TW201703855A - Printable pasty diffusion and alloying barrier for the production of highly efficient crystalline silicon solar cells - Google Patents

Abstract

The present invention relates to a printable hybrid gel which serves for the production of electronic passivation layers against aluminium. The invention furthermore encompasses the preparation and use of the paste according to the invention.

Description

Printable paste for diffusion and alloying barriers in the manufacture of highly efficient crystalline solar cells

This invention relates to a printable hybrid gel for use in the manufacture of an aluminum resistant electronic passivation layer. The invention further comprises a process and use of the paste of the invention.

The manufacture of a simple solar cell or a solar cell that currently represents the largest market share in the market includes the basic manufacturing steps outlined below:

1) Cutting damage etching and texture

Tantalum wafers (single crystal, polycrystalline or quasi-single crystal, substrate doped p-type or n-type) are "simultaneous" textured by etching methods and generally in the same etching bath without adhesive cut damage. Texturing in this case means establishing a prioritized surface property as a result of the etching step or only a deliberate, but non-specific, roughening of the wafer surface. As a result of texturing, the wafer surface now acts as a diffuse reflector and thus reduces directional reflection, which depends on the wavelength and angle of the incident light, ultimately increasing the proportion of light incident on the surface that is absorbed and thus the conversion efficiency of the solar cell improve.

The above etching solution for treating a germanium wafer is usually composed of a dilute potassium hydroxide solution in which isopropanol has been added as a solvent in the case of a single crystal wafer. If this can achieve the desired etching results, it can be changed to add other vapor pressures or higher than isopropanol. The boiling point of the alcohol. The desired etch results obtained are typically characterized by the morphology of a square-bottomed cone that is randomly or specifically etched from the original surface. The density, height, and thus the bottom area of the cones can be partially affected by suitably selecting the composition of the etching solution, the etching temperature, and the residence time of the wafer in the etching bath. The texturing of single crystal wafers is typically carried out at temperatures ranging from 70 to < 90 ° C with up to 10 μm of material on each side of the wafer being removed by etching.

In the case of a polycrystalline silicon wafer, the etching solution may consist of potassium hydroxide having a moderate concentration (10 to 15%). However, this etching technique is still rarely used in industrial practice. More frequently, an etching solution composed of nitric acid, hydrofluoric acid, and water is used. The etching solution can be modified by various additives (such as, for example, sulfuric acid, phosphoric acid, acetic acid, N-methylpyrrolidone) and surfactants, which can specifically affect the wetting properties of the etching solution and its etching rate. . These acidic etching mixtures produce a pattern of nested etched trenches on the surface. The etching is usually carried out at a temperature between 4 ° C and < 10 ° C, and the amount of material removed by etching herein is usually 4 μm to 6 μm.

Shortly after texturing, the germanium wafer is sufficiently cleaned with water and treated with dilute hydrofluoric acid to remove the chemical oxide layer formed by the aforementioned processing steps and the contaminants absorbed and adsorbed therein as well as subsequent high temperature treatment. Prepare.

2) Diffusion and doping

The etched and cleaned wafers in the foregoing steps (in the case of this p-type substrate doping) are treated with a vapor consisting of phosphorus oxide at elevated temperatures (typically at temperatures between 750 ° C and < 1000 ° C). During this operation, the wafers were exposed to a controlled atmosphere of dry nitrogen, dry oxygen, and phosphonium chloride in a tubular furnace in a quartz tube. To this end, the wafers are introduced into a quartz tube at a temperature between 600 and 700 °C. The gas mixture is conveyed through the quartz tube. During transport of the gas mixture through the strongly heated tube, the phosphonium chloride decomposes to produce a vapor consisting of phosphorus oxide (e.g., P ₂ O ₅ ) and chlorine. The phosphorus oxide vapor precipitates, in particular, on the wafer surface (coating). At the same time, the surface of the crucible oxidizes at these temperatures and forms a thin oxide layer. The precipitated phosphorous oxide is embedded in the layer, causing a mixed oxide of cerium oxide and phosphorus oxide to form on the surface of the wafer. This mixed oxide is called phosphosilicate glass (PSG). This PSG has a softening point and a diffusion constant different from that of phosphorus oxide, depending on the concentration of phosphorus oxide present. The mixed oxide acts as a diffusion source for the germanium wafer, wherein the phosphorous oxide diffuses along the interface between the PSG and the germanium wafer during the diffusion process, in which case it is reduced to phosphorus by reaction with germanium at the surface of the wafer. (矽热法). The phosphorus formed in this way has a higher order of solubility in the crucible than the glass substrate on which it has been formed and is therefore preferentially dissolved in the crucible due to the extremely high segregation coefficient. After dissolution, phosphorus diffuses into the enthalpy and enters the enthalpy volume with concentration gradients. In this diffusion process, a concentration gradient of about 10 ⁵ is formed between a typical surface concentration of 10 ²¹ atoms/cm ² and a substrate doping in a region of 10 ¹⁶ atoms/cm ² . Typical diffusion depths range from 250 to 500 nm and depend on the selected diffusion temperature (eg, 880 ° C) and the total exposure duration of the wafer under intense heating (heating and coating stages and drive-in stages and cooling). During the coating phase, a PSG layer is formed which typically has a layer thickness of 40 to 60 nm. The wafer is coated with PSG, during which time it has also spread within the volume of the crucible, followed by the drive-in phase. This can be decoupled from the coating stage, but in practice it is generally coupled directly to the coating in terms of time and therefore typically also at the same temperature. The composition of the gas mixture herein is adjusted in such a manner as to further provide that the phosphorus oxychloride is inhibited. During the driving in, the surface of the crucible is further oxidized by the oxygen present in the gas mixture, causing the cerium oxide layer which also contains the depleted phosphorous oxide of phosphorus oxide to be produced from the actual doping source, the phosphorous-rich PSG and the twin crystal. Round room. The growth of this layer becomes very fast in terms of the mass flow rate of the dopant from the source (PSG) because the oxide growth is accelerated by the high surface doping of the wafer itself (accelerated by one to two orders of magnitude). This may result in depletion or separation of the dopant source in a particular manner, as the penetration of the phosphorus oxide diffusion is affected by the material flow, which material depends on the temperature and thus the diffusion coefficient. In this way, the doping of germanium can be controlled within a specific range. A typical diffusion duration consisting of a coating phase and a drive-in phase is, for example, 25 minutes. After this treatment, the tubular furnace is automatically cooled, and the wafers can be removed at a temperature between 600 ° C and 700 ° C.

In the case of boron doping of a wafer in an n-type substrate doped form, the method is not separately described herein using a different method. The doping in such cases is carried out, for example, with boron trichloride or boron tribromide. Depending on the choice of composition of the gas atmosphere used for doping, the formation of so-called boron skin on the wafer can be observed. This boron skin depends on various influencing factors: the doping atmosphere, temperature, doping duration, source concentration, and the above coupling (or linear combination) parameters.

In such a diffusion process, it goes without saying that if the substrate has not been previously pretreated accordingly (for example, it is diffusion-suppressed and/or the diffusion-suppressing layer and material is structured), the wafer used does not contain any better diffusion. And doped regions (except for those formed by a non-uniform gas stream and the resulting non-uniform composition of the gas pockets).

