KR20130099229A - Laser processing methods for photovoltaic solar cells - Google Patents

Abstract

Various laser processing schemes have been disclosed for producing various heterojunction and homojunction solar cells. The method includes base and emitter contact openings, selective doping, metal removal, annealing to improve passivation, and selective emitter doping through laser heating of aluminum. It is also disclosed that the laser processing scheme is suitable for selective amorphous silicon removal and selective doping for heterojunction solar cells. Laser ablation technology discloses leaving the underlying silicon substantially intact. Such laser processing techniques include crystalline silicon substrates, and are also fabricated through wire saw wafering or through other decomposition techniques such as epitaxial deposition processes or ion implantation and heating, and planar or textured / 3 dimensional silicon substrates. It can be applied to a semiconductor substrate comprising a. This technique is well suited for thin crystalline semiconductors containing thin crystalline silicon films.

Description

Laser processing method of photovoltaic solar cell {LASER PROCESSING METHODS FOR PHOTOVOLTAIC SOLAR CELLS}

Cross Reference of Related Application

The present invention discloses US Provisional Application No. 61 / 417,181, filed November 24, 2010, 61 / 428,600, filed December 30, 2010, 61 / 428,953, filed December 31, 2010, Priority claim 61 / 428,957, filed December 31, 2010, the contents of which are incorporated herein by reference. In addition, the present application is filed on partial US application Ser. No. 13 / 303,488, filed on November 23, 2011, US patent application Ser. No. 13 / 118,295, filed on May 27, 2011, on October 11, 2011. US Patent Application No. 13 / 271,212, filed Oct. 6, 2007 US Patent Application No. 11 / 868,488, filed Oct. 6, 2007, US Patent Application No. 11 / 868,492, May 2010 US Patent Application No. 12 / 774,713, filed on May 5, and US Patent Application No. 13 / 057,104, filed on February 1, 2011, the contents of which are incorporated herein by reference.

FIELD OF THE INVENTION The present invention generally relates to solar photovoltaic regions, and more particularly to laser processing techniques for manufacturing crystalline semiconductors comprising crystalline silicon, and other types of photovoltaic solar cells.

Laser processing offers several advantages in terms of increased efficiency and reduced manufacturing costs for high performance, high efficiency solar cell processing. First, advanced crystalline silicon solar cells can benefit from having important feature sizes such as very small electrical contact than in current industrial practice. In front contacted solar cells, the contact area of the front metallization to the emitter and the back metal contact area to the base need to be small (or the contact area ratio should be quite small, preferably 10%). Must be much lower). For all back contact, back junction solar cells, the emitter and base regions that form the p / n junction and metallization are on the same side (the back side of the cell is opposite the solar side), and the size of the various features provides high efficiency. Generally small. In such a battery, the emitter and the base region generally form alternative stripes, and the width of this region (particularly the width of the base contact) tends to be small. In addition, the magnitude of the metal contact to these regions tends to be proportionately small. Subsequently, the metallization that is connected to the emitter and the base region needs to be patterned correspondingly at finer scale. In general, lithography and laser processing are techniques with relatively fine resolution to provide small size and the required control. Among these techniques, only laser processing provides the low cost advantage required for solar cell manufacturing. Lithography requires consumables such as photoresist and subsequent resist developer and stripper (which increases process cost and complexity), but laser processing is a non-contact, dry, direct write method and requires no consumables. This allows for simpler, lower cost processing of solar cell manufacturing. In addition, laser processing is the best choice for environmentally friendly manufacturing, as it is a predrying process that does not use any consumables such as chemicals.

In addition, to reduce the cost of solar cells, there is a desire to reduce the thickness of the crystalline silicon used while simultaneously increasing the cell area for greater power per cell and lower manufacturing cost per watt. Laser processing is a completely non-contact, dry process and can be easily adjusted to a large battery size, which is suitable for thin wafers and thin film battery substrates.

Laser processing is also attractive because it is generally a "green", environmentally friendly process that does not require or use toxic chemicals or gases. With the proper choice of laser and processing system, laser processing offers the possibility of very high productivity at very low cost for the owner.

Despite these advantages, the use of laser processing in the production of crystalline silicon solar cells is limited because no laser processing has been developed to provide high performance cells. Laser processing using schemes tailored to each major application for manufacturing high efficiency solar cells is described herein. Also described is a particular embodiment of an application of laser processing to fabricate thin film crystalline silicon solar cells, such as formed using a sub-50-micron silicon substrate formed by epitaxial silicon growth.

Various laser processing schemes for making heterojunction and homojunction solar cells are described herein. The method includes base and emitter contact openings, front and back surface field formation, selective doping, metal ablation, annealing, and passivation. Also described are laser processing schemes suitable for selective doping and selective amorphous silicon removal for heterojunction solar cells. Such laser processing techniques include crystalline silicon substrates, including crystalline silicon substrates, and also made via a wire saw wafer method or through an epitaxial deposition process, and planar or textured / 3-dimensional semiconductor substrates can be applied. This technique is very suitable for thin crystalline semiconductors containing crystalline silicon films.

Emitter isolation (including but not limited to shallow trench isolation) for all back contact homojunction emitter solar cells (eg, high efficiency back contact crystalline silicon solar cells), Openings to base doping, base and emitter contact openings (contact area ratio substantially lower than 10% for controlled contact small ratio, such as reduced contact recombination loss and increased cell efficiency), selective doping ( For example, for base and / or emitter contact doping, and metal removal for front end and all back contact / back junctions (formation of patterned metallization layers, such as subsequent adhesion of the backplane to the cell and from reusable host templates Forming a patterned metallized seed layer on the thin film single crystal silicon solar cell prior to release. The laser processing schemes which satisfy the requirements of the bus are described. Also, selective amorphous silicon removal and oxide (eg transparent conductive oxide (TCO)) removal, and metal patterning for heterojunction solar cells (eg, back contact solar cells comprising heterojunction amorphous silicon emitters on a single crystal silicon base). A laser processing scheme suitable for the removal of metals is described. Such laser processing techniques include crystalline silicon substrates, and also include crystalline silicon substrates produced via wire saw wafer method or through an epitaxial deposition process, and planar or textured / 3-dimensional semiconductor substrates may be applied, Three-dimensional substrates can be obtained using epitaxial silicon liftoff techniques using porous silicon seed / separation layers or other types of sacrificial separation layers. This technique is well suited for thin crystal semiconductors comprising crystalline silicon films obtained using epitaxial silicon deposition on templates comprising porous silicon separation layers or other techniques known in the industry.

All back contact homojunction solar cells can be formed on a crystalline silicon substrate, and laser processing is used to do one or a combination of the following: a base comprising a base during micromachining or separating the openings and emitters of the base. Pattern the regions and emitters (micromachine or pattern the emitter and base regions including base to emitter isolation as well as openings for base), provide selective doping of emitters and bases, and And provide metal patterning, provide annealing, and provide passivation. Front contact homojunction (emitter) solar cells can be fabricated using laser processing to produce openings in the metal contacts for selective doping of the emitter and metallization of the front and back sides. Heterojunction All back contact back contact solar cells can be fabricated using laser processing to define base region and conductive oxide separation.

The features, nature and advantages of the described subject matter will become more apparent from the detailed description of the invention set forth below when taken in conjunction with the drawings, in which like reference numerals indicate similar features:
1 shows an SEM image of a shallow trench made of silicon for application in all back contact back junction solar cells in accordance with the present invention.
2 shows the profile of the shallow trench in silicon for application in all back contact back junction solar cells.
3A-3D show the procedure for selecting laser fluence to obtain less damaged silicon dioxide (or oxide) removal. 3A shows the size dependence of the removal spot on laser fluence; 3B shows irregular exfoliation of oxides; 3C shows the spot without damage; 3D shows very damaged silicon in the spot opening.
4 shows a substantially parallel row of open contacts in the oxide using pulsed laser ablation according to the present invention.
5 shows a screenshot with oxide removal spots for metal contact.
6A and 6B show laser-ablated areas formed by fabricating removal spots that overlap in the x and y directions; FIG. 6A shows a 180 micron wide strip opened at 1000 A BSG (boron doped oxide) / 500 A USG (undoped oxide) for the base isolation region; 6B shows a 90-micron wide strip opened with 1000A USG (undoped oxide) for the base region.
7A shows the threshold of oxide damage below which metals can be removed without metal penetration of the oxide layer.
7B shows that after 20 scans, the metal runner is completely disconnected.
7C shows optical micrographs of trenches formed in such a metal stack.
8A and 8B show top and cross-sectional views of a pyramidal TFSC.
9A and 9B show top and cross-sectional views of the prismatic TFSC.
10A and 10B show process flows for the formation and separation of planar epitaxial thin film silicon solar cell substrates (TFSS).
11A and 11B show the process flow for a planar epitaxial thin film silicon solar cell substrate when the TFSS is too thin to free stand or self-supporting.
12A and 12B show the process flow for micromold templates (or reusable templates) for making 3-D TFSS.
12C and 12D show the process flow of a 3-D TFSS using reusable micromolded templates.
FIG. 13 shows TFSS alone, in accordance with the present invention, is thick enough to self-support (eg, thicker than about 50 microns for a smaller 100 mm × 100 mm substrate, and about 80 for a 156 mm × 156 mm substrate). Thicker than microns), a process flow for making planar front contact solar cells.
FIG. 14 shows a process flow for making a very thin, planar front contact solar cell such that the TFSS is self supporting in accordance with the present invention.
15 shows a process flow for making a 3-D front contact solar cell, in accordance with the present invention.
16A-16D illustrate a process flow for producing an interdigitated back contact back junction solar cell thick enough to allow the TESS to self-support, in accordance with the present invention.
FIG. 17 shows a process flow for fabricating interfacing back contact back junction solar cells using thick TFSS where no in-situ emitters are deposited. Instead, in accordance with the present invention, a boron-doped oxide (BSG) layer is deposited on the epitaxial silicon film and patterned to open the base isolation region.
18 is a process for fabricating an interdigitated back contact back junction solar cell in which the TFSS is not thick enough to self-support, and wherein laser ablation and in situ emitters of silicon are used to form the base isolation openings, in accordance with the present invention. Shows the flow.
19A-19H are not thick enough for TFSS to self-support in accordance with the present invention, and instead of in-situ emitters, boron-doped oxide (BSG) deposition and selective laser etchback are used to form the base isolation openings. A process flow for manufacturing an interlocked back contact back junction solar cell is shown.
20 illustrates a process flow for fabricating back contact back junction solar cells using 3-D TFSS, in accordance with the present invention.
Figure 21 illustrates a process flow for producing an interdigitated back contact back junction heterojunction solar cell, in accordance with the present invention.
22-30 were not found in US Patent Application 13 / 118,295 "LASER PROCESSING FOR HIGH-EFFICIENCY THIN CRYSTALLINE SILICON SOLAR CELL FABRICATION" of Virendra V. Rana, filed May 27, 2011.
22A and 22B are diagrams illustrating profiles of a Gaussian beam and a flat-top beam, respectively.
23 is a cross-sectional view of the back contact / back junction battery.
24A-24F are back / backside views of back contact solar cells during manufacture.
FIG. 25 is a back / back side view of the back contact solar cell of FIG. 24A with alternating metal wires in contact with the emitter and base region.
26A-26C are diagrams illustrating three ways in which a flattop beam profile can be formed.
27A and 27B are diagrams illustrating profiles of a Gaussian beam and a flat top beam emphasizing the removal threshold.
28A and 28B are diagrams illustrating that the Gaussian beam and the flattop beam remove the area profile / footprint, respectively.
28C is a graph of overlap and scan rate.
29A and 29B are diagrams illustrating a Gaussian beam and a flat top beam alignment window, respectively.
29C and 29D are diagrams illustrating a Gaussian beam area profile and a flat top beam area profile, respectively.
29E illustrates the results of Table 1 graphically.
30 shows a process flow for an NBLAC cell.
31 shows a sectional view of an NBLAC cell.
32 shows a graph of minority carrier life with or without laser annealing.
33A and 33B show process flows for all back contact solar cells with oxide removal.
33C and 33D show process flows for all back contact solar cells with oxide removal.
34A and 34B illustrate an oxide removal process.
35A and 35B show an oxide removal process using an amorphous silicon layer.
36 shows a process for forming FSF and passivation.
37 shows a process for forming FSF and passivation with amorphous silicon.
38A and 38B show schematics of selective laser scanning and formation of selective emitters of metal on emitter regions, respectively.
39 shows a P ⁺⁺ selective emitter.
40 shows an aluminum-silicon phase diagram.
41 shows an optional emitter in the front contact cell.
42 shows aluminum BSF.

