WO2011136659A1

WO2011136659A1 - Method and apparatus for removing a defect from a solar cell

Info

Publication number: WO2011136659A1
Application number: PCT/NO2011/000140
Authority: WO
Inventors: Lucheng Zhang; Timothy Charles Lommasson; Vaithianathan Veeramuthu; Dilip Chithambaranadhan; Anna Malou Petersson; Tobias BOSTRÖM
Original assignee: Innotech Solar Asa
Priority date: 2010-04-28
Filing date: 2011-04-28
Publication date: 2011-11-03
Also published as: NO20100616A1; TW201205664A

Abstract

A method for removing a defect 200 from a solar cell 100 wherein a chemical etchant 310 is used to etch a region of the cell until the defect has been removed from an electric circuit representing the cell. The etching is controlled by selecting an appropriate combination of etchant 310 and temperature, and the temperature is kept below a level where it may cause new and possibly more severe defects. An apparatus to carry out the method is also disclosed, in which a voltage imposed over the cell 100 and/or heater 400 controls the etching.

Description

Method and apparatus for removing a defect from a solar cell

BACKGROUND OF THE INVENTION

Field of the invention

The present invention relates to a method and apparatus for removing a defect from a solar cell.

Related and prior art

Silicon based photovoltaic devices, or solar cells, can be manufactured from solar grade bulk silicon of purity 99.999%-99.9999999% (5N-9N Si) which is wire sawed into self supporting wafers with a thickness between 150 and 330 micrometers (pm). The Si-wafers are preferably as thin as possible, often having a thickness less than 240 pm, to save solar grade Si while keeping a reasonable mechanical strength for handling. The wafers are then processed into solar cells and soldered together to form a solar cell module or solar panel typically having a sheet of tempered glass on the front, and a polymer encapsulation on the back. The glass and/or polymer encapsulation provides mechanical strength to the solar panel.

To create a solar cell, a lightly p-doped Si wafer is provided. The p-type dopant is typically boron (B) or another element from group IMA which creates an excess of holes in the lattice. The cell is then n-doped, typically with phosphorus (P) or another element from group V to create an excess of loosely bound electrons close to the front surface. The n-type dopant is diffused from the front surface, and form a p-n junction a few hundred nanometers (nm), e.g. 0.3 pm, below the front surface. This p-n junction is essential for converting photo energy to electrical energy.

Once the p-n junction is made, typically an antireflective surface coating of about 80 nm of silicon nitride (SiN) is added. SiN also prevents recombination at the surface, and gives the solar cell a characteristic bluish appearance.

The wafer is then provided with metal contacts on the front and back surfaces. On the front face, light must be allowed to enter into the cell. Hence, the front contacts are typically provided as gridlines formed in a comb-shaped pattern of fine

"fingers" and larger "busbars", e.g. by screen-printing on the front surface using a silver (Ag) paste. The rear contact may also be formed by screen-printing a metal paste, typically aluminum (Al), in a form of gridlines and busbars similar to the ones on the front surface. The rear contact may alternatively cover the entire back surface, as the back surface need not be transparent to light. The paste is then fired at several hundred °C to form metal electrodes. Silver electrodes on the front are typically 10-25 μητι thick.

In the manufacturing process, a certain number of defects must be accepted to keep the cost of manufacture at a reasonable level. Such defects may be caused by impurities and inclusions in the bulk silicon, cracks caused by wire sawing or handling, contamination during the diffusion to create a p-n junction etc. Some defects, e.g. cracks filled with metal when the metal contacts are made or metal inclusions, may form local areas or paths with low resistivity called 'shunts', which is an important class of defects. However, it should be understood that the present invention concerns any defect, regardless of cause and regardless of whether the defect causes a local decrease or increase of resistivity. The defects are usually not uniformly distributed over the entire area. A defect may be present anywhere in the material between and/or behind the gridlines of the front face contacts, as well as in the frontside and/or backside contacts themselves.

So far, the solar cell has been described in terms of its manufacture and mechanical structure. Its electrical properties are also an important background for the present invention.

When light illuminates the front surface of a solar cell, the photons excite the loosely bound electrons from the lattice in the n-type region, creating a voltage across the p-n junction. The direction of the voltage created in this manner is termed 'forward bias', and the opposite direction is termed 'reverse bias'.

The p-n junction acts as a diode. When a load is connected between the n-type front side and p-type back side of the cell, electrons flow from the n-type material, powers the load, and returns to the backside where they recombine with the holes. For this reason, the n-type material is termed the 'emitter' and the p-type material is termed the 'collector'. It is noted that the emitter, having an abundance of loosely bound electrons in the lattice, conducts electricity better than the p-type Si in the substrate below.

Some defects may cause current to flow erroneously and create hot spots, some of which may cause damage to the cell and/or the entire panel. Other defects simply reduce the efficiency of the cell. Typically, the entire cell is tested after manufacture to determine if it meets predetermined criteria for e.g. efficiency and/or electrical properties. Those cells that do not fulfil the predetermined criteria are rejected, and the functional solar cells are interconnected in series and/or parallel into modules or solar panels.

However, some cells may be repaired by removing electrical defects from an otherwise functional cell, such that the remaining cell meets the predetermined criteria for efficiency and/or electrical properties. Some common prior art techniques for shunt repair are:

Laser scribing and mechanical breaking and removal

In this method a laser is used to scribe straight lines on the back of silicon solar cells. Then the scribed solar cell is mechanically broken apart along the scribed lines into two or more smaller parts. If the locations of the scribing are appropriate, then the localized shunts are only in some of the divided parts. Those smaller parts that have no shunt can be used as good cells. However, the resulting good cells of varying shapes and sizes may cause difficulties in the later assembly of a solar module.