In terms of completeness, it should also be noted herein that there are other diffusion and doping techniques established to varying degrees in the fabrication of germanium-based crystalline solar cells. Thus, mention may be made of: erbium ion implantation, vapor deposition via mixed oxides, such as deposition of, for example, PSG and BSG (boron silicate glass), doping by means of APCVD, PECVD, MOCVD and LPCVD processes , (co)spraying of cerium mixed oxides and/or ceramic materials and hard materials (such as boron nitride), pure thermal vapor deposition starting from solid dopant sources (such as boron oxide and boron nitride), ̇ Boron is sputtered onto the surface of the crucible and thermally driven into the germanium crystal, from a dielectric passivation layer having a different composition (such as, for example, Al ₂ O ₃ , SiO _x N _y , wherein the latter contains a mixed P ₂ Dopants in the form of O ₅ and B ₂ O ₃ ) are laser doped, liquid and liquid (ink) and liquid phase deposition of paste with doping behavior.

The latter is frequently used in so-called in-line doping, in which the corresponding paste and ink are applied to the side of the wafer to be doped by means of a suitable method. After the application or even During the application, the solvent present in the composition for doping is removed by temperature and/or vacuum treatment. This leaves the actual dopant behind the wafer surface. Liquid doping sources such as dilute solutions of phosphoric acid or boric acid may also be employed as well as sol-gel based systems or solutions of also boron borazil compounds. The corresponding doped paste is substantially completely characterized by the use of additional thickening polymers and comprises dopants in a suitable form. The high temperature treatment is usually carried out after evaporating the solvent from the above doping medium, during which time undesired and interfering additives other than the additives necessary for the formulation are "burned out" and/or pyrolyzed. In either case. Solvent removal and burnout (but not required) can be performed simultaneously. The coated substrate is then passed through a direct current furnace at a temperature between 800 ° C and 1000 ° C, wherein the temperatures can be increased slightly compared to the gas phase diffusion in the tubular furnace to reduce the passage time. The main gas atmosphere in the DC furnace may vary depending on the doping requirements and may consist of dry nitrogen, dry air, a mixture of dry oxygen and dry nitrogen, and/or a furnace to be passed through, one of the above gas atmospheres or other Depending on the design of the area. Other gas mixtures are conceivable, but are currently not extremely important in the industry. The characteristic of on-line diffusion is that the coating and driving of dopants can in principle be decoupled from each other.

3) Removal of dopant source and optional edge insulation

A small amount of glass is coated on both sides of the surface on both sides of the wafer after doping. Some in this case refer to the modification that can be applied during the doping process: bilateral diffusion relative to the quasi-unilateral diffusion promoted by placing two wafers back-to-back in one position of the process vessel used. The latter variant primarily achieves one-sided doping, but does not completely inhibit diffusion on the back side. In both cases, the current best technique is to remove the glass present after doping by etching in a dilute hydrofluoric acid from the surface. For this purpose, the wafers are reloaded in batches on a wet process vessel and by means of the latter immersed in a dilute hydrofluoric acid solution (usually 2% to 5%) and left in it until the surface is completely free of glass or The process cycle duration indicating the required etch duration and the summation parameters of the process automation by the machine has expired. Complete removal of the glass can be achieved, for example, by completely dehumidifying the wafer surface with a dilute hydrofluoric acid aqueous solution. Complete removal of PSG at room temperature This is achieved within 210 seconds under such process conditions (eg, using a 2% hydrofluoric acid solution). Corresponding BSG etching is slower and requires a longer process time and the concentration of hydrofluoric acid used may also be higher. After etching, the wafers are rinsed with water.

On the other hand, etching of the glass on the surface of the wafer can also be performed in a horizontal operation process in which the wafers are introduced into the etcher at a constant flow rate, wherein the wafers pass through the corresponding process grooves (online machines). In this case, the wafer is transported onto the roll by either the process slot and the etching solution present therein or the etched medium is applied to the wafer surface by means of a roll. The typical residence time of the wafers during the etching of the PSG is about 90 seconds, and the hydrofluoric acid used has a slightly higher concentration than in the case of a batch process to compensate for the reduced residence time due to the increased etch rate. The concentration of hydrofluoric acid is usually 5%. The bath temperature may additionally increase slightly compared to room temperature (>25 ° C < 50 ° C).

In the final overview of the process, it has been achieved to perform so-called edge insulation simultaneously in sequence, resulting in a slightly modified process flow: edge insulation → glass etching. Edge insulation is a technical necessity caused by the inherent nature of the system in which the two sides diffuse (also in the case of deliberate one-sided back-to-back diffusion) in the process. A large area of parasitic p-n junctions are present on the back of the solar cell (back), which is partially, but not exclusively, removed during subsequent processing for process engineering reasons. As a result, the front and back of the solar cell will have been shorted by parasitic and residual p-n junctions (tunneling contacts), which reduces the conversion efficiency of subsequent solar cells. For the removal of the junction, the wafers pass through an etching solution consisting of nitric acid and hydrofluoric acid on one side. The etching solution may contain, for example, sulfuric acid or phosphoric acid as the second component. Alternatively, the etching solution is transported (transferred) onto the back of the wafer via a roller. About 1 μm (including the glass layer present on the surface to be treated) is usually removed by etching in this process at a temperature between 4 ° C and 8 ° C. In this process, the glass layer still present on the opposite side of the wafer acts as a mask, which can be specifically protected against over-etching onto the side. This glass layer is then removed by means of the glass etching already described.

In addition, the edge insulation can also be performed by means of a plasma etching process. This plasma is etched This is usually done before the glass is etched. To this end, a plurality of wafers are stacked one wafer on top of the other wafer, and the outer edges are exposed to the plasma. The plasma is fed with a fluorinated gas such as tetrafluoromethane. The reactive material generated during the decomposition of the plasma of the gas etches the edge of the wafer. Generally, after the plasma is etched, then glass etching is followed.

4) Coating the front surface with an anti-reflection layer

After etching the glass and optional edge insulation, the surface of the subsequent solar cell is then coated with an anti-reflective coating, typically consisting of amorphous and hydrogen-rich tantalum nitride. Alternative anti-reflective coatings are contemplated. A possible coating may consist of titanium dioxide, magnesium fluoride, tin dioxide and/or a corresponding stack of cerium oxide and tantalum nitride. However, antireflective coatings having different compositions are also technically feasible. Coating the wafer surface with the above-described tantalum nitride substantially achieves two functions: on the one hand, the layer generates an electric field due to a plurality of incorporated positive charges, which maintains the charge carriers in the crucible away from the surface and can be greatly reduced The rate of recombination of these charge carriers at the surface of the crucible (field effect passivation), on the other hand, this layer produces anti-reflective properties depending on its optical parameters such as, for example, refractive index and layer thickness, which causes more light to be coupled to subsequent In the solar cell. These two effects can increase the conversion efficiency of the solar cell. A typical property of the layer currently in use is a layer thickness of ~80 nm when only the above-described tantalum nitride is utilized, and the tantalum nitride has a refractive index of about 2.05. This anti-reflection reduction is most pronounced in the wavelength region of light of 600 nm. Orientation and non-directional reflection herein show values of about 1% to 3% of the original incident light (perpendicularly perpendicular to the surface perpendicular to the germanium wafer).

The above-described tantalum nitride layer is currently deposited on the surface by means of a direct PECVD process. To this end, a plasma in which decane and ammonia are introduced is ignited in an argon atmosphere. The decane and ammonia react in the plasma via ion and radical reactions to produce tantalum nitride and simultaneously deposit on the wafer surface. The properties of the layers can be adjusted and controlled, for example, via individual gas streams of the reactants. The deposition of the tantalum nitride layer can also be carried out using hydrogen as a carrier gas and/or only a reactant. Typical deposition temperatures are in the range between 300 ° C and 400 ° C. Alternative deposition methods can be, for example, LPCVD and/or sputtering.

