Patent Application
20150027519
The present invention relates to the manufacture of multi junction solar cell substrates, in particular, to the manufacture of multi junction solar cell substrates comprising wafer transfer processes and the manufacture of solar cell devices for terrestrial and space-related applications.
Photovoltaic or solar cells are designed for converting the solar radiation to electrical current. In concentrator solar photovoltaic applications, the incoming sunlight is optically concentrated before it is directed to solar cells. For example, the incoming sunlight is received by a primary mirror that reflects the received radiation toward a secondary mirror that, in turn, reflects the radiation toward a solar cell, which converts the concentrated radiation to electrical current by the generation of electron-hole pairs in III-V semiconductor or single crystal silicon, for example. Alternatively, the sunlight could be concentrated onto solar cells by using transmittive optics like Fresnel lenses.
Since different semiconductor material compositions show optimal absorption for different wavelengths of the incoming solar light, multi-junction solar cells have been proposed that comprise, for example, three cells showing optimal absorption in different wavelength ranges. A triple cell structure may comprise a GaInP top cell layer with a gap value of 1.8 eV, a GaAs intermediate cell layer with a gap value of 1.4 eV, and a Ge bottom cell layer with a gap value of 0.7 eV, for instance. In principle, III-V or group IV semiconductors can be used as active subcells of multi junction cell devices manufactured by layer transfer/bonding. Multi junction solar cells are usually manufactured by monolithic epitaxial growth. The monolithic growth process requires, in general, that any formed layers be substantially lattice matched to previously formed layers or the underlying substrate. However, the epitaxial growth of the solar cell layers on growth substrates still provides a demanding problem in view of lattice mismatches. For example, it is not suitable to epitaxially grow an InP solar cell layer on a Ge substrate, since the crystalline and optical characteristics of the InP solar cell layer would be heavily deteriorated due to crystal mismatch. In addition, in conventionally used transfer processes, intermediate substrates are lost after the transfer of epitaxially grown layers.
Thus, despite the recent engineering progress, there is still a need for an improved manufacturing process for multi junction solar cell devices, wherein solar cell layers with a low defect rate are achieved and an intermediate substrate can be recycled.
The present invention addresses the above-mentioned need and, accordingly, provides a method for manufacturing a multi junction solar cell device, comprising the steps of:
The resulting structure thus comprises at least one solar cell layer on either side (upper and lower surfaces, respectively) of the second substrate. Different from the art, the solar cell layers on either side of the second substrate are not all provided by epitaxial deposition, but also by bonding. Since the surfaces of the second substrate can be polished before bonding, the bond can be achieved with high quality, in particular, with a better quality as compared to bonding at two (epitaxially grown) solar cell layers. The step of bonding the first wafer structure to the second wafer structure can be a high-temperature bonding performed at a temperature of between about 400° C. and about 600° C., particularly, between 450° C. and 550° C., thereby allowing a reliable bond. The second substrate can be a relatively thin substrate of some 100 micrometers, for example, of about 200 micrometers, and transparent to the some wavelengths of the incoming solar light, as it will be explained below. In particular, the second substrate can be made of or comprise GaAs.
The first substrate can be provided in the form of an engineered substrate, in particular, comprising sapphire.
The term “engineered substrate” comprises a substrate that is different from a mere pure bulk substrate but rather includes a layer or interface that is formed in the substrate in order to facilitate its removal from the structure resulting of the bonding of the first and second wafer structures. In particular, the “engineered substrate” may comprise a zipper layer between a seed layer and a base substrate. In particular, the engineered substrate may comprise a base substrate that is detached from the seed layer in the step of removal of the engineered substrate.
Detachment by means of the zipper layer allows for recycling the detached substrate.
Herein, the expression “detachment of the engineered substrate” should be interpreted as the detachment of the base substrate. This detachment step may be followed by the removal of the possible residue of the zipper layer (if any), and of the removal of the seed layer from the remaining structure.
The zipper layer may be a weakened layer formed, e.g., by an appropriate treatment, for instance, a hydrogen or helium implantation in a substrate, that delimits an upper seed layer and a lower base substrate.
The zipper layer may be formed by a buried porous layer by anodic etching at a surface of the base substrate. Then, epitaxial growth of the seed layer can be performed on top of the porous layer.