While the present invention is described in terms of specific embodiments, those skilled in the art can apply the principles discussed herein to other areas and / or embodiments without undue experimentation.

Disclosed herein are laser processing, and more specifically pulse laser processing and schematics, which have been developed to address the varied requirements of other processes.

The disclosed method may be useful in areas of semiconductor device removal, particularly crystalline silicon removal. Removal of silicon with a laser generally includes silicon melting and evaporation which eliminates unwanted residual damage in the silicon substrate. This damage causes increased surface recombination velocity (SRV) and minority carrier lifetime degradation, which reduces the efficiency of the solar cell. Thus, wet cleaning of silicon substrates is commonly used to remove such damaged layers. Fabricated without the need for post-laser-processing wet cleaning, it provides a schematic for reducing damage to levels acceptable for high-efficiency solar cells that simplify process flow and reduce manufacturing costs. .

When removed to a predetermined thickness using a laser, the damage remaining on the silicon substrate is related to the amount of laser energy absorbed by the substrate that is not used by the removed material. After most of the laser energy is used to remove material, laser-induced substrate damage and SRV degradation are minimized if it is possible to minimize some of the minor energy that penetrates the silicon substrate. The penetration of laser energy into silicon depends on the laser pulse length (also called pulse width) and wavelength. An infrared (IR) laser with a wavelength of 1.06 microns has a long penetration depth of up to about 100 microns in silicon, but a green laser beam with a wavelength of 532 nm is only transmitted to depths of about 3 to 4 microns. The penetration of the UV laser beam with a wavelength of 355 nm is shorter, only about 10 nm. It is obvious that using ultra-short pulses of UV or EUV wavelengths will limit the penetration of laser energy into silicon. Additionally, shorter laser pulse lengths shorten the diffusion of heat into the silicon. While nanosecond pulses can lead to heat diffusion into silicon in the range of about 3 to 4 microns, picosecond pulses are reduced to about 80 to 100 nm, and femtosecond pulses are too short, typically silicon during the laser ablation process. It does not spread heat. Thus, shorter pulses with shorter wavelengths save damage to the laser-removed substrate. For higher production throughput, green or IR wavelengths can be used differently depending on the range of acceptable laser damage. Even under ideal conditions, some portion of the energy still penetrates into the substrate, and this absorption and undesirable adverse effects can be further reduced by reducing the laser power. However, this allows the smaller thickness of the silicon to be removed (or lower the silicon removal rate or lower the throughput). It has been found that reducing the pulse energy, but causing the silicon removal by increasing the overlap of the laser pulse, smoothes the silicon shallow isolation trench. This is an indication of low silicon surface damage. At very low pulse energies, the thickness of silicon removed can be small. Preferred depths can be obtained by using multiple overlap scans of the pulsed laser beam.

Pulsed laser beams having a pulse length in the picosecond range and wavelengths of about 355 nm or less are suitable for low damage silicon removal that allows for low surface recombination velocity (SRV) on the passivated removed surface. FIG. 1 shows 2.25 fabricated on a silicon substrate using a picosecond UV laser beam (M ² ≦ 1.3) with a Gaussian profile of approximately 110 microns in diameter, with laser spots overlapping nearly 15 times, 4 microjoules pulse energy. A trench of microns depth and a width of nearly 100 microns is shown. This depth of removal is obtained using 20 overlapped scans of the laser with each scan removing about 112 nm of silicon. FIG. 2 shows a smooth profile of trenches of 4 microns deep and 110 microns wide in silicon, obtained using the same picosecond laser beam with UV wavelengths. Note the smoothness of the profile. This removal of silicon is used for all back contact back junction solar cells to form an area where the base area is separated from the emitter area. The use of femtosecond lasers can further reduce laser damage when silicon is removed.

In addition, embodiments of the present invention can be used to remove amorphous silicon. Similar schemes may be used to remove the desired thickness of amorphous silicon using femtosecond pulse lengths and in some embodiments using pulsed laser beams having UV or green wavelengths. Since the removal of amorphous silicon requires much lower energy than crystalline silicon, this scheme can be efficiently used to selectively remove amorphous silicon from the crystalline silicon surface for applications to heterojunction solar cells.

In addition, the present invention is applicable to selective oxide removal in an underlying substrate, which may be crystalline or amorphous silicon. The oxide film is transparent to the laser beam down to the UV. When nanosecond pulse length lasers are used to remove the underlying oxides, the removal of the oxides occurs by heating and melting of the silicon below. Due to the pressure from the silicon removed below, the underlying oxide splits and is removed. However, since this causes significant damage to the silicon substrate, wet cleaning processing is generally used to remove such damaged layers in order to use high efficiency cells.

Provides a scheme in which the oxide layer is selectively removed from the silicon surface to the silicon surface without significant damage. In laser ablation, besides heating or depositing the material until it melts, other effects such as plasma formation occur. Sometimes complex processes can occur at the interface. Using a laser with a picosecond pulse length, it is possible to affect the oxide up to the silicon interface. By using a picosecond laser having a UV wavelength, the interfacial effect is improved to cause separation and exfoliation of the oxide from the silicon surface. The silicon surface left behind is virtually intact. In addition, picosecond laser irradiation with green or infra-red (IR) wavelengths can be used differently depending on how much penetration damage the silicon substrate is tolerable. The present invention will describe a procedure for obtaining selective removal of oxides from the silicon surface without damage.

3A-3D disclose a procedure for obtaining intact removal of oxides. FIG. 3A shows the variation of laser spot openings in a 1000A phosphorus-doped oxide (PSG) / 500A undoped oxide (USG) stack on a 35 micron thick epitaxial silicon film on a template using a picosecond UV laser beam. do. The oxide layer is deposited using atmospheric-pressure CVD (APCVD) technology. For a given thickness of oxide, the spot size depends on the laser fluence (J / cm ² ). Laser fluence is a laser pulse divided by the area of the laser beam. In this case, the laser beam is about 100 microns in diameter with a Gaussian profile (M ² <1.3). At very low fluences, the spot is irregular and there is irregular peeling of the oxide from the silicon surface as shown in FIG. 3B, but at very high fluences there is extensive damage of silicon as shown in FIG. 3D. The range of fluence, indicated by the line a-a ', represents the optimum range and damage to the silicon substrate is minimal as shown in FIG. 3C.

4 shows a row of cell contact openings that are optionally retrofitted to oxides for use in all back contact (and back junction) solar cells. 5 is a close-up of these contacts. The laser ablation spots may overlap in the x and y directions, opening up to any desired length and nod area on the wafer as shown in FIGS. 6A and 6B. FIG. 6A shows a 180 micron wide opening prepared by selectively removing boron-doped oxide (BSG) for the base removal area using a picosecond UV laser beam having overlapping spots in the x and y directions. . Likewise, FIG. 6B shows a 90 micron wide region that opens to an undoped oxide (USG) to form a base region.

As described herein, selective removal of oxides from silicon surfaces can be used in solar cells manufactured in several ways. In one application, when using an in-situ emitter for a back contact cell, this process is used to open tracks in the oxide film to expose the lower emitter. This exposed emitter may be removed using wet etching. This region is then used for base to emitter isolation and with base formed inside it.

In other applications, this process is used to open areas used to fabricate metal contacts. For front contact cells, oxide passivation can be used on the back side of the cell. The scheme described herein is then used to open the contact to the metal deposited on these contacts. In this way, the metal localizes the contacts, which is beneficial for high efficiency cells. For the back contact cell, the contact of the base and the emitter can be opened using this scheme.

In a solar cell flow process, the doped oxide may need to be removed without causing doping of silicon underneath (ie, without significant heating of the doped oxide and silicon structure). As described above, when using a picosecond laser beam, since the oxide is removed by separation at the oxide / silicon substrate interface, the removal of oxide occurs with limited pickup of the dopant from the oxide film to be removed. happens with limited pickup of the dopant from the oxide film being ablated).

Selective removal of silicon nitride (SiN _x ) is used for front contact solar cells. Using laser ablation, the contact area with the emitter surface can be reduced to maximize the area, and SiN passivation is eliminated. This leads to higher V _OC . Nanosecond UV lasers can be used, but picosecond lasers with UV or green wavelengths are suitable for this application.