5) Manufacturing of front surface electrode grid

After depositing the anti-reflective layer, the front surface electrode is defined on the surface of the tantalum nitride coated wafer. In industrial practice, it has been established to fabricate electrodes by means of screen printing using a metal sintered paste. However, this is only one of many different possibilities for making the required metal contacts.

In screen printing metallization, high-concentration silver particles (silver content) are usually used. 80%) of the paste. The sum of the remaining components is derived from rheological aids necessary for formulating the paste, such as, for example, solvents, binders, and thickeners. In addition, the silver paste comprises a special frit mixture (usually an oxide and a mixed oxide based on cerium oxide), a borosilicate glass and also lead oxide and/or cerium oxide. The frit basically performs two functions: on the one hand, it acts as a adhesion promoter between the surface of the wafer and most of the silver particles to be sintered, and on the other hand, it is responsible for the penetration of the top layer of the tantalum nitride to promote direct ohmic with the underlying crucible. contact. The penetration of tantalum nitride occurs via the etching process and the silver dissolved in the frit substrate then diffuses into the crucible surface, thereby achieving ohmic contact formation. In practice, the silver paste is deposited on the surface of the wafer by screen printing and then dried at a temperature of about 200 ° C to 300 ° C for a few minutes. In terms of completeness, it should be mentioned that a two-layer printing process is also used in the industry, which allows the second electrode grid to be printed with precise registration onto the electrode grid produced during the first printing step. The thickness of the silver metallization is thus increased, which can have a positive effect on the conductivity of the electrode grid. During this drying, the solvent present in the paste is driven out of the paste. The printed wafer is then passed through a DC furnace. This type of furnace typically has a plurality of heating zones that are activatable and independently control the temperature. The wafers are heated to a temperature of up to about 950 °C during passivation of the DC furnace. However, individual wafers typically only experience this peak temperature for a few seconds. The wafer has a temperature of 600 ° C to 800 ° C during the remainder of the DC phase. At these temperatures, organic concomitant materials present in the silver paste, such as, for example, binder burnout, and etching of the tantalum nitride layer begin. During a short time interval in which the peak temperature prevails, a contact is formed with the crucible. The wafers are then cooled.

The contact forming process outlined in this manner is typically performed simultaneously with the formation of two remaining contacts (references 6 and 7), which is also referred to as the co-firing process. The reason for this.

The front surface electrode grid itself is made up of thin fingers having a width of typically 60 μm to 140 μm (typical number in the case of a sheet resistance > 50 Ω/sqr) 68) Also consists of bus bars having a width in the range of 1.2 mm to 2.2 mm, depending on the number (usually two or three). Typical heights of printed silver components are typically between 10 μm and 25 μm. The aspect ratio is rarely greater than 0.3, but can be significantly increased by selecting alternative and/or suitable metallization processes. Alternative metallization processes may be mentioned for the dispensing of metal pastes. A suitable metallization process is based on a two-dimensional screen printing process (double printing or print-on-print) of two different metal pastes. In particular, in the case of the last mentioned process, a so-called floating bus bar can be used which confirms that the current is dissipated from the fingers to collect the charge carriers, but it is not actually in direct ohmic contact with the germanium crystal itself.

6) Manufacture of the back surface bus bar

The back surface bus bars are also typically applied and defined by means of a screen printing process. For this purpose, a silver paste similar to that used for the front surface metallization is used. This paste has a similar composition but contains an alloy of silver and aluminum, of which the proportion of aluminum usually accounts for 2%. Additionally, this paste contains a lower frit content. The bus bars (typically two cells) are printed on the back of the wafer by screen printing having a typical width of 4 mm and compressed and sintered as described in point 5.

7) Manufacture of the back surface electrode

The back surface electrode is defined after printing of the bus bar. The electrode material consists of aluminum, which is why the aluminum-containing paste is printed on the remaining usable area of the back of the wafer by screen printing of the electrodes at <1 mm edge spacing. The paste consists of 80% aluminum composition. These remaining components are those which have been mentioned according to point 5 (such as, for example, solvents, binders, etc.). The aluminum paste is bonded to the wafer during the co-firing by melting of the aluminum particles during the heating period and dissolving in the molten aluminum from the wafer. The molten mixture acts as a dopant source and releases aluminum to the ruthenium (solubility limit: 0.016 atomic percent), wherein the ruthenium is p ⁺ doped due to this drive-in. During the cooling of the wafer, a eutectic mixture of aluminum and ruthenium (which cures at 577 ° C and has a composition of 0.12 mole fraction of Si) is especially deposited on the surface of the wafer.

As the aluminum is driven into the crucible, a highly doped p-type layer (which acts as a mirror ("electron microscopy") on the portion of the free charge carriers in the crucible) is formed on the back of the wafer. These charge carriers are unable to overcome this potential wall and are therefore very effectively kept away from the back wafer surface, which is thus evidenced by the overall reduced rate of recombination of charge carriers at this surface. This potential wall is often referred to as the "back surface field."

The order of the process steps described in accordance with points 5, 6 and 7 may, but is not required to, correspond to the order outlined herein. It will be apparent to those skilled in the art that the sequence of process steps outlined may in principle be carried out in any conceivable combination.

8) Optional edge insulation

If the edge insulation of the wafer has not been carried out as described in point 3, this is usually done by means of a laser beam method after co-firing. To this end, the laser beam is directed towards the front of the solar cell and the front surface p-n junction is separated by the energy coupled by the laser beam. Due to the laser action, a cutting groove having a depth of up to 15 μm is produced here.矽 is removed from the treated site via the ablation mechanism or thrown from the laser trench. This laser trench typically has a width of 30 μm to 60 μm and is about 200 μm from the edge of the solar cell.

After fabrication, solar cells are characterized and classified into individual performance categories based on their individual properties.

Those skilled in the art are aware of solar cell architectures that use both n-type and p-type base materials. These types of solar cells include, inter alia:

̇PERC solar cell

̇PERT solar cell

̇PERL solar cell

̇MWT solar cell

MWT-PERC, MWT-PERT and MWT-PERL solar cells derived from them

双面Double-sided solar cell with uniform and selective back surface field

̇Back surface contact battery

接触 Use the back surface of the interdigitated contact to contact the battery.

The so-called PERC (passive emitter back contact) or LBSF (partial back surface field) cells represent a further development of a more efficient standard aluminum BSF (back surface field) solar cell. The standard aluminum BSF battery has been fabricated as explained above. This type of battery is presented in cross section (not shown in actual scale) as shown in FIG. The basic element of the battery consists of a front surface electrode grid that delivers charge carriers to the outside of the battery via an external current circuit. Directly connected to the electrode grid is a so-called emitter that collects charge carriers (electrons and holes) generated by the incidence of light and ultimately due to its absorption. In general, if a manufacturing technique that is still prevalent in a solar cell having a p-type substrate is used as a starting point, then these are electrons. The emitter is produced in a volume, a substrate or an absorber (also referred to in simplified terms) (in this case, a minority of charge carriers have a short finite lifetime in the region of a few microseconds to hundreds of microseconds) The electrons are collected, causing them to become the majority of charge carriers in this region at this time and to cause them to be dissipated through the electrode grid via the potential difference prevailing in the latter. In the substrate, or also in the absorber, most of the light incident on the cell has been absorbed so far, resulting in the above-mentioned electron/hole pair. In the case of a standard aluminum BSF battery, the back surface of the substrate has a highly doped p-type region (back surface field). The highly doped region acts on the electrons generated in the substrate in a manner comparable to a type of mirror: in this region it is thrown back (→ "reflected") due to an increase in the dopant concentration gradient, Because it can't overcome this potential wall. This back surface field can be produced in a variety of ways. However, in industrial practice, the technology of alloying aluminum alloys into germanium crystals has become a substantial and substantial position. For this purpose, as has also been described, a screen printable aluminum paste is printed on almost the entire back surface of the tantalum wafer and alloyed to tantalum due to heat treatment (co-firing) occurring at a later point in time. In the crystal. Form various areas and phases here: In addition to the eutectic mixture consisting of aluminum and ruthenium, the lanthanum-rich aluminum also has a highly doped p-type region in the ruthenium crystal composed of aluminum which has diffused therein. This region may have a thickness in the range of 6 μm to 8 μm. In addition to the "reflection" of electrons, this area collects holes. These can theoretically then traverse the back surface field into adjacent aluminum containing stages of different compositions and dissipate via externally coupled current circuits via the remaining aluminum layer on the solar cell.