The zipper layer may be provided in the form of an absorbing layer for laser lift-off, chemical lift-off or mechanical splitting in an intermediate strained layer during an epitaxy sequence: SiGe in Si matrix, in particular, an intermediate strained layer of SiGe at 20% in an Si substrate. In this alternative, the zipper layer may be formed by the seed layer itself; for instance, the seed layer can be selectively and chemically etched away to detach the engineered substrate.
The zipper layer may also be formed of an SiN absorbing layer for laser lift-off inserted between a seed layer and a transparent base substrate, as known, for example, from WO2010015878.
Another possibility for an engineered substrate reads as follows: A removable (presenting a low bonding energy of less than 1.5 J/m2, and preferentially less than 1 J/m2) bonding interface is formed between facing surfaces of a seed layer and a base support. In that possibility, the zipper layer is formed by the removable bonding interface. A first solar cell layer may be grown by epitaxy on the seed layer while preserving the removable character of the bonding interface, with the engineered substrate being heated to an epitaxial growth temperature. The low energy bonding is obtained by performing a treatment for augmenting the roughness of the facing surface of one of the seed layer or the substrate, in particular, carried out by chemical attack or etching, by effecting a treatment for decreasing hydrophilicity of the facing surface of one of the seed layer or the substrate (or the bonding layer in SiO2 or Si3N4 on each of them). Moreover, a different material for the bonding layer can be chosen such that weak intrinsic mutual chemical affinity of the interface materials is achieved. The detachment of the base substrate may be performed by application of a thermal treatment or mechanical stresses applied from a jet of fluid or a blade, for example. This is disclosed, for instance, in WO03/063214.
As already mentioned, the engineered substrate includes a seed layer formed at the top of the zipper layer or removable bonding interface. The seed layer is transferred from the seed substrate to a base substrate by layer transfer from wafer to wafer, for example, by the S
The seed layer is not necessarily used as a part of a solar cell layer because its band gap may not be adapted for an optimal efficiency in the sequence of subcell layers. Therefore, in an embodiment of the present invention, the seed layer is only used for the growth of the first solar cell layer, thereby resulting in a multi junction solar cell device comprising a first solar cell layer (as a bottom cell) with both satisfying crystal and electrical characteristics. In fact, according to the above example, one might remove the seed layer, e.g., by chemical treatment as etching.
In the above-described examples, the solar cell layers can be grown with high crystal and electrical quality. For example, a dislocation density of less than 106/cm2 can be achieved.
The detachment at the zipper layer may be performed by laser lift-off: a laser beam is radiated to the engineered substrate, for example, comprising a sapphire substrate, and provokes the decomposition of the material in the zipper layer, thereby allowing for detachment at the zipper layer.
Further, the method may comprise forming a contact on the at least one first solar cell layer after detachment of the first substrate and/or forming a contact on the at least one second solar cell layer. Furthermore, before forming a contact, the corresponding solar cell layer may be patterned to form a mesa whereupon the contact can be provided.
The structure comprising the second substrate and at least one solar cell layer on either main surface of the same may be bonded to a base substrate. The bonding to the base substrate can be performed by means of an electrically conductive (metallic, or with conductor oxides, transparent conductor oxide TCO, silicide or by direct bonding, etc.) bonding layer formed on the first solar cell layer and/or the base substrate. An electrically conductive layer also particularly helps to provide for thermal conductivity to the base substrate that may function as a heat sink or may be connected to a heat sink and preferably is itself thermally and electrically conductive. The conductive bonding may be obtained by forming a eutectic Au/Sn or by thermocompression.
In this embodiment, bonding of the first and second wafer structures is performed while the (final) base substrate is not present in the structure. Thus, risk of metallic contaminations is avoided during bonding thermal treatment, such that high temperature may be used. Moreover, the expansion of the base substrate that would be different from the expansion of the first and second substrates at such a temperature due to different coefficient of thermal expansion (CTE) would damage the structure, in particular, it would weaken the bonding interface.
A final base substrate made of tungsten or molybdenum or doped semiconductors like Ge, GaAs or InP may be particularly suitable for receiving the stack of solar cell layers provided on a second substrate made of, for instance, GaAs or GaAsOS (see below). In particular, the difference in CTE of the final base substrate to the CTE of the second substrate should be less than 30% in order to avoid problems related to the bonding to the final substrate.