Selective metal removal from oxide surfaces has historically been difficult using lasers. This is because at the high pulse energy required to remove the metal, the energy is high enough to damage the oxide below, allowing the metal to penetrate into the oxide. In fact, this is the basis of the LFC ("laser fired contacts") process used in solar cells.

Metal schemes are selectively removed from oxide (or other dielectric) surfaces without metal penetration of oxide (or other dielectrics such as silicon nitride), and three schemes for breaking or cracking oxides are disclosed. In both of these schemes, aluminum is the first metal in contact with the base and the emitter (aluminum is used as the light trapping rear mirror layer). Picosecond pulse length lasers are suitable for this application. For high metal removal rates, IR wavelengths are very suitable. According to the first scheme, the metal is removed at a pulse energy lower than the threshold of oxide removal. If the thickness of the metal removed in one scan is lower than the desired thickness, multiple overlapped scans are used to remove the entire thickness of the metal. Since the pulse energy is below the oxide removal threshold, clean removal of the metal from the oxide surface is obtained. However, the exact recipe used depends on the metal form of the stack, their thickness, surface roughness, and the like.

7A-7C show the removal results when patterning a PVD-deposited bilayer of 2400A NiV on 1200A Al on oxide. It is desirable to completely remove the metal between the runners without breaking through the oxide layer below (to prevent short circuit of the cell). Figure 7A shows the threshold of pulse energy, below which such a metal stack can be removed without penetration of oxide. This limit, which depends on the metal stack characteristics described above, depends on the laser parameters such as the scan rate and the spot overlap obtained using the desired pulse repetition rate of the laser. While increasing the pulse overlap, the threshold pulse energy will decrease due to energy accumulation in the metal. FIG. 7B shows that using pulse energy below the threshold of oxide damage, more than 20 scans provide complete separation of metal runners as determined by the 100 M-ohm resistance between parallel lines. energy below the threshold for oxide damage, more than twenty scans provided complete isolation of metal runners as determined by the 100M-ohm resistance between parallel lines). 7C shows a clean 75 micron trench formed in a 2400 A NiV / 1200 Al metal stack.

In accordance with the second high throughput scheme, higher energy is used since a substantial portion of the minor energy is absorbed while the metal is removed to reduce damage to the oxide. This approach allows laser removal of metal to be a very high throughput process. Using this scheme, a two step process was used to successfully remove NiV of 1250 A Al / 100-250 A with or without a tin (Sn) overlayer to a thickness of 2500 A. In the first step, the soft metal is removed using 15 micro Joules, followed by overlapping 15 30 micro Joules pulses. For thicker aluminum such as 2000A, the second step can be performed at 50 micro Joules with the overlap of the same number of pulses.

The third scheme of metal removal is applicable to a highly reflective film, such as an Al / Ag stack (with Al on top of Al and Al in contact with the cell) so that the minor energy of the picosecond laser is reflected and the removal is drastically Is reduced. In such a case, the surface of the reflective metal (Ag) is first dented using a long pulse length nanosecond laser with a pulse length of 10 to 800 nanoseconds, followed by picosecond cleanup of the underlying aluminum.

The invention is also applicable to selective doping of substrates. For successful doping of silicon using an overlay layer of dopant containing material, the pulse energy should be high enough to melt the silicon, but not high enough to remove it or the dopant layer thereon. While the silicon is melting, the dopant dissolves in it. Upon recrystallization of this silicon layer, a doped layer is obtained. For this application, because of limited penetration into silicon, nanosecond pulse length lasers with green wavelengths are quite suitable.

The laser processing techniques described herein are applicable to planar and 3-D thin film crystalline silicon substrates. Laser processing described herein is suitable for any thickness of silicon commercially available. This includes the current standard wafer thickness of > 150 microns used in crystalline silicon solar cells. However, because it is a process performed without any contact with the substrate, it is more advantageous for a thin, brittle wafer or substrate. It is thinner than 150 microns, or porous silicon, obtained from monocrystalline CZ ingots or multi-crystalline bricks using other techniques such as annealing after hydrogen implantation or advanced wire sawing techniques to separate wafers of desired thickness. Thin film single crystal substrates obtained using epitaxial deposition of silicon and subsequent liftoff on a sacrificial separation / release layer such as (eg, a thickness in the range of several microns to 80 microns).

Laser processing is particularly suitable for three-dimensional substrates obtained using reusable pre-structuring and silicon micromachining techniques. One such method is described in the '713 application (published as US2010 / 0304522). 8A-9B illustrate 3-D thin silicon substrates obtained using the techniques described in this publication. Figure 8A shows a top view, but Figure 8B shows a cross-sectional view of the TFSS thus obtained. For pyramidal substrates, the tip may be flat and may end with a sharp tip. 9A and 9B show TFSS with prisms obtained using the reusable pre-structured 3D template described above.

The laser processes and process flows described herein are applicable to, but are not limited to, silicon substrates of any thickness (less than 1 micron to greater than 100 microns), but are not limited to porous silicon (or other) of reusable templates as described in the '713 application. Sacrificial layer) discloses applications for solar cells fabricated using thin silicon substrates in the range of less than 1 micron to about 80 microns, including those obtained using epitaxial silicon on the surface. To facilitate understanding of this application, the process flow for obtaining single crystal TFSSs of planar thickness of desired thickness (e.g., less than about 10 microns up to about 120 microns), in general, is greater than about 50 microns so that 10A and 10B for a planar TFSS that can be processed as a supported substrate, generally less than about 50 microns to a planar TFSS that does not self-support during cell processing (and thus is strengthened prior to separation from the host template). Are shown in FIGS. 11A and 11B. 12A-12D show process flows for obtaining three-dimensional pyramidal silicon substrates. Three-dimensional prismatic substrates can be obtained in a similar process, except using lithography or screen printed patterns that provide for this structure.

The thin planar substrate obtained using the process flows of FIGS. 10A and 10B can be processed according to the process flow of FIG. 13 to obtain a high efficiency front contact solar cell. It should be noted for self-supporting TFSSs it is advantageous to process the template side of the TFSS first before proceeding before proceeding to the other side. to the other side). During the removal of quasi-monocrystalline silicon remaining on the TFSS prior to separation from the template, the front side or sunnyside of the solar cell is preferred because the template side of the TFSS is textured. Laser processing of selective removal of silicon oxide and silicon nitride (SiN) is useful for making such front contact solar cells.

FIG. 14 illustrates the application of various laser processes for fabricating high efficiency front contact solar cells using planar TFSSs, where the TFSS is so thin that it stands free or self-supporting during cell processing. Can't be. In this case, it should be noted that the non-template side surface is first machined with TFSS on the template. Once this processing is complete, the TFSS is first attached to a reinforcing plate or sheet (also referred to as a backplane) on the exposed working surface and then separated from the template. After separation of the backplane-attached (or backplane-laminated) thin film crystalline silicon solar cell, removal of residual porous silicon, texture etching, and SiN passivation on the separated side of the TFSS (which in turn will be the front of the solar cell). / ARC deposition and forming-gas anneal (FGA) operating processes are performed.

FIG. 15 illustrates applications of various laser processes for manufacturing high efficiency front contact solar cells using 3-D front side TFSS. For this application, it is useful to have pyramid tips on the template side not be sharp but end in flat ledges.

The process described herein is more uniquely tailored to simplify all back contact cell process flows.

16A-16D illustrate laser processing used on planar epitaxial substrates to fabricate back contact / back junction solar cells, with the TFSS self supporting (ie, no backplane attachment to the cell). In this application, epitaxial emitters are deposited in-situ during silicon epitaxy after deposition of the epitaxial silicon base. Thereafter, removal of silicon is used to remove the emitter from the base isolation region. At the same time, four origins are etched with oxide to adjust subsequent removal with this pattern. The thermal oxide is then grown to passivate the silicon surface to be the backside of the back contact back junction solar cell. The epitaxial silicon film is then disconnected or separated from the template (by mechanical separation from the porous silicon interface). The residual porous silicon layer is then wet etched and the surface is textured (which can be done using an alkaline etch process). This will be the textured front or solar side of the solar cell. Thermal oxide is now removed using a Picosecond UV laser to form a base opening inside the base isolation region. The base opening is aligned inside the base isolation region (trench) first formed by silicon removal using the origins first etched into silicon as described above. The phosphorus containing oxide layer (PSG) is then a blanket deposited on the surface. Scanning with a nanosecond green or IR laser aligned with the base openings using the origin in silicon allows the base to be doped. In addition, the region that will have the contact openings to emitter is doped in a similar manner using an aligned scan of nanosecond green or IR laser. The contact openings are then brought into these doped base and emitter regions using a picosecond UV laser. Again, the alignment of these contact openings is made using a starting point in silicon. Now, a metal stack layer (eg, a stack of 1250A Al / 100-250A NiV / 2250 Sn) comprising aluminum as the first layer in contact with the cell is deposited using a suitable method such as physical vapor deposition (PVD) technology. . This layer is then patterned using a picosecond IR laser so that the metal runner is connected to the base and emitter regions, respectively. After the selective forming gas anneal (FGA), the cell is with or without embedded (Al or Cu) highly conductive connectors (in the latter case, the final cell metallization can be formed by a copper plating process), It is connected to the backplane and reinforced by the backplane. The cell is ready for testing and use. FIG. 17 illustrates a laser process used on planar epitaxial substrates to fabricate back contact solar cells, wherein the epitaxial silicon base is not deposited with an emitter layer. Instead, a boron containing oxide (BSG) layer is deposited and patterned to open the base isolation region. The similar process described here follows, except that the emitter and base are formed simultaneously during the thermal oxidation step in accordance with the process flow described in FIG. 17.

18 shows a process flow using a laser process on an epitaxial substrate to fabricate a planar back contact / back junction solar cell, and the TFSS is not self supporting (hence the backplane is used). This flow uses silicon removal of the in-situ doped emitter to form the base isolation region.

19A-19H show a process flow using a laser process on an epitaxial substrate to fabricate a planar back contact solar cell, and the TFSS is not self supporting. In this flow, instead of the in-situ emitter layer, thermal oxidation (or thermal annealing or thermal oxidation annealing) after BSG deposition and selective laser ablation is used to form the base isolation region and emitter.

20 shows a process flow for making a back contact 3-D solar cell, and it is useful to have the template side of the pyramid on a relatively sharp tip. The process flow is similar to that shown in FIG. 16, because the 3-D TFSS can self support at relatively low thicknesses (eg, silicon as thin as about 25 microns). It is evident that one can choose to use laser ablation of silicon after in-situ emitter, or thermal oxidation (or thermal annealing, or thermal oxidation annealing) after BSG deposition and selective laser ablation.