Figure 1a shows a graphical cross section (not to scale) through a standard aluminum BSF solar cell (back surface bus bar not shown).

Figure 1b shows a graphical cross section (not to scale) through a PERC solar cell (back surface bus bar not shown).

PERC batteries now represent a further development of this standard aluminum BSF battery in which at least another layer (dielectric) is added to the aluminum metallization applied to the back surface of the solar cell, which in this case is located in tantalum crystals and printed aluminum between. The dielectric layer is partially opened to promote contact of the crucible with the aluminum paste. In the case of using an aluminum paste, this local metallization function is the same as is known in the case of alloying the entire area of a standard aluminum BSF solar cell. For this reason, the dielectric applied to the back surface must be sufficiently stable to the process conditions for alloying the aluminum alloy into the crucible so that the hot paste is impermeable to the dielectric (at least incomplete). When used in the manufacture of PERC solar cells, the back surface dielectric achieves at least two important functions: 1) it is responsible for the electronic surface passivation of the wafer back surface of the solar cell and thus reduces the surface recombination rate. The latter is between 500 cm/s and 1000 cm/s in the case of standard aluminum BSF batteries. It can be reduced to 10 cm/s or even less due to the use of a suitable dielectric passivation layer. A decrease in the rate of surface recombination results in a decrease in dark saturation current (I ₀ ) or dark saturation current density in the solar cell. This dark current can be thought of as a parasitic current whose direction is opposite to the photocurrent due to absorption of solar radiation. Therefore, the effective maximum current (short circuit current density (I _SC )) that can be taken from the solar cell is therefore higher, the dark current is lower, or the photocurrent generated is higher. If a concentrating element is not used, the maximum current cannot be increased as needed because it is pre-specified by the sunshine conditions and the solar cell architecture and hence its characteristics. Therefore, many optimizations of existing solar cell types have the goal of minimizing losses and thus minimizing dark current. If (by definition) the dark current is minimized, the maximum voltage that can be achieved (open clamp voltage (V _OC )) is increased according to equation [1]. Passivation of the back surface (substrate) of the solar cell thus results in an increase in voltage. Considering that the maximum achievable short circuit current density will remain constant herein, the increase in efficiency in the case of solar cells fabricated in this manner will therefore be apparent. In general, however, the reduction in surface recombination rate also has a slightly lesser effect on the maximum achievable short circuit current density, such that an increase in the efficiency of the solar cell is expected to be attributed to an improvement or optimization of the two parameters, wherein The increase in voltage will be the dominant factor and has been the subject of more attention so far. If the equation [1] is compared for two solar cells (standard aluminum BSF solar cells and PERC solar cells), the relative increase in the voltage of the PERC cell compared to the voltage of a standard aluminum BSF cell can then be estimated from the equation [2]. And the characteristic data given in [3] and its equal ratio. It is assumed that the same maximum short-circuit current (corresponding to a battery area of 37.5 mA/cm ² ) for both battery types and assuming that the standard aluminum BSF battery has a dark current of 3*10 ^-13 A/cm ² and that the PERC battery has 2 *10 ^-13 A/cm ² dark current, the dark current of the PERC battery is reduced by 33% compared to the standard aluminum BSF battery, and the voltage of the PERC battery is increased by 1.5% compared to the standard aluminum BSF battery. . This corresponds to a 650 mV voltage of the PERC battery, which is an increase in efficiency of 640 mV of the standard aluminum BSF battery, or 0.3% absolute value (19.16% vs. 19.46%). Although in the case of some general choices, the latter corresponds to the examples as described in the relevant literature.

In general, however, in some cases the efficiency is described to increase even more. This occurs if the increase in short-circuit current density is also considered. This is in some cases derived from the situation outlined above. Furthermore, this is the second important function of the dielectric passivation layer, which acts like an optical mirror on the back surface of the solar cell and thus particularly improves the back surface reflectivity in the solar cell. The light that can be emitted by the solar cell is better reflected on the germanium/dielectric interface than the interface on the aluminum. This is for long-wave radiation (λ 900nm). The indirect semiconductor has a relatively low absorption coefficient for radiation having a relatively long wavelength, in other words, it causes the absorption length to increase as the wavelength of the incident light increases. The absorption length can be greatly increased in this manner until the light can completely illuminate through the thickness of the wafer (refer to Figure 2).

Figure 2 shows the wavelength dependent transmission (calculated by the transfer matrix method) of a polished germanium wafer covering the front surface with 80 nm of SiN _x (280 to 1100 nm). The thickness of the germanium wafer was set to 180 μm in the calculation. The transmission of the wafer is 1% at a wavelength of 925 nm.

The reflectance in the closed model calculation at the tantalum/cerium dioxide interface is not necessarily better than the reflectance at the tantalum/aluminum interface in the case of vertical light incidence (ie, light incident perpendicular to the surface). However, this state varies with the angle of incidence to the interfaces, where it becomes apparent that the reflectivity of the yttrium/yttria interface (refer to Figure 3) is better than the reflectance of the yttrium/aluminum interface (refer to Figure 4). . The deviation from the normal incidence on the surface of the crucible is due to the surface texture: in the case of a CZ wafer, the surface is characterized by an irregularly arranged cone whose equilateral surface encloses 54° with a theoretically flat crucible surface. angle. In the case where the perpendicular light is incident on the theoretically flattened surface, this encloses an angle of 36° between the light ray and the side surface of the cone, which further results in a 54° angle of incidence with respect to the surface perpendicular to the side surfaces of the cones. Due to the refraction of the light ray (13°), an incident angle of 32° with respect to the surface perpendicular is present on the theoretically polished back surface of the wafer in the 1000 nm region. In the case of the above incident angle, the reflectance at the tantalum/yttria interface is significantly greater than the reflectance at the tantalum/aluminum interface. Due to the large reflectivity of the tantalum/niobium dioxide interface, light rays that normally penetrate through the entire wafer thickness and that can exit again on their back surface (wavelength) 900 nm) can be reflected more efficiently at this interface. The rays reflected at the interface can thus pass through the germanium crystal again, so that the likelihood of its (all) absorption is increased. When it is absorbed in the crucible, an additional electron/hole pair is generated which, after successful separation and collection, contributes to the current of the solar cell. Therefore, the maximum achievable short circuit current density of the solar cell will increase.

Figure 3 shows the calculated reflectance of the germanium/cerium oxide interface as a function of incident wavelength and angle of incidence of the polished germanium wafer. This calculation was carried out by means of a transfer matrix method (280 to 1100 nm). Light having a wavelength significantly less than 800 nm will not reach the ytterbium/yttrium dioxide interface in the case of a typical thickness of a germanium wafer (e.g., 180 μm or less) because it is completely absorbed in the ruthenium.