According to a particular variant, in the above-described examples, the at least one first solar cell layer comprises two layers, a first layer and a second layer, that have been grown by epitaxy on the first layer, and the at least one second solar cell layer also comprises two layers, namely a third layer and a fourth layer that have been grown by epitaxy on the third layer. According to a particular example, the first layer (bottom cell) comprises or consists of GaInAs, and/or the second layer comprises or consists of GaInAsP, and/or the fourth layer (top cell in the finished multi junction solar cell device) comprises or consists of GaInP, and/or the third layer comprises or consists of GaAs. Thus, a four-cell multi junction solar cell device is achieved (comprising two solar cell layers on either the main surface of the second substrate) wherein the material of each cell is optimized for a particular wavelength of the incoming solar light. The first and second solar cell layers are bonded to one side (main surface) of the second substrate, whereas the third and fourth solar cell layers are grown on the other side (main surface) of the second substrate. No inversion of layers is necessary during the entire manufacturing process, i.e., each series of layers are grown in the final order on their respective substrates, the second layer is grown over the first layer and the fourth layer is grown over the third layer.
It should be noted that in all of the above-described examples, only one single detachment step (detaching the first substrate) is involved.
The present invention also provides an intermediate semiconductor device, comprising:
The intermediate semiconductor substrate according to an example further comprises a further substrate comprising InP, in particular, an engineered substrate comprising InP, in particular, InP on sapphire, whereupon the first solar cell layer is formed.
The intermediate semiconductor substrate may further comprise a base substrate, in particular, a base substrate made of Cu, Mo, W or Si or Al, in particular, aluminium sheet bonded to the first solar cell layer, in particular, by a conductive bonding layer.
A final base substrate made of tungsten or molybdenum or doped semiconductors like Ge, GaAs or InP may be particularly suitable for receiving the stack of solar cell layers provided on a second substrate made of, for instance, GaAs or GaAsOS (see below). In particular, the difference in CTE of the final base substrate to the CTE of the second substrate should be less than 30% in order to avoid problems related to the bonding to the final substrate.
In addition, a multi junction solar cell device is provided that is obtainable by one of the above-described examples of the method for the manufacture of a multi junction solar cell device according to the present invention.
Additional features and advantages of the present invention will be described with reference to the drawings. In the description, reference is made to the accompanying figure that is meant to illustrate an embodiment of the invention. It is understood that such an embodiment does not represent the full scope of the invention.
An example for the inventive method for the manufacturing of a multi junction solar cell comprising four solar cell layers is shown in
The advantage of a first base substrate in sapphire and a second substrate in GaAs is that both are perfectly matched in CTE with the whole stack of solar cell layers, which makes it possible to apply bonding temperatures as high as 600° C. without degrading the solar cell. A similar reasoning applies to materials such as, for instance, Ge, GaAs or InP, which may be used as a first base substrate 1 and a second substrate 6. The used first base substrate 11 may not be conductive, such as, for instance, sapphire, which makes the bonding to a final base substrate that is conductive necessary. Preferentially, the materials of the first base and second substrates 1, 6 may be different, but the difference in their CTE should be less than 30% in order to allow high bonding temperatures. The use of Si for the first base substrate, however, is not suitable due to the difference in CTE between the seed layer used for growth of the solar cell layers and the base substrate with regard to the high temperatures used for epitaxial growth of the solar cell layers.
In principle, the engineered substrate can be fabricated by means of layer transfer of the respective seed layers to the base substrates. For example, an InP epitaxial layer or, as another example, a strained or relaxed InGaAs layer may be the surface layer of an initial bulk InP seed substrate, which is subsequently transferred to a sapphire substrate by the well-known S
The zipper layer 2 is comprised of SiN or the material of the seed layer. The seed layer 3 will not function as a solar cell layer but rather for the epitaxial growth of a first solar cell layer 4. The thickness of the InP seed layer may be in the range of 50 nm to 1μm.
A first solar cell is formed on the first engineered substrate; and on the first solar cell layer 4, a second solar cell layer 5 is epitaxially grown.
Furthermore, a second substrate 6 is provided, whereupon a third solar cell layer 7 and a fourth solar cell layer 8 are epitaxially grown. The second substrate 6 is preferably a bulk GaAs substrate, since this material is transparent to the wavelengths of the incoming solar light that will correspond to the absorption wavelength of the solar cell layers 4 and 5. The second substrate 6, according to the shown example, is polished on both (lower and upper) surfaces in order to facilitate the subsequently performed bonding (see below).
