„Interdigitated back contact" solar cells , known and further on also referred to as IBC solar cells , have the advantage that the solar cell is not shaded on the front side by metal contacts . The n-and p-type doped regions and their respective metal contacts required to separate and utili ze the hole and electron pairs generated by the solar irradiation are placed at the backside of the solar cell in an interdigitated fashion leaving the front side completely free from metal contacts . This enables particularly high conversion rates ( ef ficiencies ) of light into electricity to be achieved .

A common way of producing the n- and p-type regions in solar cells are homogeneous di f fusion processes from gas or liquid phase or solid source creating highly doped regions inside the semiconductor . Alternatively, it has proven to be advantageous to deposit a stack of thin insulation layers with a doped polycrystalline silicon layer on top instead of in-di f fusing dopants into the semiconductor obtaining so-called "passivating contacts" or "carrier selective passivating contacts" .

According to the so-called TOPCon principle (wherein TOPCon is used as an abbreviation of the term „tunnel-oxide passivated contact" ) , a thin silicon oxide layer is first produced, on which a layer of highly doped polycrystalline silicon is formed, usually with a dielectric hydrogen-providing amorphous passivation layer on top, onto which the metalli zation ( e . g . by screen printing) is applied according to common methods . This drastically reduces the charge carrier recombination under the metal contacts and signi ficantly increases the solar cell ef ficiency . IBC solar cells in accordance with this principle and various methods for their production are well known, e . g . from US 8 790 957 B2 , US 9 196 758 B2 , US 8 603 851 B2 , EP 2 980 858 A2 , and US 10 658 529 B2 .

Apart from solar ef ficiency, the behaviour of a solar cell under reverse current conditions is crucial for the solar module performance . I f single solar cells within a string of solar cells are shaded, the current of the whole string may set the shaded cells into breakdown condition ( reverse bias ) and, i f the current flows through single weak spots in the diode , hot spot formation may occur . To avoid this , commonly, bypass diodes are included in the solar module . In case of shading of single cells the whole string is bypassed and is lost for electrical energy generation . IBC solar cells , in particular, when only di f fused regions are applied on the rear side , may have so-called " soft break-down behaviour" . The breakdown voltage is relatively low and the current flows through the whole area of the cell and thus hot spot formation is avoided . The lower the break down voltage , the lower the dissipated power in form of heat , so that the bypass diode is not required in the circuit of the solar module .

In passivating contact IBC solar cells this bypass-diode-like behaviour has been shown to be implemented in various , more or less complicated ways . As it is known that p+ and n+ polysilicon layers may not touch to avoid recombination currents , however, touching regions of the two polarities can also work as bypass diode . This behaviour is made use of by locally shortcircuiting the two polarities in certain regions distributed over the wafer surface , as described in US 7 851 698 B2 , thus compromising solar cell ef ficiency against reverse bias behaviour . Other, more complicated approaches e . g . as demonstrated by the same applicants in US 8 134 217 B2 or US 10 217 880 B2 , indicate that good reverse bias behaviour without compromising solar cell ef ficiency to date requires complicated structures obtained by very complicated processing sequences .

The complicated process sequences that often involve nonstandard production equipment make the IBC solar cells premium market products .

Accordingly, the problem to be solved by this invention is providing a more simple solar cell structure for an IBC solar cell with good reverse bias behavior ( low breakdown voltage ) that can be fabricated more easily and a method for producing such a solar cell .

This problem is solved by an interdigitated back contact solar cell with the features of claim 1 and a method for producing such a solar cell with the features of claim 6 . The respective dependent claims are directed at advantageous embodiments of the invention .

An interdigitated back contact solar cell according to the invention comprises a monocrystalline , n-doped wafer, a first contact area which is formed by a first stack on the surface of said monocrystalline wafer, said first stack comprising a thin silicon oxide layer and a highly n-doped polycrystalline silicon layer, and a second contact area which is formed by a second stack on the same surface of said monocrystalline wafer as said first stack, said second stack comprising a thin silicon oxide layer and a highly p-doped polycrystalline silicon layer . Furthermore , according to the invention there is an additional p-doped monocrystalline silicon region located in a gap between said first contact area and said second contact area .

The thin silicon oxide layer represents a tunneling oxide layer, thus "thin" means typical thicknesses of less than 5nm . The term "highly doped" is used for layers doped with at least 7E18 atoms/cm ³ .