Laser isolation

A second method of shunt repair is proposed in the article written by M. D. Abbott, T. Trupke, H. P. Hartmann, R. Gupta and O. Breitenstein, entitled "Laser isolation of shunted regions in industrial solar cells", in Progress in Photovoltaics: Research and Applications 2007; 15: 613-620. The method involves using a laser to ablate away material to provide a closed groove around the shunt region. When the groove extends below the emitter, the shunt is isolated from the functional circuit. This may improve the conversion efficiency for the repaired solar cell, as discussed above.

A main benefit of the method is its ability to remove defects of any size and shape. However, it should be understood that there is a limit to area which is practical or economical to remove. For example, if the dimension of the shunted region is more than several square centimeters, the repaired cell still has a low working current and provides a poor current match if the repaired cell is incorporated in a module. In addition, a large isolated area not contributing to the output of a solar module is a waste of space and encapsulation material. Hence, solar cells having defects exceeding a predetermined limit may be, but does not have to be, discarded rather than repaired.

A problem with laser isolation is that new shunts may be produced along the scribing path when the laser scribes the emitter side of the solar cell. Especially when the laser passes the metal contact region, the newly produced shunts are much worse than that in the regions without metal contact. This makes laser isolation unsuitable for repairing shunts under the metal contacts.

Physical removal and replacement

A method for physically removing and replacing a shunt in a solar cell is disclosed in the in the article written by L. C. Zhang, H. Shen, Z. J. Yang and J. S. Jin, entitled "Shunt removal and patching for crystalline silicon solar cells using infrared imaging and laser cutting", in Progress in Photovoltaics: Research and

Applications 2010; 18: 54-60. In this article, the shunted region is first detected and its location is determined using infrared imaging techniques. Then, the detected shunt is physically removed using laser scribing. Finally, the removed shunt is replaced with a small good cell having the same dimensions as the removed shunted region, and the non-shunted majority cell and small good cell are electrically connected. This technique is obviously best suited to repair large area shunts in a solar cell.

The problem of producing new shunts also exists when a laser is used to physically remove the shunts rather than ablating a groove around them. In addition to increasing the cell efficiency and improving the performance in shade for the repaired cell, the replacement technology can also improve the current matching in a module.

The main objective of the current invention is thus to remove an electrical defect from a solar cell while overcoming the problems of prior art.

A further objective of the invention is to provide an apparatus using the method.

SUMMARY OF THE INVENTION

This is achieved by providing a method for removing a defect from a solar cell comprising a front side with first electrical contacts, a conversion layer converting photon energy to electric energy, and a back side with second electrical contacts, the method being characterized by using a chemical etchant to etch a region of the cell until the defect has been removed from an electric circuit comprising the first and second electrical contacts and electrical components representing the conversion layer.

One or more suitable etchants may be used to etch away the metal connectors leading current to or from the region containing the defect, and/or remove substrate material from the conversion layer to isolate or remove the defect. As long as the chemical etching is performed at temperatures below 300°C, there are no heat induced zones or significant processes to create new defects.

In a first embodiment of the invention a reverse bias is provided over the cell, and a suitable etchant is applied to the cell. The cell can be immersed in a bath of etchant, obviating a need for localizing the defect before etching. Alternatively, the defect can be localized and the etchant can be dispensed locally to a surface area over the defect. Due to the reverse bias, the defective region, in particular a defect constituting a shunt, is heated. This accelerates etching in the shunt region, thus confining the etching to the hotter region containing the shunt. Alternatively, a laser may be employed to heat a smaller region of the cell to accelerate etching. In a second embodiment, the surface of the cell is covered in a mask material being inert to the selected etchant, and holes are provided in the mask to bring the etchant in contact with the surface of the cell. The masking material confines the etchant, and thus makes it possible to use etchants having a viscosity and/or other properties that would otherwise cause them to spill over an undesirable large surface area.

In a third embodiment, the defect is localized before an etchant is dispensed though an inner pipe or pipette. An outer tubular element is provided around the inner pipe. The etchant may be confined by bringing a lower rim of the outer tubular close to the surface to be etched. A pressure less than the ambient pressure can be, but does not have to be, provided in the space between the inner pipe and outer tubular, causing an inflow of fluid into the space. Such a fluid flow may also help confining the etchant.

In a fourth embodiment, etchant solutions are used in combination with surface treatment so that liquids that are placed on the surface are sufficiently hydrophobic so that they do not spread but only contact the surface where they are placed. In this manner defects in cells can be removed by etching with liquid etchants without using a mask, an outer tubular, an inflow of fluid or other methods for confining the etchant.

The invention also comprises apparatus for carrying out the method.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be better understood from the following detailed description with reference to the accompanying drawings, in which like numerals refer to like parts, and in which:

Figure 1 a is a schematic cross section of a solar cell with an isolated shunt;

Figure 1 b shows the cell of Fig. 1 a where the shunt is completely removed;

Figure 2 shows the equivalent circuit of the solar cell in Fig. 1a;

Figure 3a is a schematic view of an apparatus for selective etching;

Figure 3b is a schematic view of an alternative apparatus for selective etching; Figure 4a is a drawing of an electroluminescence image before etching, and Figure 4b is a drawing of an electroluminescence image after etching, and

Fig. 5 shows a nozzle for confining etchant using an inflow of fluid.