Figure 4 shows the calculated reflectance of the germanium/aluminum interface as a function of incident wavelength and angle of incidence of the polished germanium wafer. This calculation was carried out by means of a transfer matrix method (280 to 1100 nm). Light having a wavelength significantly less than 800 nm will not reach the tantalum/aluminum interface in the case of a typical thickness of the tantalum wafer (e.g., 180 μm or less) because it is completely absorbed in the crucible. The refractive index and absorption coefficient of aluminum are inconsistent with the refractive index and absorption coefficient of conventional screen-printed and fired aluminum screen printing pastes. In addition, the back surface of the tantalum solar cell is typically characterized by the following other regions/phases prior to the conversion to aluminum: highly doped BSF regions, eutectic aluminum/ruthenium phases.

Figure 5 shows a representation of the absorption of aluminum calculated by means of the transfer matrix method in the case of the following test structures: SiN _x (80 nm) / Si (180 μm) / Al (40 μm). For this calculation, assuming a polished silicon wafer, the incident angle on the test structure is 0° (parallel to the surface perpendicular) and the wavelength range covers 280 nm to 1100 nm. In the test structure, the regions/phases present on the back surface are ignored: the highly doped BSF region, the eutectic aluminum/矽 phase. On the front surface, the emitter area is not considered.

The dielectric interfacial layer is not only solely responsible for the improvement of the reflectivity of the back surface of the solar cell (more precisely, at the tantalum/cerium oxide interface), but also in the case of PERC cells. Parasitic absorption. The parasitic absorption reference test structure for aluminum is shown in Figure 5. This is 6.41 W/m ² over the entire wavelength range shown, or 0.80% of the incident radiation available in the wavelength range. Due to the screen or decoupling of aluminum from the surface of the germanium wafer in the case of PERC cells, the parasitic absorption of aluminum is significantly reduced. This combination increases the reflectivity at the germanium/cerium oxide interface and ultimately contributes to the short-circuit current. The increase in density and thus the increase in the efficiency of the solar cell (refer to the model in Table 1). If (for this) the equations [3] and the parameters used for this purpose are finally used again, the achievable voltage of the PERC battery is increased in the case of an increase in the short-circuit current density of 0.5 mA/cm ² . Only increase by 0.3mV. Referring to this example, it can be seen that the increase in efficiency in the case of the PERC structure is substantially due to surface passivation and thus dark current saturation density.

effect). In practice, a large difference in short circuit current density between two cells is expected. It can be as high as 0.5 mA/cm ² (herein: 0.13 mA/cm ² ).

Inserting a dielectric layer between the crucible and the aluminum thus causes the two main positive effects to be achieved: on the one hand, the electronic passivation of the wafer surface, and on the other hand, the reflectivity of the incident light at the back surface of the wafer is increased and thus it is in aluminum The parasitic absorption in the layer is reduced. The first effect helps to reduce the dark current saturation density in the battery substrate, which primarily results in a higher achievable voltage of the battery and a higher current to a lesser extent. The second effect contributes to the higher current of the battery. To contact the battery on the back surface, the dielectric layer must be partially open to the back surface to make contact with the aluminum at the top of the dielectric. Various methods can be used to facilitate such local contact openings: photolithography, resist printing and subsequent immersion etching, printing of etched media, dielectric deposition via shadow masking, and laser ablation of the layer.

In general, it has proven advantageous to have a dielectric on the back surface of the wafer consisting of more than one layer (corresponding to a layer stack). This typically involves two layers, the first of which is deposited directly onto the wafer surface to perform the function of dielectric surface passivation. This layer typically has a low thickness of a few nanometers (5 to 10 nm). Since such a thin layer is generally incapable of withstanding the metallization (ie, its alloying-in) process using an aluminum paste, it is also melted and dissolved in aluminum at the same time and thus its passivation is of course significant, The actual passivation layer is covered by at least one other cover layer having a layer thickness that is several times greater than the layer thickness of the actual passivation layer itself. These coatings must be sufficiently resistant on the one hand to withstand the alloying process using an aluminum paste, and on the other hand must ensure that the aluminum paste promotes sufficient adhesion. Typical thicknesses of such cap layers are between 70 nm and 200 nm. In industrial practice, the use of SiN _x is merely has established itself as a covering material, wherein the means of PECVD processes typically SiNx is deposited on the passivation layer. Conversely, the passivation layers are typically composed of SiO ₂ , Al ₂ O ₃ , and in some cases amorphous yttrium (a-Si), and have also described or described amorphous lanthanum carbide (a-SiC). Use. Moreover, the benefit of the passivation layer being affected by the overlying cap layer is that the hydrogen included in the cap layer can be released into the dielectric passivation layer below the cap layer. This hydrogen is saturable and thus the passivation defect sites are present at the interface between the ruthenium and the passivation layer.

In the context of the implicit formula described above, the manufacture of a PERC solar cell thus includes the following process steps:

1. Cutting damage removal and texture

2. Diffusion and doping

3. One side of the back surface is polished and etched

4. The anti-reflection layer is deposited on the front surface

5. Dielectric passivation is deposited on the back surface

6. The cover layer is deposited on the dielectric passivation

7. Local contact openings in the cap layer and also in the dielectric passivation layer

8. Metallization printing and co-firing

Needless to say, the modification of the above program (which however gives the same result (ie PERC solar cells)) may be possible. The individual steps indicated in the program column have different ratios of the total manufacturing cost of the PERC solar cell. The most expensive individual steps are as follows: metallization printing and co-firing, deposition of a dielectric passivation layer and deposition of a capping material and deposition of a wet chemical texture and wet chemical polishing. There are many alternatives to the industry standard, the industry standard for metallizing solar cell wafers by means of a screen printing process, but these have not been established for large-scale manufacturing to date for cost reasons. Metallized printing can therefore be difficult. Regarding the efficiency of the subsequent solar cell, the texturing and polishing of the wafer can be omitted as well, leaving the deposition of the passivation layer as the final main cost lever, which can be provided in industrial large-scale manufacturing due to targeted optimization. The possibility of cost savings. This block cost will primarily drive the capital expenditures for the vacuum deposition units necessary for these steps. In the case where Al ₂ O _{3 is} used as the dielectric passivation layer, the consumption of the precursor gas before the generation of Al ₂ O ₃ by vacuum deposition on the wafer surface can be recognized as another significant cost factor. In the case of the deposition of the Al ₂ O _3, Al ₂ O ₃ is typically generated in the surface of the wafer using trimethylaluminum (TMA). The use of other alanes has also been described in the literature, but it has not been as popular as trimethylaluminum so far. The industrial acceptability of the manufacture of PERC solar cells is therefore associated with the manufacturing costs of such components, and is therefore further associated with the possibility of profitably selling such solar cells in the market, characterized by sustained pressures of reduced costs.

Purpose of the invention

Accordingly, it is an object of the present invention to provide a method for meeting the requirements of industrial large-scale manufacturing: in a simple manner by means of an inexpensive and robust printing process, it is non-corrosive in the preparation of equipment necessary for plants and their applications and at high temperatures (up to 900 ° C) It is possible to produce a hybrid gel which does not allow the aluminum to diffuse and diffuse, and which is resistant to alloying (and has a passivation effect) on the surface of the crucible.

The problems described above can be unexpectedly solved by a novel printable hybrid gel based on a group selected from the group consisting of ceria, alumina, tin oxide, tin dioxide and titanium dioxide. Driven; produced by a sol-gel technique; printed on a crucible surface in a structured manner by means of a screen printing process or another application process for the purpose of manufacturing a solar cell, preferably a so-called PERC solar cell Or printing onto the surface of the electronically passivated crucible and then dried, and then optionally coated by means of a screen printable aluminum paste and processed in a subsequent (co)firing process, wherein the dried The hybrid gel inhibits the aluminum paste alloying and diffusing into the germanium wafer at the site where the hybrid gel has been printed, and the electronic passivation layer is additionally obtained in a functionalized form at exactly the same.