For applications in heterojunction solar cells, the heterojunction emitter may be formed of a doped amorphous silicon layer in contact with the doped crystalline silicon base. For interdigitated back contact solar cells, the amorphous layer and the transparent conductive oxide (TCO) are patterned using a crystalline layer and optionally laser ablation. Femtosecond pulse-width lasers with UV or green wavelengths are suitable for this application. The flow process is described in FIG. 21. Various variations of this flow process are possible.

Various embodiments and methods of the present invention include at least the following aspects: a process for obtaining silicon removal of crystalline and amorphous silicon with reduced damage; To obtain oxide removal for doped oxides and undoped oxides with or without damage to silicon; A process for obtaining a fully separated metal pattern on the dielectric surface for solar cell metallization; Selectively doping the emitter and base contact areas; The use of pulsed laser processing on ultra-thin wafers, including planar and 3-D silicon substrates; The use of pulsed laser processing on substrates obtained using epitaxial deposition on reusable templates made using template pre-structuring techniques; The use of various pulsed laser processes in the manufacture of front contact homojunction solar cells; The use of various pulsed laser processes in the manufacture of all back contact homojunction solar cells; And the use of various pulsed laser processes in the manufacture of heterojunction solar cells.

Although the front contact solar cell is described as a p-type base and the back contact back junction solar cell is described as an n-type base, the laser process described here is for the front contact solar cell with the opposite doping, ie P ⁺ emitter. n-type, p-type base and the n ⁺ p-type base to the emitter back contacts the back-junction solar cell with (n-type for front contact solar cell with P ⁺ emitter, and p-type base for back-contact back-junction solar cells with p-type base and n ⁺ emitter are equally suitable for substrates.

The application of flat top laser beams is described below with laser processing for the description, tables and or interdigitated back contact cells (IBC). The following description relates to a method of forming a back contact solar cell using a flat top laser beam as compared to a conventional Gaussian laser beam. In addition, the implementation of flat-top laser beams with the laser processing described through this application can result in substantial reduction of silicon damage, improved solar cell manufacturing throughput, and patterns that fit inside other patterns (eg, patterns in emitter and base regions). Provides a larger alignment window for defining.

22A and 22B are schematic diagrams illustrating the Gaussian beam, FIG. 22A and flat top beam, and the profile of FIG. 22B. The beam intensity of the Gaussian beam decreases smoothly from the maximum from the center of the beam to the outside of the beam. In contrast, the intensity is "flat" or uniform for a flat top beam over most of its profile (from the center to the outside).

As described herein, high efficiency back contact, back junction cells with interlocked back contact (IBC) metallization benefit from at least one or several stages of pulse laser processing. Laser processing can be used for several processing during the formation of back contact cells, including: defining emitter and base regions (or base-to-emitter separation), defining back field (BSF) regions, and Doping to form a backside electric field, and contacting the openings in the dielectric up to the base and emitter, metal patterning. Some of these steps overlap with a Gaussian beam laser spot, requiring large area laser processing, which is typically manufactured. Overlap severely reduces battery processing speed and can cause silicon damage, resulting in degradation of battery performance and yield. By replacing the smaller diameter Gaussian spot with a relatively wide flat top laser beam, substantial improvement in throughput is obtained. In addition, since the overlap of spots is greatly reduced, semiconductor (such as crystalline silicon) substrate damage is significantly reduced. 23-25 illustrate embodiments of back contact solar cells that may be formed according to the described flat top laser beam processing method.

FIG. 23 is a cross-sectional view of a back contact / back junction cell with interlocked back contact (IBC) metallization formed from an n-type substrate, as described herein. FIG. As shown in FIG. 23, the alternating emitter and base regions are relatively lightly separated into n-doped substrate regions (n-type base). The back / back sides are, for example, thermal silicon dioxide deposited on silicon dioxide, or silicon oxide / silicon nitride layers (and / or aluminum deposited by atomic layer deposition or ALD) that may be deposited using techniques such as PECVD or APCVD. Oxide), which is applied with a surface passivation layer that provides good surface passivation with low back recombination rate. This surface passivation process can then produce openings in this passivation layer that act as 'localized contacts' to the emitter and base regions. Conductor deposition and patterning (eg, aluminum as shown in FIG. 23) can then be done to connect the emitter and base regions separately.

FIG. 24A is a backside / backside view of a back contact solar cell with an interlocked back contact base and emitter design in which the emitter and base regions are laid out in parallel rows. This backside may be formed, for example, by starting with a surface that is completely applied to the emitter region and then describing the formed base region of the patterned emitter region. covered by an emitter region, then delineating a base region resulting in the formation of the patterned emitter regions.) Then, doping of the base contact region with phosphorus is performed, and the contact is opened to the base and emitter regions in preparation for metallization. .

24B-24F are back / backside views of a back contact solar cell illustrating a back contact cell after a major processing step, and any one or combination of steps may be performed in accordance with a laser process with or without a flat top beam. Can be. Various laser patterning steps of a particular illustrative method are described in FIGS. 24B-24E. Starting with an n-type silicon substrate, a BSG layer is deposited over the entire surface. The emitter to BSF isolation region is then defined using laser ablation of the BSG as shown in FIG. 24B. This stage of the description of the base and emitter regions is also referred to herein as the "BSG Opening" stage. Alternatively, the in situ boron doped layer may be deposited during silicon epitaxy, and the BSF region may be defined during silicon epitaxy and the BSF region defined using laser ablation of silicon.

After the emitter is defined in the BSG opening step up to the BSF isolation area, the USG layer is deposited on the wafer, as shown in FIG. . This patterning step is also referred to herein as a BSF Opening step or a base opening step. Since short circuits are detrimental to solar cell efficiency, to prevent short circuit formation, the BSF openings must be separated from the edges of the BSG openings.

The PSG layer is then deposited on the wafer, and the silicon exposed to the PSG upon opening the BSG is doped using a selective laser scan of this region. The doped BSF region (base region) is described in FIG. 24D.

Subsequently, contact to the base and emitter is made using laser ablation as shown in FIG. 24E. It should be noted that the contact may be a point contact as shown in FIG. 24E or a line contact as shown in FIG. 24F. In addition, the number of contacts or the number of lines should be optimized for the minimum series resistance of the current conducting path to the solar cell—thus the design and method of the disclosed subject matter is not limited to the embodiments shown herein. It is also important that the contact openings are properly aligned within the particular doped region, so that there is no short circuit.

As previously described, picosecond pulse length lasers can be used in the oxide removal process of BSG opening, BSF opening, and contact opening. In addition, IR wavelengths may be used, but green or UV or smaller wavelengths are more suitable because of the reduced penetration into silicon.

In particular, for BSF doping, nanosecond pulse length lasers may be more suitable because of penetration into silicon. In addition, IR wavelengths may be used, but because of the reduced penetration compared to IR, green wavelengths may generally be more suitable with the desired depth of doping.

FIG. 25 is a back / back side view of the back contact solar cell of FIG. 24A with alternating metal lines in contact with the emitter and the base region. Note that the metal lines of the emitter and base regions are each connected with busbars not shown in FIG. 25 for the sake of simplicity. This metal pattern can be formed by laser ablation of the metal to separate the base contact from the emitter contact after the blanket deposition of the metal. Since relatively thick metal wires require good current conductivity (usually 20 μm thick or thicker wires), thinner metal stacks, such as aluminum / nickel-vanadium / tin / tin, are first deposited and laser patterned, followed by electrical Alternatively, electroless plating can be used for selective deposition of thick metals such as copper. Alternatively, a backplane with relatively thick conductors can be applied and attached to thin conductor wires. Picosecond pulse length lasers with IR wavelengths may be best suited to remove metal stacks with good selectivity to the underlying oxide layer.

The described flat top laser beam processing steps that can be used to enable this structure include, but are not limited to: deposited boron doped dielectrics (such as boro-silicate glass BSG deposited by APCVD) or emitters. Delineation of the base region and emitter (BSF and emitter to BSF isolation) by laser ablation; The technique of BSF by opening a dielectric agent for applying openings made in the BSG; N + doping of the base (eg with phosphorus); Opening the metallization contacts to the base and emitter regions; And metal patterning using metal laser separation to separate base and emitter contacts. 26A-26C are schematics illustrating three methods, and a flattop beam profile can be formed ("Laser Beam Shaping: Theory" by Mercel Dekker Inc., FM Dickey and SC Holswade, NY, which is hereby incorporated by reference in its entirety). and Techniques "(reproduced). FIG. 26A illustrates one technique for forming a flat top beam profile, also known as "aperturing of the beam." Using this method, the Gaussian beam can be flattened by expanding it, and the aperture selects the appropriate flat portion of the beam and cuts down the progressively decreasing "sidewall" area of the beam. Used to cut out. However, this method can cause significant loss of beam power.

As shown in FIG. 26B, a second exemplary method for forming a flat top beam uses beam integration, and multi-aperture optical components, such as microlens arrays, break the beam into many smaller beams and recombine them on a stator plate. do. This beam integration method can work very well with high M ² beams.

As shown in FIG. 26C, a third beam forming system for forming a flat top beam redistributes energy and uses a diffraction grating or refractive lens to map it to the output plane. Any known method, including the three example techniques described in FIGS. 26A-26C, can be used to obtain a flat top beam profile for the applications described herein. Suitability and selection of the flat top beam forming method depends on various factors including the available beam characteristics and the desired result.

27A and 27B are diagrams illustrating profiles of a Gaussian beam and a flattop beam with emphasis on removal thresholds. As shown in Figures 27A and 27B, in particular compared to a Gaussian beam, a flat top laser beam can substantially reduce laser damage during removal and doping processing. For a Gaussian beam, there is a substantial excess laser intensity beyond that required for removal, particularly in the center portion of the beam, which can cause damage to the silicon (as shown in Figure 27A). Flat top beams can be configured so that the peak intensity is slightly higher than what is required to remove the material (removal threshold as shown in FIG. 27B) and avoids damage that may be caused by the high intensity of the Gaussian beam.

Whether having a square or rectangular cross-sectional area, flat top beams offer particularly throughput advantages compared to Gaussian beams. 28A is a diagram illustrating Gaussian beam canceled area profile / footprint. The circular spot of the Gaussian beam needs to be substantially overlapped, typically with as much as 50% overlap, to minimize the zigzag outline of the pattern (FIG. 28A). 28B is a diagram showing a flat top beam rejection area profile / footprint. Square or rectangular flat top beams have a flat edge, forming a flat outline, and overlap can be significantly reduced (FIG. 28B). 28C is a graph showing improvement in scan speed as beam overlap is reduced. Note that 30% overlap and 33% increase in scan speed can be realized.