DETAILED DESCRIPTION

Figure 1 a is a schematic view of a part of a typical solar cell 100, wherein a groove 210 is etched in a path around a defect 200. The groove extends from the front side, through an antireflective layer 140 consisting of SiN_x and/or SiO_x, an n-doped emitter layer 130 and into a p-doped substrate or collector 120. The emitter 130 and collector 120 form a p-n junction which is essential in converting the photon energy of incident light to electrical energy. Because the conductivity of the p- doped substrate 120 is less than the conductivity of the emitter layer 130, the defect 200 is isolated from an electric circuit comprising the first 1 10 and second 1 11 electrical contacts and electrical components representing the conversion layer. It should be understood that the groove is shown circular for illustrative purposes only, and that a real groove may have any shape.

On the front face of the solar cell, first electrical contacts 1 10 are typically provided in a comb shaped pattern of fine gridlines called fingers and larger bus bars, so that light may illuminate the active parts of the photovoltaic device. The gridlines on the front face may typically be made of silver (Ag). As can be seen from Fig 1a, the front side gridlines 1 10 contacts the emitter 130, which is a layer of n-doped Si (n-Si) having an excess of loosely bound electrons.

Underneath or behind the n-type layer is the p-doped silicon (p-Si) layer 120. This layer has an excess of vacancies in its lattice, or holes, and constitutes the collector of a diode. The layer 120 forms a p-n junction with the emitter 130 on the side facing the front of the solar cell, and contacts a second electrical conductor 1 1 1 on the back face of the cell. The second electrical conductor 1 1 1 can be made of e.g. aluminium (Al), and it may be provided as a grid similar to the front face contacts or a contact covering the entire back face. It should be understood that the figures are not to scale. In a practical cell, the entire substrate is typically 150-330 μηι thick, the gridlines 1 10 are between 10 and 25 μηι thick, and may be considerably broader compared to their height than depicted in the figures. The emitter 130 is typically about 0.3-0.4 μη thick, and is shown comparatively much thicker in the figures for illustrative purposes.

When the first and second electrical contacts are connected to a load and the cell is illuminated, incident light with sufficiently small wavelengths, i.e. photons with sufficient energy, will release or emit the loosely bound electrons and holes from the lattice in the emitter 130 and collector region 120 respectively. So the excess carriers can be separated from the majority carriers in their corresponding regions by the space charge region before they recombine. The separated electrons are free to move into the gridlines on the front face of the cell. The electrons then pass though the load and are recombined with the holes in the collector 120.

The topmost layer illustrates a passivation layer 140, typically comprising a coating of SiN_xand possibly SiO_x. The passivation layer 140 is anti reflective and also helps preventing recombination of electrons and holes on the front surface of the device. The passivation layer 140 can be composed of more than one sub layers that have different compositions. Some of these sub layers may be made of SiO_x.

From the above, it should be understood that the cell depends on a functioning p-n junction pulling electrons from the p-Si layer 120, into the n-Si layer 130 and holes in the opposite way. Both the electrons and holes are excited from the lattice by the incident light.

In Fig 1a, a defect 200 is shown in the p-n junction. This defect, which may be an inclusion extending just through the p-n junction, a crack extending though the entire thickness of the cell or any other defect, disturbs the p-n junction in a certain region and reduces the efficiency of the cell. In some cases it may also form a shunt causing a local current that is large enough to create a hot spot, which in turn may harm the cell and even the finished module. In Fig. 1a, the potentially harmful defect 200 is isolated from the rest of the cell by a groove 210 that encircles the defect 200 and extends from the front face to below the emitter 130 and into the p-doped Si-substrate having a low electrical conductivity compared to the metal gridlines 110 and the emitter 130. Hence, the groove 210 prevents current from flowing through the defect 200, and thus increases the efficiency of the p-n junction. It also prevents a shunt defect 200 from forming a hot spot.

Because it is considered sufficient to etch, at most, past the p-n junction and a few pm into the p-doped layer 120, a 'defective region' may for simplicity and without loss of generality be defined as an area on the front surface that is large enough to enclose a major part of practical defects and yet small enough to avoid that too much of the functional p-n junction is removed with the defect. Thus, in some embodiments, the 'defective region' may arbitrarily be defined as e.g. an area of about 2 mm x 2 mm, or a circle having a 2 mm diameter. In other embodiments or contexts, a 'defective region' may be defined otherwise.

Fig 1 b is a schematic view of a repaired cell where the defect 200 is etched away rather than being isolated by a path-shaped groove as in Fig. 1 a.

Fig. 2 illustrates an equivalent circuit for a solar cell. The dotted box on the right hand side of Fig. 2 shows the equivalent circuit of a functional cell.

As discussed above, when light illuminates the front surface of a solar cell, the photons excite electrons from the lattice, creating a voltage across the p-n junction, which acts as a diode. When a load is connected between the n-type front side and p-type back side of the cell, electrons flow from the n-type material, powers the load, and returns to the backside where they recombine with the holes. An ideal solar cell may thus be modeled by a current source in parallel with a diode. However, a practical solar cell model has to account for smaller or larger imperfections, so the equivalent circuit of a solar cell includes a shunt resistance RSH in parallel with the diode, and a series resistance Rs- The fundamental equation of a solar cell may be written:

where

/ is the output current flowing through the series resistance Rs,

II is the current generated by illumination, i.e. the photons

ID is the current through the diode, and

ISH is the current through the shunt resistance RSH.

Alternatively, a similar equation for voltages Vor an equation using current densities J ean be used.

The direction of current and voltage generated by the illuminated cell is called 'forward bias', and the opposite direction is called 'reverse bias'.