The presence of a gap, i . e . a distance in a direction parallel to the surface of the monocrystalline , n-doped wafer, between each part of the first and second contact area, respectively, implies and makes sure that direct contact between n+ doped polysilicon and p+ doped polysilicon is completely avoided . The gap thus prevents the recombination of minority charge carriers where highly n-doped polysilicon ( also referred to as n+ polysilicon) touches highly p-doped polysilicon ( also referred to as p+ polysilicon) because no such contact is present due to the presence of the gap .

Instead, it induces a charge carrier di f fusion barrier, preventing parasitic shunting (which would increase the recombination current ) but allowing for favorable low voltage breakthrough . This is the case , because by doping the gap with emitter type , i . e . p-type dopant , the bypass-diode function is achieved as this doping produces the highly n-doped polysilicon - highly p-di f fused doping profile in bulk silicon - highly p-doped polysilicon structure . Each of the doped polysilicon regions is contacted by a metalli zation finger and separated by a p-di f fused doped region in the bulk silicon .

Furthermore , the di f fusion of emitter type dopant into the gap regions decreases the proportion of electrically shaded, nondoped areas . Also , the use of the excellent passivating properties of a boron doped high sheet resistance emitter profile created in the gap region eases the needs for highest precision of alignment between the respective structuring processes needed during fabrication of the solar cell .

The second contacting region width may be of smaller width compared to the width of the initially etched trench so that even with a slightly imprecise positioning of the second contacting region, the two contacting regions will not touch . The gap region width in turn can be chosen to be wider than in other state of the art IBC solar cells .

The width of the gap and/or the doped region located therein can preferably be equal or even wider than hal f of the widths of at least one of the first and second area regions or even wider than both of these widths . A motivation for such wide gap and/or doped region is that the incident light collection ef ficiency of the rear side of the solar cell is improved, as it widens the regions without polysilicon layers which show strong parasitic absorption of incident light . Thus very high bi faciality factors ( rear side light collection ef ficiency divided by front side collection ef ficiency) can be achieved . Preferably, the entire width of the gap region between the first and second contact area, respectively, is doped in this way, but direct contact between this doped region and the n- doped polysilicon region is not required .

It should be noted that typically, the stacks that form the respective first and second contact area comprise further layers , speci fically dielectric layers like SiN _x or A10 _x /SiN _x or AlO _x /SiO _x N _y and metal contacts . Still , the term contact area is used in this description for a respective region of the wafer as soon as the thin silicon oxide layer and the highly doped polysilicon layer are present .

Whereas in principle this gap may be a filled gap or a partly filled gap in which other material is arranged that does not belong either of the first contact area and the second contact area, in a preferred embodiment of the interdigitated back contact solar cell , the gap is formed by a trench, which means that the level of the surface of the interdigitated solar cell is lower in the region of the gap than in other regions of the solar cell , especially the first and second contact area .

This arrangement is favorable from a processing point of view, as it allows for creation of the gap by removal of layers that are applied to the surface of the wafer to create the first and/or second contact area . In this embodiment , the additional p-doped monocrystalline silicon region in the gap is formed in parts of the monocrystalline silicon wafer that form the side and bottom walls of the trench, preferably in the entire side and bottom walls of the trench .

In a preferred variant of this embodiment , the height of the opposite side walls of the trench is di f ferent . This feature implies that the first contact area and the second contact area, which form these walls , are located at di f ferent height . This alternating height of the position of the p+ and n+ layers , which can be adj usted according to the choice of process parameters , allows for separate application of di f ferent screen printing layers for each polarity individually .

In another preferred embodiment , the interdigitated back contact solar cell further comprises a p-doped floating emitter on the side of the wafer that is opposite to the side on which the first and second contact areas are located . This floating emitter can be formed simultaneously with the doping of the gap region .

The method for producing such an interdigitated back contact solar cell comprises the steps of

- creating a first contact area which is formed by a first stack on the surface of a monocrystalline wafer, said first stack comprising a thin silicon oxide layer and a highly n- doped polycrystalline silicon layer,

- creating a second contact area which is formed by a second stack on the same surface of said monocrystalline wafer as said first stack, said second stack comprising a thin silicon oxide layer and a highly p-doped polycrystalline silicon layer,

- creating a gap between the said first contact area and said second contact area, and

-creating a p-doped monocrystalline silicon region located in a gap between said first contact area and said second contact area .