Mixing gel printable precursor of the present invention may be manufactured having the following oxide-based materials: a silicon dioxide: by symmetric and asymmetric mono- to tetrasubstituted of carboxyl, alkoxy and alkoxyalkyl Silane, And the like, the alkyl alkoxy decane, wherein the central ruthenium atom may have at least one [void] substitution degree directly bonded to the hydrogen atom of the ruthenium atom, such as, for example, triethoxy decane, and wherein the degree of substitution is further With respect to the amount of possible carboxyl groups and/or alkoxy groups present, which are equal to the alkyl group and/or the alkoxy group and/or the carboxyl group, each of which contains an individual or different saturated, unsaturated branched chain, unbranched chain fat Groups, alicyclic and aromatic groups, which may be further functionalized at any desired position of the alkyl, alkoxide or carboxyl group via a hetero atom selected from the group consisting of O, N, S, Cl and Br, And a mixture of the foregoing precursors, b. Alumina : a symmetrically and asymmetrically substituted aluminum alkoxide (alkoxide) such as triethoxyaluminum, triisopropoxyaluminum, tris-secondary butoxide aluminum , tributoxide aluminum, aluminum trispentoxide and aluminum triisolylate, ginseng -diketone)aluminum, such as aluminum acetylacetonate or 1,3-cyclohexanedicarboxylate, ruthenium (β-ketoester) aluminum, aluminum monoacetate monoalcoholate, ginseng (hydroxyquinoline) Alkyte, aluminum soap, such as monobasic and dibasic aluminum stearate and aluminum tristearate, aluminum carboxylate, such as alkaline aluminum acetate, aluminum triacetate, basic aluminum formate, aluminum triacetate and Aluminum trioctanoate, aluminum hydroxide, aluminum hydroxide and aluminum trichloride and the like, and mixtures thereof, c. tin oxide (II, IV) : tin alkoxide, such as tetraisopropoxy tin and tetra- Third-butoxytin, tin carboxylate, such as tin diacetate, tin oxalate, tin tetraethoxide, alkyl tin carboxylate, such as dibutyltin diacetate, tin hydroxide and the like, and mixtures thereof, d. titanium dioxide : titanium alkoxides, such as titanium ethoxide, titanium tetraisopropoxide, titanium tetrabutoxide, (triethanolamine) titanium triisopropoxide, bis(triethanolamine) titanium diisopropoxide and Tetraoctyloxytitanium, titanium hydroxide, β-diketone titanium, such as acetonitrile oxytitanium oxide, monoethyl hydrazine acetone triisopropoxy titanium, bis(acetonitrile) titanium diisopropoxide, titanium hydroxide Carboxy Titanium, such as di-titanium diisopropoxydiacetate and the like, and mixtures thereof, wherein the precursors and mixtures thereof are in the presence of water or under anhydrous conditions by means of the sol-gel technique Partial or complete intra- and/or inter-species condensation (simultaneous or sequential), and the degree of gelation of the resulting hybrid gel may be due to the set condensation conditions (such as precursor concentration, water content, catalyst content, reaction) Temperature and time, the addition of condensation control agents (such as, for example, various combinations of the above-described various chelating agents and chelating agents, various solvents) and their individual volume fractions, as well as the specific elimination of volatile reaction auxiliaries and unfavorable by-products, are specifically controlled and It is affected in the desired manner to produce a formulation that is stable in storage, extremely easy to screen printable, and pressure stable and therefore sufficiently shear stable.

The printable hybrid gel of the present invention can be affected in terms of its degree of condensation by selecting suitable reaction conditions such that it is present in the form of a high viscosity mixture which can be used in the manner claimed. A printing process (preferably a screen printing process) of the mixture (usually in the form of a paste or paste) is applied and applied to the substrate. The hybrid gel may additionally comprise a polymeric thickener and an inorganic particulate additive such as, for example, SnO ₂ , SiC, BN, Al ₂ O ₃ , SiO ₂ , Al ₂ TiO ₅ , TiO ₂ , TiC, Si ₃ N ₄ , TiN and Ti _x C _y N _z , which are used to positively influence the rheological properties and to adjust the layer thickness, which additionally have an advantageous effect on the layer resistance and scratch resistance properties of the resulting layer.

According to the invention, the printable hybrid gel-based paste can be printed in a structured manner for the purpose of producing a solar cell, preferably a so-called PERC solar cell, preferably by means of a screen printing process. Surface or electronically passivated tantalum surface, and then dried, and which can also be applied to the entire area by a thin layer of aluminum by means of a PVD process and then subject the known structure to a local laser melting process, such that aluminum Specific alloying to unprotected crucibles (laser-fired contact cells).

A printable hybrid gel which can be prepared based on an oxide precursor of only tin dioxide and aluminum oxide can be printed on the surface of the crucible and dried, and the gel has both electronic surface passivation and also acts as an alloy of aluminum and A barrier that spreads into the crucible below the layer.

In a preferred embodiment, the printable hybrid gel of the present invention is based solely on oxide precursors of ceria, alumina, and titania and is printed on the surface of the crucible as described above and dried, and has both an electronic surface. Passivation and also acts as a barrier to the alloying of aluminum and diffusion into the crucible below the layer. In particular, the printable hybrid gel of the present invention improves and increases the internal back surface reflectivity in a solar cell, preferably a so-called PERC solar cell, wherein the reflectance can be based on the primary oxide precursor used to make it. The concentration ratio is specifically adjusted within a wide range. Thus, the printable hybrid gel forms an electrically insulating barrier between the two electrical contacts on the surface of the tantalum wafer after drying, or typically on the surface. Floor. In particular, the layers obtained after hybrid gel printing and drying are in microelectronics and microelectromechanical (MEMS) components, thin film solar cells, thin film solar modules, organic solar cells, printed circuits and organic electronic devices, based on thin films. The display elements of the technology of a transistor (TFT), a liquid crystal (LCD), an organic light-emitting diode (OLED), and a contact-sensitive capacitive and resistive sensor act as a scratch-resistant layer and a corrosion-resistant layer and a reflection-reducing layer. The hybrid gels of the present invention can thus be advantageously used in the manufacture of such components.

Figure 1a shows a graphical cross section (not to scale) through a standard aluminum BSF solar cell (back surface bus bar not shown).

Figure 1b shows a graphical cross section (not to scale) through a PERC solar cell (back surface bus bar not shown).

Figure 2 shows the wavelength dependent transmission (calculated by the transfer matrix method) of a polished germanium wafer covering the front surface with 80 nm of SiN _x (280 to 1100 nm). The thickness of the germanium wafer was set to 180 μm in the calculation. The transmission of the wafer is 1% at a wavelength of 925 nm.

Figure 3 shows the calculated reflectance of the germanium/cerium oxide interface as a function of incident wavelength and angle of incidence of the polished germanium wafer. This calculation was carried out by means of a transfer matrix method (280 to 1100 nm). Light having a wavelength significantly less than 800 nm will not reach the ytterbium/yttrium dioxide interface in the case of a typical thickness of a germanium wafer (e.g., 180 μm or less) because it is completely absorbed in the ruthenium.

Figure 4 shows the calculated reflectance of the germanium/aluminum interface as a function of incident wavelength and angle of incidence of the polished germanium wafer. This calculation was carried out by means of a transfer matrix method (280 to 1100 nm). Light having a wavelength significantly less than 800 nm will not reach the tantalum/aluminum interface in the case of a typical thickness of the tantalum wafer (e.g., 180 μm or less) because it is completely absorbed in the crucible. The refractive index and absorption coefficient of aluminum are inconsistent with the refractive index and absorption coefficient of conventional screen-printed and fired aluminum screen printing pastes. In addition, the back surface of a tantalum solar cell is typically characterized by the following other regions/phases prior to conversion to aluminum: highly doped BSF region, eutectic aluminum/矽 phase.