As discussed in the introduction, a certain level of impurities and defects resulting from the manufacturing process is allowed because it would be too expensive to reduce them below that certain level. Some of these defects have low electrical resistance and so constitute shunts. It should be understood that the term 'shunt' as used herein thus includes any defect affecting the shunt resistance of the equivalent circuit, and that a shunt therefore is not a particular defect like a crack or impurity. However, the origin of the shunt can be material- or process related. For example, a crack caused by wire sawing or handling may be filed with metal during screen printing and sintering. Other major sources of shunt defects are inclusions of SiN, SiC or a metal. Such shunts reduce the overall RSH of the solar cell. This in turn increases the corresponding current ISH and decreases the output current / as shown by equation (1 ). Hence, the overall efficiency of a solar cell depends on the number and severity of the shunts caused by such defects.

Severe shunts, causing standard devices with less than 1 Ω shunt resistance for a 156 mm by 156 mm cell, have poor fill factors and hence produce low output current. Standard devices with shunt resistance RSH between 1 Ω and about 4Ω do not have significantly affected electrical output, but can create a hot spot in a module. The relatively high voltage generated by an array of cells in a module can drive a reverse current through the low shunt resistance and cause a hot spot. Hot spots are potentially destructive for modules and systems, e.g. by destroying the lamination material in a solar module.

The part enclosed by a dotted line on the left hand side of Fig. 2 is a similar equivalent circuit representing a shunt defect. If the defective region is connected to the cell, a hot spot may be caused by a low resistance represented by Rs_∑- From Fig. 2, it is readily seen that when a reverse bias is provided over the shunted region, most of the current passes through the shunt resistance Rs2 rather than through the diode in the shunted region. This represents the cause of a local heating or hot spot in the shunted region, especially when a reverse bias is imposed.

Finally, in Fig. 2, the shunted region represented by the left hand dotted box is disconnected from the functional cell represented by the right hand dotted box. By this, the remaining shunt resistance RSH is brought within acceptable limits, the hot spot caused by the shunt is removed, and the solar cell becomes functional.

When the solar cell is electrically biased, forward or reverse, more current flows through a shunt defect 200 than through other regions of the p-n junction. So, as shown in Figures 1 a and 1 b, etching away the shunted region from the top by removing part of the antireflection layer^"! 40, emitter layer 130, and collector layer 120 in the region of the defect 200, typically to a few μητι below the p-n junction, can completely remove the shunt from the solar cell.

As noted in the introduction, laser ablation has been proposed to remove the material. However, heating the material by a laser introduces thermal stresses that may cause new and potentially more severe shunts. In addition, excessive heat may cause undesired diffusion or other changes to the structure. Here, chemical etching is proposed as an improved method for removing defects. Chemical etching

The wet chemical reaction processes described herein preferably occur at a temperature below 300 °C in the reaction area. When the temperature in the reaction area is below 300 °C, there is no atom diffusion in the solid silicon.

Hence, there is no doping process, and no degradation in the quality of the p-n junction in the reaction region. Therefore, no new shunts are produced during the chemical etching process. However, acceptable results may be achieved with temperatures above this level, e.g. to about 350 °C, for a shorter period of time.

Shunt detection

In some, but not all, of the embodiments below, a defect must be detected and localized before etching. This may be achieved using electroluminescence, photo- luminescence or thermograph imaging techniques, by using a liquid crystal sheet, by processing information from high resolution l-R Images of a cell under reverse bias, EL Image of a cell under forward bias or any other shunt detection method or by any combination of methods known in the art. Shunt detection is not part of the present invention.

Masking

In some embodiments, it may be necessary or desirous to confine an etchant using a mask. The masking can be achieved for example by a coating of fluori- nated polymer, chemical resistant wax, by lithography, by an electro mechanical device, by a ceramic coating, by vacuum sealing or by a mechanical device.

Masking as such or masking materials are not part of the present invention, and hence are not described in great detail here. Any masking material or masking technique known in the art can be used with the present invention.

Etchants

A chemical etchant can be used to etch a region of the cell until a shunt defect has been removed. The front face gridlines are made of silver, and hence any etchant used for etching silver, e.g. nitric acid, ferric nitrate, piranha solution, dilute aqua regia, CR-7 or a mixture of relevant inorganic and organic compounds can be used to isolate the cell by etching away the metal contacts on the front face. Nitric acid (HN0₃) is preferred as it does not etch the Al-contacts at the back of the device, it does not disturb the antireflective coating on the front surface, and hence does not deteriorate the optical properties of the cell, and it etches Si at a considerably slower rate than it etches Ag.

Experiments have shown that the etching rates tend to increase with concentration and temperature as expected, and further that better results and more controllable processes tend to be achieved from higher temperatures and moderate concentrations than by lower temperatures and higher concentrations.

For etching Ag gridlines, concentrations of HNO3 in the range 10-90wt%t may be used at temperatures from below room temperature to about 350 °C. At this temperature, undesired diffusion processes and potentially harmful thermal stresses tend to occur.

A concentration range of 45-50wt% HN0₃ at temperatures above 70 °C has been found to produce etching times for a 15-16 μηι bus bar of a few seconds. It should be understood that similar etching times may be achieved using higher concentrations at lower temperatures, and that the etching time can be adjusted by using a fixed concentration and adjusting the temperature.

Similarly, HF, H3PO4 or another composition can be used to etch the SiN_x passivation layer. An HF solution can be used to etch a SiO_x layer.

To etch crystalline silicon, a variety of etchants may be used. Examples include KOH, NaOH, mixtures of HF and HN0₃, XeF₂, mixtures of phosphoric and sulphuric acid, etc.