FIG. 29A is a diagram showing a beam alignment window of a Gaussian beam, and FIG. 29B is a diagram showing a beam alignment window of a flat top beam. As shown in Figures 29A and 29B, another advantage of using a flat top beam to make a pattern is that the alignment window provided by the flat top beam is larger. The circular spot resulting from the Gaussian beam forms a zigzag edge of the removed region (FIG. 29A). As shown in FIG. 29A, the alignment margin of M is reduced due to the waveform of the zigzag edge profile, and is limited to Mab.

However, the ablation region edges created using a flat top beam are straight allowing the alignment margin to stay at M. For the back contact back junction solar cells described herein, the BSF openings are formed in the BSG open area, and the contact openings are formed in the BSF area. Therefore, wider alignment margins are important because they make the BSG open, BSF, and contact areas smaller. Thus, it reduces electrical shading and improves solar cell performance.

Since the overlap of the square or rectangular flat top beam can be reduced in the x and y directions with large area removal or doping, the throughput is significantly enhanced. Also, the throughput is further increased because the size of the square or rectangular flat top can be increased without causing excessive zigzag of the perimeter. Table 1 shows the reduction in the number of scans required to open the 150 um width line, as used for the removal of the base region by removing the BSG film.

1 shows the throughput of a Gaussian to flattop beam for forming a 90 μm wide base opening. The results in Table 1 are shown graphically in Figure 29E.

fair Line width ( um ) Spot
size ( um ) Overlap% scan
pitch ( um ) Per line
Scans Remove BSG with Gauss 150 30 50 15 9 Remove BSG with Flat Top 150 30 20 24 6 Remove BSG with Flat Top 150 60 20 48 3

FIG. 29E shows a flat top beam (60 μm flat top beam area profile compared to a Gaussian beam (30 μm flat top beam area profile is shown in FIG. 29C) for a high productivity laser system capable of processing four wafers at one time. Throughput advantages), as shown in FIG. 29D. To further reduce the cost, for example, two lasers can be used with each laser beam split into two. However, various modifications of these flat top laser beam hardware and manufacturing schemes are possible.

In addition, when using a flat top beam, since the overlap is significantly reduced in the x and y directions, the laser induced damage of silicon is greatly reduced compared to the Gaussian beam.

In addition, similar throughput advantages include opening the oxide regions of the BSF, doping the BSF regions using overlay PSG, forming base and metal contact openings when they are in line contact, and flat top beams for metal removal separation lines. In the case of using-with the simultaneous advantage of reduced silicon damage-can be drawn. In addition, using a flat top beam provides the advantage of an increased alignment window of the BSF openings inside the BSG openings and the contact openings inside the BSF. In addition, the flat top laser processing method can increase throughput to form a back surface field. For example, the backside field can be formed by doping an open base region with an n-type dopant, such as phosphorous, as described. For this process, the base is coated with a phosphorus-doped silicon oxide (PSG) layer, and doping is done by irradiating this area with a laser beam. Evenly doping this area with a Gaussian beam requires an overlap, but the overlap can be minimized or completely reduced using a flat top beam. In addition, using the flat-top laser beam, in conjunction with the base and emitter region removal and back field removal described herein, provides the advantage of substantial throughput and reduced damage since the required overlap is reduced. To form the back field, the beam needs to be a flat top beam in one direction that is typical for the scan, but it should be noted that the direction of the scan may be Gaussian. beam need to be flat top beam only in one direction-normal to the scan, whereas it may be Gaussian in the direction of the scan). This type of beam is called a hybrid flat top beam.

Importantly, to form a separate base or emitter contact, overlap is not a problem, but unlike Gaussian, since there is no high intensity peak in the center of the beam, silicon damage is still reduced using flat top beams (Fig. As shown in 27A and 27B).

Another embodiment of the present invention improves the passivation properties of crystalline semiconductor solar cells, generally dielectric-coated surfaces, thereby improving the conversion efficiency performance of crystalline silicon solar cells, more particularly silicon nitride (SiN) -coated surfaces. It relates to using laser annealing to improve. Improved front passivation properties are indicated by reduced front recombination rate (or reduced FSRV) and increased efficient minority carrier life. This technique is particularly useful for high efficiency back junction, back contact cells with interlocked metallization (IBC), and SiN-coated front annealing can be used to simultaneously anneal emitter and base metal contacts to the solar cell back side, Lower specific contact resistivity and improve solar cell fill factor (FF). The laser annealing method of the present invention is directed to crystalline semiconductor solar cells using a wide range of semiconductor absorber layers, ie thick wafer-based solar cells, such as crystalline silicon wafer solar cells having a wafer thickness of 10 to 100 microns. Applicable Also, more specifically, the non-contact laser annealing process and method thereof of the present invention are applicable to ultra-thin (eg, a crystalline semiconductor layer of several microns to ˜50 microns thick) crystalline silicon solar cells, and unsupported cell mechanical processing May cause breakage. It is also an in-line replacement for batch furnace annealing processes. Laser annealing processes and methods can be used immediately after the last step of the cell manufacturing process flow or after deposition of the front passivation and antireflective coating (ARC) layers. The process and methods of the present invention enable the formation of high quality surface passivation and ARC layers using low temperature, low thermal budget deposition processes for passivation and ARC layers such as silicon nitride deposited by low temperature PECVD.

Passivation of the surface of phosphorus rich N + emitters with silicon nitride to standard front contact solar cells with p-type silicon bulk (or p-type base) is well known and widely used in the solar industry. . Although the SiN film acts as an antireflective coating to reduce optical reflection loss and increase solar capture, it plays an important role in passivating the surface of phosphorus rich N + emitters by well known hydrogenation processes. Hydrogen released from the hydrogen containing SiN layer satisfies open bonds on the silicon surface, reducing the rate of surface recombination or the number of carriers by this dangling bond site. For cells made from polycrystalline or polycrystalline silicon, this hydrogen provided by the SiN layer reacts further with impurities, defects in the bulk of the silicon wafer, and eliminates grain boundary trap sites. Thereby reducing total minority carrier recombination and increasing efficient minority carrier life in bulk of the material.

The release of hydrogen and thus the surface and bulk passivation of silicon are generally obtained during the so-called "metal firing" process in the standard front junction / front contact solar cell manufacturing process flow which is now widely used in the solar cell manufacturing industry. The screen printing metal firing process consists of multi-stage heating of solar cells using a carefully designed temperature and time sequence with a final dwell at about 850-900 ° C. before the desired cooling sequence. This firing cycle is optimized after careful experiments. Since hydrogen is a small element, it can be dispersed out of the wafer if the wafer temperature is too high or the annealing time is too long. Hydrogen passivation, on the other hand, can be unsatisfactory if the temperature is too low or the annealing time is too short. Thus, hydrogen passivation phenomena are the subject of intense research and research in the solar cell industry and are considered science and technology by many (as there are many areas not fully understood). It is clear that a process that can provide a high degree of control is desirable.

For standard mainstream front contact solar cells with p-type silicon bulk (or p-type boron-doped base) and n + in-doped emitters, the front contact surface is contacted with silver, but the backside is screen printed with a blanket layer or Contact is made of aluminum through which openings are made in the backside dielectric face to produce a selective contact. In order to obtain a low resistance contact, the mixing of silver with silicon at the front and aluminum at the back is enhanced during the metal firing process described above. Based on the above description of the metal firing process, the implementation of low resistance contact and high FF of the solar cell is complicated. Again, a process is desired that can provide a high degree of control.

In addition, the same metal in contact with the n ⁺ and p ⁺ contacts on the back side, all back contact, back junction solar cells using aluminum, cannot be heated too high, such as doping of n + contacts by aluminum, Dopants increase contact resistance, lowering the charge rate of the battery. In addition, overheating of aluminum above 450 ° C. can cause deterioration of the optical reflectivity of aluminum (thus increasing optical loss of the infrared photons of the cell). The controlled low temperature heating of the contact, preferably in the range of 200-450 ° C., of aluminum making intimate contact with silicon by reducing and absorbing oxides at the silicon surface is highly desirable.

Where the front surface or sunnyside of the solar cell is substantially uniformly or in selected areas, where the front or photovoltaic side of the solar cell is substantially uniform or where the semiconductor (eg silicon) is selectively heated in a selected area irradiated with a laser beam. irradiated with the laser beam, selectively heating the semiconductor (eg, silicon), hydrogen atoms are released from SiN to efficiently passivate the silicon surface, reduce surface state density, and reduce front recombination rate (FSRV) A method for increasing the effective minority carrier lifetime of the solar cell is disclosed. In addition, the process and method of the present invention can reduce bulk tap density and improve bulk minority minority carrier life. One embodiment of the described method is based on using a pulsed laser source having a wavelength smaller than the semiconductor (eg silicon) bandgap. In this embodiment (e.g., using a pulsed green or UV laser source of crystalline silicon surface annealing), the front side is selectively heated using pulsed laser source irradiation, but the rear side of the cell is substantially cooler than the front side of the cell. maintain. Another embodiment of the described method is based on using a pulsed laser source having a wavelength similar to or greater than the semiconductor bandgap. In this embodiment (eg, using a pulsed IR laser source of crystalline silicon surface annealing), the front side is heated using pulsed laser source irradiation, while the back side of the cell is heated and annealed. Using this other embodiment, the laser beam is transmitted to the back side of the solar cell which simultaneously heats the Al / silicon contact to reduce contact resistance and improve overall cell charge rate and efficiency. The laser annealing process and method of the present invention can be performed immediately after the formation of the passivation / ARC layer at the end of the manufacturing process flow of the solar cell or before the cell is tested and sorted for module packaging. Alternatively, the laser annealing process and method of the present invention may be done through the front glass cover of the module assembly after assembling and packaging the cell in the PV module. In this case, the wavelength needs to be used to be able to pass through the glass like infrared.

When the laser annealing process is optimized (including laser source wavelength, pulse width, power, etc.), the passivation layer (e.g. PECVD SiN layer) does not degrade during this process, so sunlight does not have a significant optical absorption loss, resulting in an antireflective coating. It is important to be able to pass. In addition, the surface texture is not affected, so light capture is not reduced. It is clear that the shape of the pulsed laser source and the laser process parameters must be carefully chosen to meet all these requirements.