It should be noted that the sequence of execution of these processing steps is not necessarily the order in which they are listed . Speci fically, whereas the gap may in principle be created by subsequent removal of an area in which a previously formed first contact area and a previously formed second contact area touch each other, according to a preferred embodiment of the invention after the first contact area has been created, alkaline etching is performed in an area of the surface of the monocrystalline wafer outside of the first contact area but comprising the area in which the second contact area is to be created, so that the thickness of the monocrystalline silicon wafer is reduced in the area where the alkaline etching is performed .

The gap is then formed by a subsequent creation of the second contact area in the region in which alkaline etching has been performed, but at a distance from the first contact area . This can be done e . g . by creation of a local thin silicon oxide tunneling layer, creation of a local poly-silicon layer on top of this tunneling layer and locali zed doping, as described in the European patent application with application number 20201147 . 4 , the content of which is incorporated by reference , which locali zed doping is preferably performed as a locali zed laser induced doping . The addition of these layers provides the gap which is formed as a trench whose sidewalls that are formed by the first and second contact have a di f ferent height due to the local reduction of the thickness of the monocrystalline silicon wafer .

Speci fically, in some embodiments of the invention, the second contact area is created by modi fying the etching resistance of an intrinsic polycrystalline layer . This can be achieved e . g . by locally di f fusing p-dopant into the layer only in a region where the second contact area is to be located to create a local highly doped region, which is not etched in a subsequent alkaline etching step creating a gap region .

Likewise , the step of creating a p-doped monocrystalline silicon region located in a gap between said first contact area and said second contact area can be performed at di f ferent stages of the method . E . g . , according to one preferred embodiment , when creating the p-doped monocrystalline silicon region located in the gap between said first contact area and said second contact area a front floating emitter structure is created simultaneously, which reduces the number of processing steps and allows for excellent front surface passivation . The front floating emitter structure also eases the need for high purity material , as the lateral transport of minority charge carriers is improved .

Likewise , in another preferred embodiment when creating the p- doped monocrystalline silicon region located in the gap between said first contact areas and said second contact area, the p-doping of the highly p-doped polycrystalline silicon layer of the second contact area is also modi fied . This can be done , e . g . , by applying a tube furnace di f fusion to the gap and the non-protected highly p-doped polycrystalline silicon region and allows for the fine-tuning of the polysilicon region dopant profile .

A speci fic embodiment of the invention is now discussed based on figures . The figures show :

Fig . 1 : A possible solar cell processing sequence according to one embodiment of the invention,

Fig . 2 : three bypass diode functional structures of IBC solar cells according to embodiments of the invention, and

Fig . 3 : reverse bias curves of a solar cell according to one embodiment of the invention and reverse bias curves of a homogeneous di f fused IBC solar cell ( ZEBRA™) and of a solar cell without a doped p+ region in a gap between the two polarity polysilicon regions . It should be noted that in the Figures schematic cross sections , which are not to scale , of the solar cells or solar cell precursors , respectively, are displayed with their front side , i . e . the side that is to be facing the sun, pointing down and their back side , i . e . the side that is opposite to the side that is to be facing the sun, is pointing up .

Likewise , it should be noted that in Figure 1 and 2 only a part of a complete wafer 101 is shown, but only a "unit cell" comprising one second contact area as well of hal f of each of the two neighboring first contact areas arranged next to it and the gaps separating the first contact areas from the second contact areas . Placing a large number of these unit cells next to each other leads to a corresponding schematic cross section, which is not to scale , through a complete solar cell wafer .

Figure 1 schematically describes various intermediate stages 1- 9 ( i . e . solar cell precursors ) and the final interdigitated back contact solar cell 10 obtained by one possible embodiment of a method according to the invention .

In this example , the processing of the solar cell according to this invention starts with providing a pre-processed solar wafer 101 , typically an n-type mono-crystalline substrate , after saw damage removal and cleaning . Intermediate stage 1 shows this wafer 101 on which a first layer stack, a part of which will be used to form the first contact area, has been deposited on at least one side .

The layer stack consists in this example of a thin thermal silicon oxide layer 201 , an n-type amorphous or polycrystalline silicon layer 301 and an etching and di f fusion mask 401 , e . g . SiO _x or SiN _x deposited by PECVD . Depending on the processing sequence that is applied, at this stage the n-type doping does not yet have to be completed; it is also possible to provide the Si layer and a dopant source layer and drive in the dopant atoms from the dopants source layer at a later point of time .