The four solar cell layers 4, 5, 7, 8 show absorption maxima for incident solar light for different wavelengths. It should be noted that the first solar cell layer 4 becomes the bottom cell and the fourth solar cell layer 7 becomes the top cell in the finished multi junction solar cell device (see description below). According to the present example, all of the four monocrystal solar cell layers 4, 5, 7, 8 are formed by epitaxial growth. In principle, the material of the solar cell layers can be selected form III-V semiconductors of the group consisting of InGaAs, GaAS, AlGaAs, InGaP, InP and InGaAsP. For example, the first solar cell layer 4 may be comprised of InGaAs; the second solar cell layer 5 may be comprised of InGaP, InGaAsP or InP; the third solar cell layer 7 may be comprised of GaAsP or GaAs; and the fourth solar cell layer 8 may be comprised of InGaP or InAlAs. Appropriate tunnel junction layers may be provided between particular ones of the solar cell layers by deposition or growth on a respective solar cell layer.
As it is illustrated in
Subsequently, the first base substrate 1 is detached from the seed layer 3 by means of the zipper layer 2, for example, the first engineered substrate is cleaved at the zipper layer 2 allowing the detachment of first base substrate 1 from the seed layer 3. For instance, the zipper layer 2 may be provided in the form of a weakened layer formed by the implantation of ionic species, for example, hydrogen or helium ions. In this case, the first substrate 1 can be provided as a bulk InP substrate, the implantation through the main top surface of the bulk substrate forms the weakened layer 2 that creates the seed layer 3 and the first base substrate 1, respectively, in the top portion and the bottom portion of the bulk substrate. Then, detachment of the first base substrate 1 can be mediated by the application of mechanical forces to delaminate at the weakened layer formed by hydrogen or helium ions.
Moreover, residue of the base substrate and/or bonding layer and the seed layer 3 may be removed by etching and/or polishing, for example, in particular, if the seed layer 3 shows poor electrical and/or thermal properties and could not be used in the subsequent device. It is noted that the detached first base substrate 1 can be recycled to form a new engineered substrate for the re-usage as a growth substrate for epitaxial layers.
The second substrate is ensuring sufficient stability and rigidity to the structure, such that it is not mandatory to provide this structure onto a final substrate. However, in an optional approach, the thus-obtained structure comprising two solar cell layers on either main surface of the second substrate 6 can be bonded to a base substrate 10 (shown in
A final base substrate made of tungsten or molybdenum or doped semiconductors like Ge, GaAs or InP may be particularly suitable for receiving the stack of solar cell layers provided on a second substrate made of, for instance, GaAs or GaAsOS (see below). In particular, the difference in CTE of the final base substrate to the CTE of the second substrate should be less than 30% in order to avoid problems related to the bonding to the final substrate.
In the next step illustrated in
The finish processing could thereby be performed, either before the bonding to the final base substrate, which may be advantageous in terms of the efficiency of the workflow and ease of the processing. However, if it is preferred, the finish processing may be performed after transfer to the final base substrate.
All previously discussed embodiments are not intended as limitations but serve as examples illustrating features and advantages of the invention. It has to be understood that some or all of the above-described features can also be combined in different ways. It has to be understood that some or all of the above described features can also be combined in different ways. In particular, it is possible, according to the invention, to form multi-junction solar cells not only composed of four junctions (as generally disclosed in the previous embodiments) but also 2, 3, 5 or more.
| Number | Date | Country | Kind |
|---|---|---|---|
| 12290106.9 | Mar 2012 | EP | regional |
This application is a national phase entry under 35 U.S.C. §371 of International Patent Application PCT/EP2013/055133, filed Mar. 13, 2013, designating the United States of America and published in English as International Patent Publication WO 2013/143850 A1 on Oct. 3, 2013, which claims the benefit under Article 8 of the Patent Cooperation Treaty and under 35 U.S.C. §119(e) to European Patent Application Serial No. 12290106.9, filed Mar. 28, 2012, the disclosure of each of which is hereby incorporated herein in its entirety by this reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2013/055133 | 3/13/2013 | WO | 00 |