In a preferred embodiment, KOH may be used to etch through the relatively conductive emitter of n-doped Si close to the surface and into a p-doped Si substrate below. Any solution of 10-90 wt% KOH can be used. As with the HNO₃ solution above, the concentration of KOH may be selected from a narrower range. Concentrations in the range 30-40 wt% KOH at temperatures above 100 °C have been found favourable in instances where etching rates in the range 2-8 μηι per minute are desired. An etching rate suitable for the application at hand can be achieved by controlling the temperature and/or by selecting a different concentration.

Finally, any etchant used with the invention may be provided in any form suitable for the application, e.g as an aqueous or liquid solution, it may comprise additives forming a paste or gel, it can be atomized or vaporized, etc. Obviously, providing the etchant as a paste or gel can reduce or obviate a requirement for masking. It should also be understood that the chemical constituents of an etchant are not part of the present invention. Hence, any suitable etchant, including any suitable mixture of the compositions above with or without additives, may be used with the present invention.

Stopping the etching

Any suitable way of stopping the etching may be used with the invention. Such methods include, but are not limited to, neutralizing a base with an acid or vice versa, neutralizing the activity using de-ionized water, using a dry gas like air or nitrogen to blow away and/or evaporate liquid, etc.

Heating through hot spots

Figs. 3a and 3b illustrate a first embodiment of the invention, where the defect is etched in a selective manner. In this case, the selectivity is achieved by heating the defect using electric power. By heating a defect, a difference in temperature is created between the defect and the surrounding material, which in turn causes a difference in etch rate and hence selective etching of the defect.

In Fig. 3a a reverse bias is applied to a solar cell. The voltage over the cell is preferentially kept between 1 and 10V.

The selective etching can be achieved by selecting an etchant solution which etches the defect faster than the surrounding material. For example, HNO₃ is known to etch Ag considerably faster than Si, SiN_x, SiO_x and Al. Hence, HNO₃ can be used to etch front side fingers and bus bars without significantly affecting the anti reflective layer, the silicon substrate or Al-contacts at the back of the cell. Accordingly this method of eliminating a defect can be performed without identifying the location of the defect prior to etching. However, locating the defect prior to etching will reduce the required amount of etchant and reduce the area of contact between the etchant and the device.

Figure 3b shows an alternative embodiment in which the solar cell 100 is oriented horizontally with the side to be etched facing upwards. In Fig. 3b, a smaller amount of etchant 310 is dispensed on the uppermost side. The etchant 310 may have a sufficiently high viscosity to stay essentially where it is dispensed. The etchant may also have hydrostatic properties preventing it from spilling out into a larger patch. For example, a drop of an aqueous KOH-solution is known to stay in place on the front surface of a solar cell as manufactured. The etchant may optionally be confined by a mask or a mechanical device to avoid that it affects other parts of the cell 100.

In Figs. 3a and 3b, the solar cell is held by a holder 400 (not shown in Fig. 3a). The holder may clamp the edges of the cell for use in the embodiment of Fig. 3a.

In Fig. 3b, the cell is placed on or clamped to a holder 400 with the front side facing up. The holder 400 is a horizontal metallic plate, which act as a terminal contacting the back side contacts 1 1 1 of the cell 100. In this embodiment a relatively small amount of etchant 310 can be dispensed on the upper surface using a small pipe or pipette. Obviously, the back side contacts 11 1 could be etched in a similar manner by turning the cell such that the backside faces up. One or more other suitable etchants can optionally be applied to the front side of the cell to etch the antireflection layer or the silicon below.

Some types of shunts are only active under reverse bias, while others will also be active also under forward bias. By applying a reverse bias, both types of shunts will generate significant amount of heat compared to the rest of the cell. Hence, the chemical reaction between the etchant and the cell will accelerate at the position of the shunt. Different etchants are used for different purposes. For example:

• Isolation by using HNO3. HNO3 etches Ag, but does not significantly etch the surface coating/antiref lection layer 140, Al contacts 1 1 1 or Si substrate 130 and 120 as described above. Hence, the gridlines and/or bus bars in the vicinity of the shunt are etched away. In this way, the shunt is isolated from the rest of the cell. As long as there is no significant electrical contact with the defective region, the cell will behave like a fully functional cell. This etching can also be made without heat assisting, but it takes much longer time, and it is not selective.

• Repair by using NaOH. NaOH etches Si and Al. The reverse bias induces a chemical reaction that repairs the shunt itself. It is possible that the current through the shunt and/or the heating is sufficient to actually cure the shunt. It may not be possible to repair the shunt by simply applying the etchant and no other measures.

Thus, as shown in Figs. 3a and 3b, the cell 100 can be exposed to the etchant in different ways:

1. The entire cell, except possibly for the edge along which the cell is connected to the power supply, is lowered into a bath of etchant as shown in Fig. 3a. This may require more refined parameter settings than method 2, since the etchant might otherwise affect not only the shunt, but also other parts of the cell, e.g. the gridlines or backside Al contact. In this embodiment it would be possible to stop the etching by immersing the cell in another bath containing a neutralizing or buffering chemical.

2. A small volume of the etchant is dispensed on the surface of the cell, which is lying horizontally, at the position of the shunt. See example 5 below. This approach requires previous knowledge of the position of the shunt. Depending on the etchant and type of shunt, it might also require confinement of the shunted area not to let it spread out on the rest of the cell. See example 4 below. In this embodiment, it may be practical to stop the etching by neutralizing the chemical etchant, by removing the etchant e.g. by means of a flow of fluid, e.g. de-ionized water, dry nitrogen or air or some other suitable liquid or gas. Both the concentration of the etchant, voltage and etching time can be adjusted. It is possible to use different combinations of settings to repair/isolate a shunt successfully.