Figure 5 shows a representation of the absorption of aluminum calculated by means of the transfer matrix method in the case of the following test structures: SiN _x (80 nm) / Si (180 μm) / Al (40 μm). For this calculation, assuming a polished silicon wafer, the incident angle on the test structure is 0° (parallel to the surface perpendicular) and the wavelength range covers 280 nm to 1100 nm. In the test structure, the regions/phases present on the back surface are ignored: the highly doped BSF region, the eutectic aluminum/矽 phase. On the front surface, the emitter area is not considered.

Figure 6 shows an arrangement (center) printed on a wafer etched in a smooth manner using a screen 1 (according to Table 2). The line (right) and the gap between the surface or in the plurality of lines (left) were dried in 10 ° C for 10 min, in each case of light micrographs.

Figure 7 shows the arrangement of printing on a polished wafer by means of a screen 2 (according to Table 2) on a hot plate at 400 ° C for 10 minutes after printing.

Unexpectedly, it has been found that a hybrid gel composed of a mixture of cerium oxide, aluminum oxide, tin oxide, tin dioxide and titanium dioxide precursors is excellently resistant to aluminum after it is deposited on the surface of the crucible and dried. Alloying of the paste. These mixed sols and gels can be prepared by means of a classical sol-gel process and controlled in such a way that the complete bandwidth between the low viscosity and high viscosity formulations can be set. In the following connections, low viscosity formulations (where low viscosity is intended to mean The arbitrarily selected limit of the dynamic viscosity of 100 mPa*s is called ink. Thus, high viscosity formulations (i.e., those whose dynamic viscosity is then above the above limit of 100 mPa*s) are referred to as pastes.

The mixed sol and the gel may contain an alkoxide mixed in any ratio in the above-mentioned compounds (cerium oxide, aluminum oxide, tin oxide, tin dioxide, titanium dioxide precursor), but this is not necessarily the case. Combinations of only three or two of these classes of compounds can be envisaged in the same manner. Moreover, the possibility of combination is not limited by these examples: in the case of additional components, other materials which provide a sol having a favorable property and a gel may be present in the mixture. In sols and gels. These may be: oxides, basic oxides, hydroxides, alkoxides, carboxylates, β-diketones, β-ketoesters, cerates, and cerium, zirconium, hafnium, zinc, lanthanum, gallium Analogs of ruthenium, osmium, boron and phosphorus, which may be used directly or precondensed in sol-gel synthesis. The hybrid sols and gels can be applied to the surface of the electronically passivated germanium wafer or onto the surface of the germanium wafer by means of a printing and coating process. Suitable methods for this purpose may be: rotary or dip coating, drop coating, curtain coating or slot die coating, screen or flexographic printing, gravure printing, ink jet printing or aerosol inkjet Printing, lithography, microcontact printing, electrohydrodynamic dosing, roll or spray, ultrasonic spray, pipe spray, laser transfer printing, pad printing or rotary screen printing. The printing of the hybrid gels is preferably carried out using a screen printing process. The hybrid gels printed on the surface of the electronically passivated germanium wafer or printed on the surface of the germanium wafer are subjected to a drying step after they are deposited. This drying can be, but is not required to be, carried out in a DC oven. During the drying of the gels, these are compressed to produce a uniform and impermeable glass layer due to the removal of the solvent, as well as thermal degradation of the formulation aid and thermal degradation of the oxide precursor. This drying can be achieved at temperatures up to 600 ° C (but preferably from 200 ° C to 400 ° C). The layers printed on the surface of the electronically passivated germanium wafer or printed on the surface of the germanium wafer and dried may be applied over the entire area during the process. After drying, the layers are resistant to alloying of the aluminum paste, which itself is printed on the layers during the co-firing process and then compressed and sintered. During this process, the hybrid gel printed on the layer can be subjected to another drying or compression. The hybrid gels are preferably applied to the surface of the electronically passivated wafer or applied to the surface of the germanium wafer during screen printing using a structured screen that has been found to be suitable. The structured screens herein preferably have structural features that are important for subsequent formation of contact between the aluminum paste and the tantalum wafer, i.e., opening in the barrier layer to be created, wherein the electrons are The surface of the passivated germanium wafer or the surface of the germanium wafer can be locally contacted (LBSF structure) by alloying of the aluminum paste applied in the subsequent step. Needless to say, this structure is freely selectable and is only limited by the requirements for making a sufficiently efficient solar cell. Structural printing of the barrier layer The brush makes the step of partially contacting the opening as necessary in the case of dielectric passivation and deposition of the cover layer redundant. The electronic surface passivation that is already present under the barrier layer is infiltrated by the alloyed aluminum paste during the contact formation step. If the barrier layer that can be printed based on the hybrid gel can also electronically passivate the surface of the crucible, the deposition of the dielectric passivation layer can be further omitted.

The simple structured printing as a hybrid gel for the barrier layer of alloying resistant to aluminum therefore has a significant cost advantage over the vacuum deposition already described: inexpensive structured printing of the barrier layer and then The process for achieving the local contact opening is omitted. If the printed barrier layer has sufficient electronic surface passivation properties, the deposition of dielectric passivation or its deposition can be further omitted by a significantly less expensive printing step and drying step.

These mixed gels can be prepared by means of sol-gel synthesis which is anhydrous and also contains water. As other auxiliaries in the formulation of such gels, the following materials can be advantageously used: ̇ surfactants, surface-active compounds for influencing the wetting and drying behavior, 消 defoamers for affecting drying behavior and Deaerator, 高 used to influence particle size distribution, pre-condensation degree, condensation, wetting and drying behavior and high-boiling and low-boiling polar protic and aprotic solvents, erbium and chelate compounds, Such as, for example, acetoacetone, 1,3-cyclohexanedione, dihydroxybenzoic acid, acetaldehyde oxime isomeric compounds, and further mentioned in the patent applications EP 12703458.5 and EP 12704232.3, etc. All of them may be separated or used in a mixture of such materials, granule additives for affecting rheological properties, granule additives (such as hydrogen) for affecting the dry film thickness and morphology thereof after drying. Alumina and alumina, colloidal precipitated or highly dispersed cerium oxide, tin dioxide, boron nitride, tantalum carbide, tantalum nitride, aluminum titanate, titanium dioxide, titanium carbide, titanium nitride, titanium carbonitride ), ̇ is used for influence The particulate additive scratch resistance of the film (such as aluminum hydroxide and aluminum, or by gummy precipitate of highly dispersed silicon dioxide, tin dioxide, boron carbide, Antimony, tantalum nitride, aluminum titanate, titanium dioxide, titanium carbide, titanium nitride, titanium carbonitride), cerium oxide, hydroxide, basic oxide, acetate, alkoxylation Pre-condensed alkoxides of sulphate, bismuth, gallium, gallium, antimony, bismuth, tin, antimony, antimony, bismuth, antimony, bismuth, antimony, bismuth, antimony, bismuth, bismuth Such as, for example, polyvinylpyrrolidone, hydroxyethyl cellulose, methyl cellulose and ethyl cellulose, polyacrylate and polyacrylic acid, polyvinyl alcohol, polyvinyl acetate, polyvinyl isobutyrate and other.