The laser pulse length should be long enough so that there is no nonlinear optical interaction with the passivation / ARC layer (such as the SiN layer), so that the passivation / ARC layer is not affected. Although a laser or continuous wave having a pulse length of 1 nanosecond to microsecond can be used in the present invention, the choice depends on the depth at which heat penetration is desired. Using shorter pulse lengths, heat is limited to shallow depths. In addition, the wavelength should be selected based on the depth of the semiconductor (eg crystalline silicon) required to be heated. For the use of single crystal solar cells, front-side passivation is required to be improved, and green wavelengths may be more suitable. For applications where improved bulk silicon passivation is needed and / or back contact annealing is desired, the IR wavelength may be more suitable. It is clear that the range and wavelength of the laser pulse length can be used based on the desired use.

The process of back contact cells with interlocked metallization, also called NBLAC cells, is disclosed in the related application (see, eg, US patent application Ser. No. 13 / 057,104).

30 shows one embodiment of an NBLAC process flow, and FIG. 31 is a schematic of a cross-sectional view of the cell (back plate not shown for clarity). The low temperature front passivation / ARC: PECVD (silicon nitride) + laser annealing process step of FIG. 30 involves depositing SiN at a lower temperature than used in industry (<350C). The surface is then subjected to pulsed laser irradiation, resulting in preferential silicon front side annealing which results in improved passivation of the silicon surface with hydrogen from SiN. In particular, the laser annealing processes and methods of the present invention enable the formation of high quality passivation and ARC layers deposited at temperatures as low as 90 ° C., more generally at deposition temperatures in the range from 90 ° C. to 250 ° C.

In some embodiments, the SiN to be annealed may contain a desired amount of phosphorous dopant. In this case, the annealing step causes silicon doping with phosphorus. This process is discussed below in connection with FIG.

In addition, a SiN bilayer stack, silicon oxide (SiO ₂ ) phase on SiN, silicon oxynitride (Si _x O _y N _z ), or silicon carbide (Si _x C _y ) monolayer or amorphous silicon (α-Si) An SiN bilayer stack, or a SiN bilayer stack on silicon oxynitride, can be used for silicon surface passivation. For example, it is known that an amorphous silicon layer can passivate a silicon surface well. However, in current industrial processes, there is a need for surface cleaning of silicon and process optimization of α-Si deposition processes. Laser annealing of α-Si films coated with hydrogenated SiN can activate hydrogen in SiN and radical improvement in passivation, as measured by substantially increased effective minority minority carrier lifetime and substantially reduced frontal recombination rate. Can lead.

30 The pulsed picosecond laser ablation of Al for interdigitated cell base & emitter for PVD Al / NiV / Sn contact & backside enhanced BSR stage and interlocked cell base & emitter Al lines. Al lines step forms metal contacts to the base and emitter on the backside of the solar cell. This contact is shown in cross section in FIG. It is evident that the laser beam penetrating the backside of the silicon film anneals the backside contact simultaneously, resulting in reduced contact resistance and increased charge rate of the solar cell.

The results obtained using laser annealing are shown in FIG. It can be seen that an effective lifetime improvement of up to 100 times is obtained on the low temperature deposited passivation layer of SiN without resorting to hot metal firing. deposited passivation layer of SiN without resorting to high temperature metal firin). In the NBLAC process, thin epitaxial silicon is supported on the backplane. If this backplane cannot tolerate high temperatures, the SiN deposition temperature is reduced to facilitate the epitaxial / backplate assembly process and to process the integration applying the thermosensitive backplane assembly. For such thermosensitive backplanes, with a suitable choice of laser pulse length and wavelength, heating can be limited to the front side of the silicon, but the laser annealing is maintained because it keeps the back side of the silicon within an acceptable value for the backplane. Very suitable

The non-contact laser annealing process is very suitable for NBLAC cells using epitaxial films with thicknesses in the range of about several to 50 microns, which are weak to handle.

For improved throughput and improved process control, the laser source used in the present application provides relatively uniform beam power at the top-hat profile (at least 100 microns or more) in order to reduce the overall surface irradiation scan time. Having). It also eliminates the possibility of damage in the beam overlap area.

This laser annealing process is an attractive alternative to furnace annealing because it can be an in-line cost efficient process.

According to another embodiment of the present invention, electrically insulated patterning and selective laser ablation, such as chemically-vapor-deposited silicon oxide or thermal growth on silicon, is a crystalline silicon solar cell process to obtain relatively high cell efficiency values. Used for flow In such applications, any substantial elimination-induced damage can lead to increased minority carrier recombination losses, leading to further losses in cell conversion efficiency, so that no damage is introduced to the underlying silicon substrate or, to a large extent, negligible. Damage is introduced. The solar cell semiconductor (such as silicon) surface provides a novel scheme to ensure that the surface is not damaged during pattern-selective removal of dielectric (such as silicon oxide) overlays. The present invention involves introducing a thin intermediate layer of silicon that interrupts the laser beam from reaching the silicon substrate. This thin intermediate silicon layer may be located close to the underlying silicon surface, separated only by a thin buffer of silicon oxide. The layer of oxide on this intermediate silicon layer is removed by a laser beam that interacts and separates the silicon oxide-intermediate silicon layer interface. A very thin (eg, 3 nm to 100 nm or 3 to 30 nm in some embodiments) layer of silicon oxide under this intermediate silicon layer is responsible for the significant damage-causing effect of laser action at this interface leading to the silicon substrate. Suppress The intermediate silicon layer is then oxidized (using either a thermal oxidation process or an oxidation annealing process) to remove unwanted interactions in subsequent laser processes. This scheme is particularly suitable for use in all back contact back junction solar cell designs, and laser ablation of dielectric layers such as silicon oxide is used several times like NBLAC solar cells.

In one embodiment of the process flow, the oxide removal process is performed three times to form an oxide pattern, namely BSG (or BSG / USG stack) removal to describe the emitter and base region, USG (to define the base region) Or PSG / USG stack) removal, and finally removal of PSG (phosphosilicate glass-oxide) to open contact to the base and removal of BSG / USG / PSG removal to open contact to the emitter region. . The technique described herein can be usefully used in the first step of the removal of the BSG layer to define the patterned emitter and base region (for solar cells using n-type base). If desired, this technique can be further used during the removal of the USG to define the opening of the N ⁺ base region. (Their polarity will be reversed in solar cells using p-type bases.)

33A shows the process flow for all back contact solar cells including oxide removal in three different steps. FIG. 33B shows a slight variation of the BSG / USG (USG undoped silicate glass or undoped silicon oxide) deposition step, where the ultra-thin α-Si layer is a thin USG layer prior to deposition of the remaining BSG / USG stack Is deposited on top of (in some embodiments in situ within the same APCVD BSG deposition facility situ )). During the laser ablation process, the laser beam separates the BSG / α-Si interface to remove the BSG / USG stack. This thin layer of silicon is oxidized during subsequent processing as shown in FIG. 33B.

33C and 33D show other variations to the process flow of FIGS. 33A and 33B, and the USG deposition step is modified to include the deposition of an ultra-thin α-Si layer on top of the ultra-thin USG layer prior to the deposition of the thicker USG layer. During laser ablation, the laser beam separates the upper USG / α-Si layer to remove the upper USG layer. As before, this thin layer of silicon is oxidized with α-Si previously deposited as described above during the subsequent step as shown in FIG. 33D.

FIG. 34 schematically illustrates a standard oxide removal process using a laser beam having a pulse width in the range of several picoseconds. The interface to be acted upon by the laser is the surface of the silicon substrate, which can be damaged if the correct pulse energy is not used. 35 shows a schematic of depositing an ultra-thin amorphous silicon layer after deposition of an ultra-thin USG layer. As shown in FIG. 35B, the interface for laser action is the BSG / amorphous silicon interface. This interface acts as a removal stop layer, protects the crystalline silicon surface from laser irradiation, prevents or suppresses possible damage to the crystalline silicon surface, thereby making the battery more efficient.

The complete stack USG / α-Si / BSG / USG can be deposited in-situ using APCVD for solar cell manufacturing. The APCVD device may be a high productivity in-line APCVD device having multiple sequential in-line deposition regions to enable deposition of the entire stack into a single piece of APCVD device. Using an APCVD instrument, a thin undoped silicon layer is deposited in one of the APCVD deposition regions (second region after deposition of the first USG layer) using silane and argon (or silane and nitrogen) at temperatures below about 500 ° C. Can be. Alternatively, it can be deposited using PECVD technology. Wide thicknesses of thin USG and thin α-Si can be used based on the specific process flow. In general, USG in contact with the crystalline silicon surface may range from 3 nm to 100 nm, while the amorphous silicon layer may range from 3 nm to 30 nm. However, as described above, thicknesses outside these ranges will also be possible if the remainder of the process flow is varied to apply to the thickness of these films.

In addition, the same scheme can be used to open the oxide layer to the base region that is phosphorus doped to form the N ⁺ layer as needed. In this case, the process flow is changed to ensure the oxidation of this α-Si layer.

According to another embodiment of the present invention, laser doping is used to form the front field (FSF). The use of a front electric field away from the p / n junction to reduce minority carrier recombination loss and increase electrical current collection in solar cells is well known. Doping the substrate with the opposite polarity to the substrate is used to form an electrical p / n junction, but the surface of the remaining silicon can be doped with the same polarity as the substrate but with a high concentration of dopant. This creates a built-in electric field due to the degree of doping concentration gradient that 'repels' minor carriers from the base contact (or from the surface state), collecting to benefit emitter contact. Can be gainfully collected. This field is useful for the front side of the back junction back contact solar cell, and the p / n junction is on the back side of the wafer. This front electric field increases current collection at emitter contact on the back of the solar cell. This is done by suppressing the loss of minority carriers at the front surface recombination sites (such as the surface state at the front passivation layer / silicon interface).

The front field (FSF) is a back junction using a particular pulsed laser doping technique, including using a passivation layer such as silicon nitride containing a dopant of the desired polarity (e.g., FSF on n-type base and boron FSF on p-type base). A textured crystal (monocrystalline in some embodiments) of a back contact solar cell is created on the front side of a semiconductor (silicon in some embodiments). In this case, the front side of the silicon needs to be heated to a temperature sufficient for dopant dispersion and dopant activity in the semiconductor layer. Again, with proper selection of laser parameters such as pulse length and wavelength, the front side of the semiconductor is selectively heated to the desired temperature (or at least with reduced heating) without noticeable heating of the solar cell bulk or back side. See FIG. 36. This makes it possible to use a thermal back plate for supporting the thin silicon film.