Example 1: Repair by selective etching of Si

Applying the inventive concept to a solar cell, electrical terminals are applied to both the front 1 10 and back 1 1 1 contacts or bus bars with the cell in a vertical orientation as shown in Fig. 3a, or a horizontal orientation as shown in figure 3b. The front- and back contacts 1 10 and 1 1 1 of the silicon solar cell is connected to a power supply. The shunt defect is then exposed to an appropriate etchant. In this example, we assume that an etching solution of NaOH 310 is dispensed on an area covering at least the shunt defect, either by immersing the cell in etchant as shown in Fig. 3a or dispensed in a local area as shown in Fig. 3b. A reverse bias is then applied to the electrical contacts on the front and back sides of the cell. This drives a current through the shunt defect, heating the defect. The shunt defect 200 will react vigorously with etchant 310. As the defect is etched, the electrical resistance increases, the current decreases and the reaction subsides. The voltage and current can be controlled to optimize the process time and extent of reaction.

Referring once more to Figure 3b, it is noted that the horizontal holder 400 may also be used as a heating plate heating the cell 100 to a predetermined temperature. Thus, the holder 400 in Fig. 3b can also be a heating plate, an electrical terminal, both of those or just a simple holder.

Example 2: Isolation etching

The cell in Figures 4a and 4b has a shunt (e.g. inclusions of SiN and/or SiC, a metal filled crack or some other defect) in the lower left part, as illustrated by the hatched area 201. The figures 4a and 4b are drawings of electroluminescence (EL) pictures. By running a reversed bias through the cell, the areas that would produce electron-hole pairs (i.e. contribute to the electricity gained from the cell) under illumination will instead emit photons, which can be detected by a camera in a dark room. In the EL pictures, areas that will contribute to the electricity production in a solar module will appear lighter and areas that will not contribute will appear darker. In the picture of the cell before etching, Figure 4a, a large dark patch 201 can be seen. The shunt defect itself is in the approximate middle of this. The reason that a larger part of the cell appears dark is that the current passes through the shortcut from not only the area of the shunt, but also from the area around it. Hence, no photons are produced in this part of the cell. The efficiency is low, about 8.5%, whereas well processed cells should be around 15%. An estimate of the shunt strength is the reverse current. This is the current measured at a certain reverse bias, in this case at -10V. The value before etching, 8.9 A, is high.

In this case, the etching was performed after confining the area with a material that will not let the etchant spread out. HN0₃ of 65wt% concentration was applied within the confinement. Firstly, a reverse bias of 3 V was applied to the cell. The reverse current, as read of the power supply utilized, was first 4.8 A, but decreased to less than 1 A in less than 30 s. After that, the reverse voltage was increased in steps to 10 V. The reason not to increase to 10 V immediately is that the shunt could have become very hot at lower voltages, which might have caused the etchant to bubble out over the cell and vaporize. To etch the silver gridlines within the confinement completely, a total of 3.5 minutes was used in this case.

It should be understood that other voltages, chemicals and concentrations can be used. However, the range 1-10V is believed to give the desired reaction rates while avoiding adversities such as boiling of the etchant or even breakdown of the cell due to severe hot spots in an unrepaired cell.

After the etching, the EL picture depicted in figure 4b still shows a dark, but smaller, area 202 in the lower left corner. It is in exactly the same position as where the gridlines are etched away. Now, the area 202 is not dark because the current flows through the shunt, but because no current enters this area. This is confirmed by the reverse current at -10 V, which has now decreased to 1.4 A. This value is generally considered to be within the limits of a functional cell. The efficiency has increased to 14.9%. Example 3: Heat induced by laser

With reference to example 2 above, it is noted that the area with inclusions is successfully isolated by etching of the fingers in the area off the defect. Etching only small parts of the fingers at the borders of the black rectangle 202 in the EL picture (Figure 4b, after etching) would give the same result, since it also prevents current from entering or leaving the shunted area.

Selective etching of the fingers may advantageously be carried out by dispensing small amounts of HNO3 on the fingers or gridlines at the borders of the black rectangle of figure 5b, and then heat the area using a laser. The HN0₃ can be provided in a solution of a proper concentration as described above. Because nitric acid does not wet a solar cell as described in Example 5 below, an aqueous or liquid solution can be used for this purpose. Alternatively, additives may be used to provide the etchant as a paste or gel. Obviously, a similar effect may in some cases be achieved by etching away Al-gridlines at the back using NaOH.

The laser light and etching chemical can be combined so that the laser beam propagates through a liquid jet. In this embodiment, the energy supplied by the laser is limited to avoid thermal stresses that may cause new, and potentially more severe, defects.

The result is a repaired solar cell, which started as a cell which does not meet industrial specifications for shunt resistance and reverse current, to a solar cell that does meet industrial standards for shunt resistance and reverse current.

Example 4. Two step etching

In this example, the shunt location on the cell is first identified and then the area to be etched can be sealed from other active areas of the cell by using a masking material and/or a masking technique using prior art techniques as described above. After masking, the cell can be placed on or clamped to a horizontal heating plate as described above, or heated using an IR source or any other heat source to carry out the two step etching process. In the first step silver is removed with nitric acid and in the second step the emitter is removed with potassium hydroxide. The etching processes are described in greater detail below.

