according to claim 47 wherein the step of reacting comprises heating the precursor.

49. The method according to claim 48 wherein the step of depositing comprises a wet process.

50. The method according to claim 49 wherein the wet process is electroplating.

51. The method according to claim 48 wherein the Group IB material is Cu.

52. The method according to claim 51 wherein the Group IHA material is one of In and Ga.

53. The method according to claim 51 wherein the Group IIIA material includes both In and Ga

54. The method according to claim 52, wherein Group VIA material is one of Se and S.

55. The method according to claim 52, wherein the Group VIA material includes both Se and S.

56. The method according to claim 53, wherein the Group VIA material is one of Se and S.

57. The method according to claim 53, wherein the Group VIA material includes both Se and S.

58. The method according to claim 50 wherein the Group IB material is Cu, Group IIIA material is at least one of In and Ga, and Group VIA material is Se.

59. The method according to claim 48 wherein the step of applying is at least one of electroplating, electroless plating, atomic layer deposition, chemical vapor deposition, metal- organic chemical vapor deposition and physical vapor deposition.

60. The method according to claim 59 wherein the step of applying is electroplating.