In the intermediate stage 2 , the mask is locally removed . This can be done , as illustrated, by laser ablation, but also in di f ferent ways , e . g . by applying a suitable etching paste .

Forming a first trench 111 by alkaline etching leads to intermediate stage 3 . In this stage , the first contact areas 10 are already pre- formed .

The intermediate stage 4 is then reached by deposition of a second stack of a thin interface silicon oxide 202 , and an intrinsic amorphous or polycrystalline silicon layer 501 with a p+ dopant layer on top 601 on at least the same side of the wafer 101 .

This stack which is subsequently treated, as illustrated in intermediate stage 5 by a laser, the width of which is adapted so that the entire laser beam is within the first trench 111 , but with a width that is smaller than the width of said trench 111 , because the first contact areas 40 and the second contact areas are arranged at di f ferent heights .

By doing so , as seen in intermediate stage 6 , a highly p- doped polysilicon layer 701 is formed on top of the thin silicon oxide layer 202 , thus forming the second contact areas 20 and defining the gaps 30 between first contact areas 40 and second contact areas 20 . Well-known alternative methods for locali zed doping application may also be considered such as masked ion implantation or local screen printing of solid doping layers

The highly p-doped polysilicon layer 701 is more etch resistant against alkaline solutions compared to the undoped polysilicon layer 501 . Therefore , an alkaline solution is then used to remove the non-treated areas including the surface of the first contact area, which leads to the formation of trenches 112 in the gap region 30 . The opposing walls of these trenches 112 have a di f ferent height . The alkaline solution may also be used to create a textured surface in a second trench 112 and also of the front side of the wafer (not shown) .

Next , intermediate stage 8 is created by a di f fusion of the same dopant type as used for the polysilicon layer 701 of the second contact areas 20 is employed ( e . g . a tube furnace di ffusion of BBr ₃ , BCI3 or another type of dopant source for the same type of doping) , thus creating a highly p-doped monocrystalline silicon region 801 located in the gaps 30 between said first contact areas 40 and said second contact areas 20 by a homogenous in-di f fusion of emitter type dopant into the bulk of wafer 101 that forms the bottom of the trench . This also creates the bypass active area .

At the same time this dopant may also be used to dope the front side of the sample creating a so called front floating emitter structure , which is not shown in intermediate stage 8 .

Then at least the back side , but typically both sides of the wafer are coated with dielectric layers e . g . SiN _x or A10 _x /SiN _x 402 , leading to intermediate stage 9 . Alternatively, a stack of SiO _x in-situ (in the diffusion furnace grown oxide) or separately grown, covered by SiN _x may be used to passivate the surfaces .

The final interdigitated back side contact solar cell 10 also comprises metal contacts 911,912 on the respective first and second contact areas 40,20. These are typically created by screen printing a metal paste onto these regions and firing. In this context, it is noted that due to the different height of first contact area 40 and second contact area 20, respectively, separate screen printing processes can be used selectively for the respective first and second contact areas 40,20.

In one embodiment of the invention, which is not shown in Figure 1, the etching and diffusion mask 401 may be removed at one stage in the process flow, preferably after intermediate stage 7 has been reached.

Small alteration in the processing sequence can lead to different embodiments of the invention, such as the interdigitated back contact solar cells a,b and c that are shown in Figure 2, using the same reference numerals of Figure 1 for the various constituent parts of the IBC solar cells.

Specifically, in Figure 2 the bypass diode functional structure 1001,1002 is indicated by a square with broken lines. Depending on choice of etching medium, the doped surface may textured as in diode functional structure 1001, or flat etched as in diode functional structure 1002 according to known alkaline etching steps. In the embodiments of the invention shown as b and c, instead of a front floating emitter 801, a front surface field diffusion (of the same type as the base) 802 may be employed. To achieve these three variations, additional masking layers or single side treatments may be necessary which are well-known to people skilled in the art.

Also, in Figure 2 the respective first contact areas 40, second contact areas 20, and gaps 30 are indicated.

Figure 3 indicates the superior reverse bias behaviour with a very early breakdown voltage without affecting the shunt resistance (and light IV parameters) of the solar cell according to an embodiment of the invention in comparison of state-of the art homogeneous diffusion based IBC solar cells, specifically a ZEBRA™ solar cell, and a structure with a trenchshaped, but undoped gap between the first and the second contact area.