A first step of isolating a shunt may comprise etching the current collecting fingers and/or bus bars represented by gridlines 1 10 in Fig. 1. Based on experimental results, a concentration of between 45-50 wt% HN0₃ and temperatures above 70 °C are used in this example for fast removal of silver from fingers and bus bars of standard crystalline silicon solar cells. The heating is in this example achieved by placing the cell on a hot plate heated to the required temperature, and the times required for etching 15-16 μη thick Ag bus bars are less than 4 s under these conditions. Lower temperatures and/or higher concentrations tend to increase the etching times due to availability of reactants and concentration of AgN0₃ in the solution.

After removal of Ag-gridlines, the silicon underneath the gridlines is exposed to the nitric acid solution. Nitric acid is known to oxidise silicon to silicon dioxide at a slow rate (0.1 nm/ minute in 40 wt% HNO₃ at 108 °C). For the reaction conditions in this example, formation of a very thin layer of silicon dioxide layer is expected.

The emitter of a typical solar cell with a sheet resistance of 45-55ohm/square is fairly conductive. Hence there will still be some leakage current flowing to the current collectors from the shunt through the emitter. Exhaustive experiments have shown that successful isolation of a shunt may be achieved by etching away conductive paths in all directions (360 degree in plane) around the shunted area.

However, the magnitude of leakage current through the cell can be further reduced by removing the emitter. Thus, a subsequent optional step involves etching the Si to a depth significantly below the emitter, e.g. to a few μητι below the front surface assuming that the emitter extends to a p-n junction about 0.3-0.4 μιτι below the front surface.

In this example, a KOH solution is used to carry out the process of etching the emitter layer. Experiments on optimization of KOH etch conditions carried out on (100) Si have shown that temperatures above 100 °C and a KOH solution of concentration between 30-40 wt % KOH can be sufficient for etching silicon. For multi-crystalline cells, experiments has shown that the etch rate varies significantly depending on the etch plane, but in general a solution with KOH concentration of 30-40 wt% and etch temperatures above 100 °C gives an etching rate of silicon of more than 2 μηι/ΐΎΐϊη. Etching rates tend to increase with temperature. Depending on the severity of the shunt, a longer etching time might be required.

As with the HNO₃- solution used for etching Ag-gridlines above, etching times can be controlled by varying the concentration in the range 10-90wt% KOH keeping the temperature fixed, and/or at concentrations in the range 30-40 wt % KOH and varying the temperature.

Example 5: Combination of liquid etchants and surface treatment

As indicated in the discussion of Fig. 3b above, a small amount of etchant 310 can be placed on a surface of the cell without confinement. This is possible e.g. when liquids that are placed on the surface are sufficiently hydrophobic so that they do not spread but only contact the surface where they are placed. In this manner defects in cells can be removed by etching with liquid etchants without using a mask or other confinement, and the etchant can be dispensed using a pipette or a dosimeter. Thus, after identifying the shunt, etchants and temperatures can be chosen as discussed previously. For example, 45-50wt% HNO3 at temperatures above 70 °C can be used to etch the front face Ag gridlines.

A nitric acid solution, e.g. the 45-50wt% HNO₃ solution, does not flow on the surface of a cell as manufactured. It will etch the Ag-contact just where it is placed. Hence, a liquid nitric acid solution is conveniently dispensed using a pipette or the like. Consequently there will be no masking to be removed, and still no large area from which spill and/or residue must be removed after etching.

After rinsing and possibly drying one or more relatively small spots where the grid- lines are etched away, a KOH-solution can be used to etch through the emitter as discussed above. A potassium hydroxide (KOH) solution does flow on the surface after the surface has been roughed or rubbed on a microscopic scale, e.g. by rubbing paper, on the area to be etched. In this example, a rubbing paper is used to rough the surface in a circular patch, and then apply the KOH-solution near the center of the patch using a pipette. The result after etching is illustrated in Fig. 1 b.

The etching rate for both etchants in this example is controlled by heating the cell.

Figure 5a illustrates yet an embodiment of the invention, where an inlet pipe or dispenser 301 is connected to an etchant supply system (not shown) from which the etchant flow into the inlet pipe or dispenser 301 as illustrated by arrow 31 1. The etchant flows out of the pipe 301 and is dispensed on the surface of the cell 100 where it reacts with the surface materials. Residues and products from the etching processes leave the system through an outlet pipe 352 as indicated by arrow 312.

An outer housing 350 is disposed around the dispenser 301. The housing 350 has a lower rim 351 that can engage the surface of the cell 100, thereby confining the etchant to the interior of the outer pipe. A pump (not shown) may be provided to help removing the residues, i.e. the products and/or material produced during etching. The rim 351 may have any shape, e.g. circular, rectangular, diamond shaped etc in order to etch an area of a desired shape.

Figure 5b is similar to Fig. 5a, the difference being a gap between the lower rim 351 and the surface of the cell 100. In Fig. 5b, a suction system reduces the pressure in the space between the dispenser 301 and the outer pipe 350 to below the ambient pressure. This causes a fluid to flow from the outside or exterior of housing 350, under the rim 351 and into the space inside the housing 350, as illustrated by arrows 360. The fluid flowing from the exterior to the interior of the housing 350 can be liquid, e.g. de-ionized water or a buffer solution, it may be air from the ambient atmosphere, or it may be any other liquid or gas. The influx of fluid helps confining the etchant, and may also stop the etching. In Fig. 5a and Fig 5b, the outer housing 350 and the outlet 352 form part of an etchant removal device. The removal device can be connected to the etchant supply system through a recycling device, thereby reducing the cost of the chemical etching process.

As the defect may be located under the antiref lection layer 140, a front contact 110, possibly also under a polymer, different etchants will be required to etch the different materials. In a preferred embodiment of the invention, the type and amount of etchant is controlled automatically.