Alternatively, the present invention provides a thin (eg 2 nm to 20 nm) amorphous silicon layer (or alternatively semistoichiometric silicon-rich oxide layer) containing the desired dopant underneath the ARC layer and a major passivation such as PECVD silicon nitride. Or a semistoichiometric silicon-rich silicon nitride layer) and then laser doping the front of the solar cell such that the silicon surface is selectively doped. Again, the temperature of laser doping needs to be high enough to cause the electrical activity of the dopant and the dispersion of the dopant in silicon. Amorphous silicon is epitaxial crystallized on single crystal silicon during cooling. In addition, this allows the quality of the front passivation to be substantially improved (in addition to the formation of a thin FSF layer, through effective heating and activation of the passivation layer, a substantial reduction in the rate of front recombination and a substantial decrease in surface state density). See FIG. 37. Also, as described above, with proper selection of laser parameters, heat penetration into the backplane can be prevented, so that the thermal backplate can be used to support the thin silicon film.

In addition, the techniques of the present invention can be used to form emitters and BSFs in front contact cells. BSF is used on the front of the cell (or alternatively on emitter contacts collected on the back side of the back contact solar cell) on the back of the front contact solar cell, which increases current collection by repelling minority carriers, which are emitter contacts Is collected by.

In some embodiments, the heating on the opposite side of the solar cell is limited to temperatures below 500 ° C., and in some embodiments below 150 ° C.

For this pulsed laser doping application, the laser pulse length is sufficiently long that there is no nonlinear optical interaction with SiN, so that the SiN ARC and passivation properties are not degraded. Lasers or CWs (continuous waves) with pulse lengths> 1 nanoseconds to microseconds may be suitable for such applications. Some embodiments of the present invention use pulsed laser sources with pulse lengths ranging from 10 nanoseconds to several microseconds; Some embodiments use a range of about 100 nanoseconds to 5 microseconds. The wavelength should be selected based on the depth of silicon required for doping. In the case where the silicon film is supported on the thermosensitive backplane, there is an additional requirement that heat is limited to the front side of the solar cell. To limit the heated zone to stay close to the surface being irradiated with the laser and still dope it sufficiently, the NIR (near infrared) wavelength is applied to this application. It can be applied to, but the green wavelength is more suitable. It is clear that a range of laser pulse lengths and wavelengths based on the desired application is suitable and can be used as various embodiments of the present invention.

In addition, SiN doped with phosphorus, SiN stack on amorphous silicon (α-Si), or Si-rich SiO _x or Si-rich SiN _x SiN stacks of phases can be used for silicon surface passivation. In this case, an amorphous silicon layer (or Si-rich SiO _x or SiN _x) Layer) is doped in-situ during laser deposition (eg by PECVD) with the desired amount of phosphorus. Laser annealing of a phosphorus-doped α-Si (or Si-rich SiO _x or SiN _x ) film applied with hydrogenated SiN improves the passivation of the silicon surface with hydrogen in PECVD SiN while also causing doping of silicon with phosphorus . The process sequence is shown graphically in FIG. 37 depicting FSF formation using a phosphorus-doped α-Si underlayer. Or a doped Si-rich SiO _x or SiN _x underlayer may be used.

Laser doping technology can be used to form FSF using doped glass (on n-type base), boron doped glass (on p-type base), which is phosphorous on a thin crystalline semiconductor film supported on a thermosensitive backplane, Heating will be limited to the front side using appropriate selection of laser wavelength and pulse length.

This process using pulsed laser doping is useful for the application, and the back contact cell with the backplane cannot withstand high temperatures after attaching the thin cell to the backplane, so the opposite side of the entire solar cell substrate and / or solar cell (I.e. solar cell back side in the case of front side passivation improvement and FSF formation) a conventional high temperature doping process cannot be performed.

In addition, this technique provides an FSF to the epitaxial film, since in-situ growth of the FSF during epitaxial deposition is not useful because the doped surface will be lost during texturing. This is the case for NBLAC batteries.

This technique has been described for the application of NBLAC cells with n-type base substrates. For the p-type base substrate, an amorphous silicon film containing boron can be used to form the FSF.

This technique can also be used to form emitters using a phosphorus containing oxide film (PSG) or boron containing oxide (BSG) on each of the p-type and n-type substrates.

The non-contact pulsed laser doping process is well suited for back contact solar cells using epitaxial films of about 80 microns or less that are weak to process. In addition, the laser doping process is an attractive alternative to furnace doping because it can be an in-line cost effective process.

According to another embodiment of the present invention, laser annealing is used to apply a silicon substrate with aluminum to selected regions, providing an acceptor-rich p ⁺ doped region of the crystalline silicon solar cell. This technique is particularly useful for IBC cells, and the emitter contacts can be selectively applied with aluminum by selectively laser annealing the emitter contacts in contact with the deposited aluminum layer. The same scheme can be applied to obtain selective emitters in back junction contacted cells using n-type silicon as the base. Other applications of this technology include using a p-type substrate (or p-type base) to provide the back field of the solar cell in front contact.

The present invention includes a laser process that can provide highly doped selective emitter contacts to these and other back contact cells with interdigitated metallization.

Doping silicon with aluminum to obtain acceptor rich (p ⁺ or p ⁺⁺ ) regions is well known in solar cell manufacturing techniques. For a standard front contact cell using p-type silicon, the back of the cell is screen printed with aluminum paste. During plastic annealing at a moderately high temperature, aluminum dissolves the silicon layer in contact. During cooling, the aluminum-rich silicon layer is precipitated to p-type (p ⁺ ) because aluminum acts as an acceptor or p-type dopant in silicon. This p-type doped p ⁺⁺ surface layer acts as the backside field to avoid minority carriers from the backside to the front face and they are collected in emitter contact. This increases the solar cell's current output (J _SC ) and efficiency. In addition, the Al / Si contact resistance is reduced, improving the charge rate and further increasing the solar cell conversion efficiency.

38A and 38B diagrammatically illustrate the disclosed laser scanning schemes. In FIG. 38A, only the emitter region is scanned to show a laser beam of the appropriate size and intensity used to heat the aluminum in contact with the emitter through an open contact to the dielectric, and FIG. 38B shows selective emitter formation after laser scanning. Indicates. When the metal and silicon in contact with the metal are heated to a eutectic temperature of Al-Si of at least 577 ° C., aluminum dissolves the silicon, and upon cooling below this temperature, the Al-rich silicon layer precipitates. This layer is deposited epitaxially on the silicon substrate and is free of crystal defects. This is the same mechanism as providing Al-BSF in standard Al paste printed cells.

FIG. 39 shows a P ⁺⁺ selective emitter with aluminum-saturated silicon formed by selective laser scanning only on the emitter region.

The mechanism of Al-rich layer formation can be understood with the aid of the Al-Si phase diagram shown in FIG. 40. The eutectic at 577 ° C. is aluminum with 12.6% silicon dissolved therein. At higher temperatures, more silicon is dissolved. During cooling, the epitaxially deposited silicon is saturated with Al by 1.6%. This Al-saturated silicon is P ⁺⁺ doped and provides low resistance contact to the emitter, but minority carrier absorption is suppressed in this area (providing selective BSF in the contact area).

It is clear that the same scheme can be used to obtain selective emitters for front contact cells using n-type silicon substrates and having p ⁺ back emitters. The schematic can be understood by following the schematic of FIG. 41.

As is well known, in the case where the entire surface aluminum contact of the base on the back side is replaced by localized contact, a significant efficiency improvement is obtained for a standard front contact cell using a p-type silicon base. The efficiency is further increased when localized contact is provided to the BSF area. The scheme of laser scanning described can be used to provide Al BSF as shown in FIG. 42. This mechanism has been described above.

Those skilled in the art will recognize that the described embodiments pertain to a wide range of areas other than the specific embodiments described above.

The foregoing description of the examples has been provided to enable any person skilled in the art to make and use the claimed subject matter. Various modifications to these examples are readily apparent to those skilled in the art, and the generic principles described herein may be applied to other embodiments without innovative capabilities. Thus, the claimed subject matter is not limited to the examples described herein, but is consistent with the widest scope retaining the principles and novel features described herein.

All additional systems, methods, features, and advantages included in this description are intended to be within the scope of the claims.

Claims (72)

As a method of removing an electrical insulation layer on a semiconductor substrate:
The method comprises:
providing a semiconductor substrate having n-type doping;
Depositing a first relatively thin layer of undoped glass or undoped oxide on a surface of the semiconductor substrate;
A first relatively thin semiconductor layer comprising a material selected from the group consisting of an amorphous semiconductor, a nanocrystalline semiconductor, a microcrystalline semiconductor, and a polycrystalline semiconductor on the relative thin layer of the undoped glass or undoped oxide. depositing a layer);
Depositing a layer of borosilicate glass or a stack of borosilicate / undoped glass on the relative thin semiconductor layer; And
Selectively removing the layer of borosilicate glass or the borosilicate / undoped glass stack with a pulsed laser,
And wherein said relatively thin semiconductor layer substantially protects said semiconductor substrate from said pulsed laser.
The method of claim 1,
And further a subsequent thermal oxidation process for oxidizing the relative thin semiconductor layer.
The method of claim 1,
And the semiconductor comprises silicon.
The method of claim 1,
Wherein the thickness of the first relative thin layer of undoped glass or undoped oxide is in the range of about 3 to 100 nanometers.
The method of claim 1,
And the thickness of the first relative thin semiconductor layer is in the range of about 3 to 30 nanometers.
The method of claim 1,
After the selectively removing step,
Depositing a second relative thin layer of undoped glass or undoped oxide on a surface of the semiconductor substrate;
Depositing a second relative thin semiconductor layer comprising a material selected from the group consisting of an amorphous semiconductor, a nanocrystalline semiconductor, a microcrystalline semiconductor, and a polycrystalline semiconductor on the relative thin layer of undoped glass or undoped oxide;
Depositing a layer of undoped glass or undoped oxide on the second relative thin semiconductor layer; And
Selectively removing the layer of undoped glass or undoped oxide with a pulsed laser,
And the second relative thin semiconductor layer substantially protects the semiconductor substrate from a pulsed laser.
The method according to claim 6,
And further a subsequent thermal oxidation process for oxidizing the relative thin semiconductor layer.
The method according to claim 6,
And the semiconductor comprises silicon.
The method according to claim 6,
Wherein the relative thickness of the undoped glass or undoped oxide ranges from about 3 to 100 nanometers.
The method according to claim 6,
And the thickness of the relative thin semiconductor layer is in the range of about 3 to 30 nanometers.
The method of claim 1,
Wherein the pulse length of the laser is about 200 picoseconds or less and the wavelength is about 1064 nanometers or less.
The method according to claim 6,
Wherein the pulse length of the laser is about 200 picoseconds or less and the wavelength is about 1064 nanometers or less.
The method of claim 1,
Further comprising process flow steps to fabricate a thin single crystal semiconductor solar cell.
The method of claim 13,
And the thin single crystal semiconductor solar cell comprises a thin single crystal silicon layer having a thickness in the range of 10 to 100 microns.
The method of claim 13,
The thin single crystal semiconductor solar cell comprises a back-contact / back-junction solar cell.
The method of claim 1,
Further comprising a process flow step for fabricating a crystalline semiconductor based photovoltaic solar cell comprising all back-contact back-junction solar cells.
The method according to claim 6,
And a process flow step for manufacturing a crystalline semiconductor based photovoltaic solar cell comprising all back contact back junction solar cells.
The method of claim 1,
Wherein said removal is used to fabricate openings to delineate all back contact, base and emitter regions of back junction solar cells.
The method according to claim 6,
Wherein said removal is used to fabricate openings to delineate all back contact, base and emitter regions of back junction solar cells.
17. The method of claim 16,
And the crystalline semiconductor photovoltaic solar cell comprises an epitaxial silicon thin film solar cell.
18. The method of claim 17,
And the crystalline semiconductor photovoltaic solar cell comprises an epitaxial silicon thin film solar cell.
17. The method of claim 16,
Wherein the thickness of the epitaxial thin film solar cell is in the range of about 10 to 100 microns.
18. The method of claim 17,
Wherein the thickness of the epitaxial thin film solar cell is in the range of about 10 to 100 microns.
21. The method of claim 20,
And the epitaxial thin film comprises a substantially planar epitaxial film formed through an epitaxial silicon liftoff process.
21. The method of claim 20,
Wherein the front surface of the epitaxial thin film comprises a three-dimensional pyramid or prism formed through a textured template liftoff process.
The method of claim 21,
And the epitaxial thin film comprises a substantially planar epitaxial film formed through an epitaxial silicon liftoff process.
The method of claim 21,
Wherein the front surface of the epitaxial thin film comprises a three-dimensional pyramid or prism formed through a textured template liftoff process.
As a method of doping a solar cell substrate:
The method comprises:
Providing a semiconductor substrate;
Depositing a passivation layer on a surface of the semiconductor substrate, the passivation layer comprising a predetermined amount of dopant, wherein the dopant comprises a p-type dopant or an n-type dopant;
Selectively heating the passivation layer through laser irradiation, wherein the selective heating causes the dopant to be dispersed into the semiconductor substrate, the dispersed dopant being a doped surface field or Forming a doped emitter region, the method comprising doping the solar cell substrate.
29. The method of claim 28,
The selective heating further results in bulk passivation and improved surface of the semiconductor substrate.
29. The method of claim 28,
The semiconductor substrate comprises an epitaxial thin film substrate having a thickness of about 5 to 100 microns.
29. The method of claim 28,
And the semiconductor comprises silicon.
32. The method of claim 31,
And the semiconductor comprises single crystal silicon.
32. The method of claim 31,
And the semiconductor comprises polycrystalline silicon.
29. The method of claim 28,
Wherein the doped passivation layer comprises a material selected from the group consisting of silicon nitride, silicon oxynitride, silicon oxide, silicon-rich silicon nitride, silicon-rich silicon oxide, and the dopant comprises phosphorus .
29. The method of claim 28,
Wherein the doped passivation layer comprises a doped amorphous silicon / silicon nitride stack and the dopant comprises phosphorus.
29. The method of claim 28,
The laser comprises a pulsed laser or continuous wave laser having a pulse length greater than about 10 nanoseconds.
29. The method of claim 28,
And the laser comprises a wavelength of about 10.6 micrometers or less.
31. The method of claim 30,
And the epitaxial film is supported on a substrate.
The method of claim 38,
The substrate is suitable for a processing temperature of up to about 300 ° C.
The method of claim 38,
The substrate is suitable for a processing temperature up to about 400 ° C.
29. The method of claim 28,
Depositing the passivation layer occurs at a temperature in the range of about 90 ° C to 400 ° C.
29. The method of claim 28,
The laser heating is substantially limited to a depth of about 20 microns or less.
43. The method of claim 42,
Wherein the laser comprises a pulsed laser having a wavelength less than about 828 nanometers.
36. The method of claim 35,
The laser annealing is performed such that the amorphous silicon is crystallized into monocrystalline silicon by epitaxy.
29. The method of claim 28,
The method is further used to form a front surface field in all back contact, back junction solar cells.
31. The method of claim 30,
And the epitaxial film comprises a three-dimensional pyramid or prism formed through a textured template liftoff process.
31. The method of claim 30,
And the epitaxial film comprises a substantially planar semiconductor layer formed through an epitaxial silicon liftoff process.
As a method of doping a solar cell substrate:
The method comprises:
providing a semiconductor substrate having n-type doping;
Depositing a borosilicate glass layer on a surface of the semiconductor substrate, wherein the borosilicate glass layer comprises a predetermined amount of boron dopant;
Selectively heating the borosilicate glass layer through laser irradiation, wherein the selective heating causes the dopant to be dispersed into the semiconductor substrate, and the dispersed dopant forms an emitter region in the semiconductor substrate. Including a method of doping a solar cell substrate.
As a method of doping a front contact solar cell substrate:
The method comprises:
providing a semiconductor substrate comprising p-type doping;
Depositing a borosilicate glass layer on a surface of the semiconductor substrate, wherein the borosilicate glass layer comprises a predetermined amount of boron dopant;
Selectively heating the borosilicate glass layer through laser irradiation, wherein the selective heating causes the dopant to be dispersed into the semiconductor substrate, and the dispersed dopant forms a surface electric field in the semiconductor substrate. A method of doping a solar cell substrate, comprising.
A method of providing a high concentration of aluminum doped p-type region in a silicon substrate:
The method comprises:
Providing a single crystal silicon substrate;
Applying aluminum metal contact to a surface of the single crystal silicon substrate;
Selectively heating the aluminum metal contact via laser irradiation such that the aluminum and a portion of the single crystal silicon substrate near the aluminum reach a temperature sufficient to dissolve at least a portion of the silicon in the aluminum; And
Cooling the aluminum and a portion of the single crystal silicon substrate near the aluminum to form an aluminum-rich doped silicon layer on the single crystal silicon substrate, providing a heavily doped aluminum doped p-type region in the silicon substrate. How to.
51. The method of claim 50,
And the single crystal silicon substrate comprises an epitaxial silicon substrate.
51. The method of claim 50,
Wherein the temperature comprises at least about 577 ° C. for eutectic aluminum-silicon melt formation.
To provide a heavily doped p-type emitter in a back contact / back junction solar cell:
The method comprises:
providing a single crystal silicon substrate having an n-type base doping;
Forming an emitter region on the surface of the single crystal silicon substrate, the emitter region having p-type doping;
Forming an aluminum metal contact on the emitter region;
Selectively heating the aluminum metal contact through laser irradiation such that a portion of the emitter region near the aluminum and the aluminum reaches a temperature sufficient to dissolve at least a portion of the silicon in the aluminum; And
Cooling the aluminum and a portion of the emitter region near the aluminum to form a heavily doped selective emitter region, providing a heavily doped p-type selective emitter in a back contact / back junction solar cell. How to.
54. The method of claim 53,
And the single crystal silicon substrate is an epitaxial silicon substrate.
54. The method of claim 53,
The temperature comprises at least about 577 ° C. for eutectic aluminum-silicon melt formation.
By providing a high concentration of aluminum doped regions in the front contact solar cell:
The method comprises:
Providing a single crystal silicon substrate having a first doped type;
Forming an aluminum metal contact on a back side of the single crystal silicon substrate;
Selectively heating the aluminum metal contact through laser irradiation such that the aluminum and a portion of the silicon near the aluminum reach a temperature sufficient to dissolve at least a portion of the silicon in the aluminum; And
Cooling the aluminum and a portion of the silicon in the vicinity of the aluminum to form a heavily doped aluminum doped region on a back side of the single crystal silicon substrate. .
57. The method of claim 56,
And the single crystal silicon substrate is an epitaxial silicon substrate.
57. The method of claim 56,
The temperature comprises at least about 577 ° C. for eutectic aluminum-silicon melt formation.
57. The method of claim 56,
The first doped type comprises n-type doping,
A portion of the silicon near the aluminum includes a doped emitter region,
Wherein the heavily doped aluminum region comprises an optional emitter region.
57. The method of claim 56,
The first doped type comprises p-type doping,
Wherein the heavily doped aluminum region comprises a backside field region.
51. The method of claim 50,
The laser comprises a pulsed laser or continuous wave laser having a pulse length greater than about 10 nanoseconds.
54. The method of claim 53,
The laser comprises a pulsed laser or continuous wave laser having a pulse length greater than about 10 nanoseconds.
57. The method of claim 56,
The laser comprises a pulsed laser or continuous wave laser having a pulse length greater than about 10 nanoseconds.
51. The method of claim 50,
The wavelength of the laser is about 10.6 micrometers or less.
54. The method of claim 53,
The wavelength of the laser is about 10.6 micrometers or less.
57. The method of claim 56,
The wavelength of the laser is about 10.6 micrometers or less.
54. The method of claim 53,
Wherein the thickness of the epitaxial thin film solar cell is in the range of about 10 to 100 microns.
68. The method of claim 67,
Wherein the front surface of the epitaxial thin film comprises a three-dimensional pyramid or prism formed through a textured template liftoff process.
68. The method of claim 67,
And the epitaxial thin film comprises a substantially planar epitaxial film formed through an epitaxial silicon liftoff process.
57. The method of claim 56,
Wherein the thickness of the epitaxial thin film solar cell is in the range of about 10 to 100 microns.
71. The method of claim 70,
Wherein the front surface of the epitaxial thin film comprises a three-dimensional pyramid or prism formed through a textured template liftoff process.
71. The method of claim 70,
And the epitaxial thin film comprises a substantially planar epitaxial film formed through an epitaxial silicon liftoff process.

Images