When the defect or shunted region has a comparatively large area (more than, for example, 10 mm x 10 mm), the etching nozzle can be moved along the boundary to etch a groove as shown in Fig. 1 a to isolate the defective region from the rest of the cell 100.

The etchant(s) dispensed by the inlet pipe 301 can be in a state of sol-gel, gas, liquid, nebulized or atomized state. The chemical can be in a different state when leaving the surface than the state it had when applied to the surface, as illustrated by the hatching on arrow 312 in Fig. 5b. Its state can be changed on the etched surface by supersonic technology or other technologies.

The active chemical can be dissolved in any suitable solvent. The temperature of the chemical can be controlled before entering the inlet pipe 301 or within the housing 350. In addition to etching, a cleaning process can be performed by dispensing an appropriate chemical, such as de-ionized water, ethanol or other inorganic or organic cleaning chemicals through the dispenser pipe 301.

It should be understood that any or all of the techniques and apparatus disclosed herein may be combined in order to isolate or remove a defect from the equivalent circuit of a solar cell, thus bringing the cell within predetermined levels for efficiency and/or electrical properties.

Claims

1. A method for removing a defect (200) from a solar cell (100) comprising a front side with first electrical contacts (110), a conversion layer converting photon energy to electric energy, and a back side with second electrical contacts (111 ), the method being characterized by

using a chemical etchant to etch a region of the cell until the defect (200) has been isolated from an electric circuit comprising the first (110) and second (1 1 1) electrical contacts and electrical components representing the conversion layer.

2. The method according to claim 1 , characterized in that etching the region involves etching through at least one part of at least one of the first (110) and second (1 11 ) electrical contacts.

3. The method according to claim 1 or 2, characterized in that etching the region involves etching a groove from the front side, through an emitter layer (130) having an excess of loosely bound electrons in the lattice and into a collector layer (120) having an excess of vacancies in the lattice, the groove forming a path around the defect (200).

4. The method according to any preceding claim, characterized in that etching the region involves removing the defect (200).

5. The method according to any preceding claim, characterized by selecting at least one etchant from a group comprising HN0₃, HF, H₃P0_4, NaOH, and KOH.

6. The method according to any preceding claim, characterized by applying the etchant in a first aggregate state having a first viscosity.

7. The method according to claim 6, characterized by changing at least one of the etchant's aggregate state and viscosity before removing it from the etched surface.

8. The method according to any preceding claim, characterized by controlling the temperature of the etchant and/or at least a part of the cell.

9. The method according to claim 8, characterized by keeping the

temperature of all parts of the cell below 350 °C.

10. The method according to any preceding claim, characterized by imposing a voltage between the first (110) and second (111) electrical contacts during etching, whereby the cell is heated locally at and near the defect (200).

11. The method according to claim 10, characterized by increasing the voltage in steps from 1 to 10 V at reverse bias.

12. The method according to claim 8 or 9, characterized by heating a surface area of the cell using a laser beam.

13. The method according to claim 12, characterized by sending the laser beam within a jet of liquid etchant.

14. The method according to any preceding claim, characterized by the further steps of:

- covering the solar cell with a masking material inert to the etchant, and

- providing an opening in the masking material over the region to be etched.

15. The method according to any one of the claims 1 -13, characterized by the further steps of:

- selecting a liquid etchant which will remain in place during etching, and

- applying the liquid etchant using a pipette or other tubing.

16. The method according to claim 15, characterized by the further step of rubbing a surface area before applying the etchant to the rubbed area.

17. An apparatus for removing a defect (200) from a solar cell (100) comprising a front side with first electrical contacts (110), a conversion layer converting photon energy to electric energy, and a back side with second electrical contacts (111), the apparatus being characterized by

- an etchant dispenser (300, 301),

- a solar cell holder (400),

- a monitoring device, and

- an etchant removal device (350, 352).

18. The apparatus according to claim 17, characterized in that

it further comprises a heat source (400, 401).

19. The apparatus according to claim 18, characterized in that

the heat source (401 ) is a source for voltage and electric current having first and second terminals adapted to be connected to the first (110) and second (111 ) electrical contacts on the cell (100).

20. The apparatus according to claim 18, characterized in that

the heat source is a heating plate.

21. The apparatus according to claim 18, characterized in that

the heat source is a laser.

22. The apparatus according to any one of the claims 17 to 21 , characterized in that the etchant dispenser is a vessel (300) sufficiently large for the cell (100) to be at least partially immersed in etchant (310) therein.

23. The apparatus according to any one of the claims 17 to 22, characterized in that the solar cell holder (400) keeps the cell (100) in a horizontal orientation.

24. The apparatus according to any one of the claims 18 to 23, characterized in that it further comprises control means for adjusting the temperature of the defective region to a temperature in the range 10 °C to 350 °C.

25. The apparatus according to any one of the claims 17 to 24 characterized in that the etchant dispenser is an inner pipe (301 ) disposed within an outer housing (350), the outer housing (350) being connected to an outlet pipe (352), the outer housing (350) and the outlet pipe (352) forming part of the etchant removal device.

26. The apparatus according to claim 25 characterized in that a pump (not shown) is connected to the outlet pipe (352), its low pressure side in fluid connection with the interior of the housing (350).

27. The apparatus according to claim 25, characterized by a recycling device coupled between the etchant removal device and the etchant dispenser.

28. The apparatus according to any one of the claims 17 to 27, characterized in that the monitoring device measures electrical parameters of the cell (100).

29. The apparatus according to any one of the claims 17 to 28, characterized in that the monitoring device measures at least one of etchant concentration, temperature and etching time.