A method for making a suitable paste of the present invention by pre-dissolving a ceria precursor such as, for example, tetraethyl orthosilicate, in a solvent or solvent mixture (preferably selected from high boiling point ethylene) It is carried out in an alcohol ether or a group of preferably high-boiling glycol ethers and alcohols. After dissolving the ceria precursor, water and acetic acid or carboxylic acid are preferably used or added to the solution in the desired amount, and after the addition, the mixture is refluxed at a temperature between 100 ° C and 130 ° C for three hours. The precondensation solution of the cerium oxide precursor is then added to the precondensed mixture having tetraethyl orthotitanate in the case of advantageous use of the titanium dioxide precursor, which itself is preferably dissolved or dissolved in the high boiling diol, A mixture of alcohol ethers or a mixture of high boiling diols and/or glycol ethers and alcohols. The complete reaction mixture was then refluxed for another hour. A chaotic or chelating agent such as, for example, 1,3-cyclohexanedione and 3,5-dihydroxybenzoic acid and possibly others, is then added to the reaction mixture in the required amount and dissolved until a completely clear and clear solution is obtained. . Alternatively, and depending on the concentration of the reactants participating in the reaction, a suitable alumina precursor (such as, for example, tri-second aluminum butoxide) may also be added first to the reaction mixture, and the latter immediately after addition It can be further carried out only with the above-mentioned intercalating agents and chelating agents. The alumina precursor is advantageously pre-dissolved in a solvent that positively affects the rheology and printability of the finished formulation, such as, for example, Texanol and the like or solvent mixture, and is transferred to the reaction mixture having the latter. The reaction mixture which was gradually completed in this manner was refluxed again. The volatile component formed in the reaction can be distilled off, for example, by means of a water separator during reflux. Alternatively, the removal thereof may also be carried out in the vacuum distillation step after completion of the reaction, here In this case, the mixture to be distilled is advantageously heated to 70 ° C and the vacuum applied to the apparatus is gradually reduced to a final pressure of 10 mbar. The gelled paste may then, but not necessarily, be added by adding a polymerization aid which advantageously affects rheology and printability (such as, for example, various polyvinylpyrrolidone, various ethylcellulose or methyl groups). Cellulose or a material already mentioned elsewhere), and also by the additional use of other particulate additives (such as aluminum hydroxide and alumina, which are also beneficially affecting rheology, colloidally precipitated or highly dispersed) Cerium oxide, tin dioxide, boron nitride, tantalum carbide, tantalum nitride, aluminum titanate, titanium dioxide, titanium carbide, titanium nitride or titanium carbonitride are further adjusted and affected in the desired manner. If the introduction and mixing of the above-mentioned auxiliaries and additives necessitate the use of low-boiling solvents to achieve rapid establishment of their homogenization, they are removed by distillation under reduced pressure after introduction and homogenization. In the case of fabricating a hybrid gel based on the use of tin oxide prior to the use of the auxiliaries and additives described herein, it is also possible to add the corresponding tin dioxide precursor during the step of adding the additive to the paste mixture. (such as, for example, dibutyltin diacetate). If the advantageous use of the additive is necessary for the non-paste formulation, the tin oxide precursor may be added to the mixture at the end of the reaction to the addition of the gelled paste or the combined alumina precursor, and thus Incorporation into the hybrid gels.

If the paste synthesis is carried out in the presence of a mixture of high-boiling diols or glycol ethers and in the presence of an alcohol, the alcohols should undergo rheology and printability after the distillation loss. Sexually exert a beneficial effect on solvent replacement. Such solvents may be (as exemplified by way of example, but not claiming completeness): dibenzyl ether, butyl benzoate, benzyl benzoate, Texanol, terpineol, 2-pyrrolidone, ethylene glycol , 1,3-propanediol, diol, glycol ether and glycol ether carboxylate, glycol ester, polyethylene glycol and others.

Alternative synthetic methods based on the same solvents and solvent mixtures as described in more detail above and the reactants to be used (such as, for example, water and acetic acid and the like) include the simultaneous addition of the desired oxide precursor to the reaction solution. (such as cerium oxide, aluminum oxide, titanium dioxide, and tin oxide precursors), adding a suitable β-diketone to the reaction solution, such as Acetylacetone, or, for example, 1,3-cyclohexanedione, α- and β-ketocarboxylic acids and esters thereof such as, for example, pyruvic acid and its esters, ethyl acetate, ethyl acetate, dihydroxybenzoic acid, And other reference compounds of this type, as well as any desired mixtures of the above-mentioned complexing agents and chelating agents and controlling agents for controlling the degree of condensation, and heating the mixture at, for example, 80 ° C for up to 24 h and causing a reaction. The volatile and desired parasitic by-products present in this type of reaction are removed from the finished reaction mixture by means of vacuum distillation. This vacuum distillation is carried out under the conditions set forth above. The hybrid gels are then adjusted for their desired properties by specific addition of a suitable solvent which facilitates the rheology and printability of the paste, and if necessary diluted. Furthermore, the paste stream change may, but need not, be adjusted and trimmed accordingly by means of the auxiliaries and additives which have also been described in detail above, depending on the particular requirements.

Instance

Example 1:

A 1:1 mixture of 3.6 g of acetic acid and 10 g of EGB (ethylene glycol monobutyl ether) and ethanol was initially introduced, and 4 g of TEOS (tetraethyl orthosilicate) was dissolved therein. Next, 1 ml of water was dissolved in another 5 g of EGB/EtOH (1:1) and added dropwise to the reaction solution. The entire reaction solution was refluxed at 100 ° C for 3.5 h. Then 9.1 g of TEOT (tetraethyl orthotitanate) was pre-dissolved in 42 g of EGB/EtOH (1:1) and slowly added dropwise, and the whole mixture was refluxed for 30 min. A solution of 2.2 g of 1,3-cyclohexanedione and 3 g of 3,5-dihydroxybenzoic acid in 20 g of EGB/EtOH (1:1) was prepared. In the final step, 9.8 g of ASB (aluminum tris-butoxide) was predissolved in 20 g of EGB/EtOH (1:1) and slowly added dropwise. Immediately thereafter, the resulting solution (solution of the wrong agent) was added, and the mixture was refluxed for another 30 min. In the distillation solvent exchange step, the ethanol was replaced with Texanol in a rotary evaporator. Immediately thereafter, a thickener solution consisting of 10 g of ethanol and 2 g of PVP (polyvinylpyrrolidone) K30 was added. Stir vigorously to make the mixture uniform. The thickener solvent was stripped again in a rotary evaporator. The viscosity of the paste was then 4 Pa*s to 5 Pa*s (shear rate: 25 s ^-1 , T = 23 ° C). The screen printable paste was printed on a <100> CZ alkaline textured n-type wafer. The screen printing was performed using a stainless steel wire mesh having 280 mesh counts per inch, 25 μm wire thickness, and 12 μm emulsion thickness (ISAR). The squeegee used was a trailing end blade squeegee having a Shore hardness of 35°. The printing was carried out at a speed of 170 mm/s, a squeegee pressure of 1.1 bar and an interval of 1 mm. The chosen arrangement is 5 x 5 cm squared. After printing, the wafers were pre-dried at 100 ° C for 5 min on a standard laboratory hot plate. The thickness of the layer formed was determined to be 2 μm by means of scanning electron microscopy.

Example 2:

10 g of EGB, 4.08 g of TEOS and 3.6 g of acetic acid were mixed, and a mixture of 0.75 g of water and 5 g of EGB was slowly added dropwise to the reaction solution at room temperature. The resulting solution was refluxed at 130 ° C for 3 h. Then 9.12 g of TEOT diluted in 10 g of EGB was added and the mixture was refluxed for a further 2 h. Initially, 40 g of EGB and 10.2 g of Texanol were introduced into a stirring apparatus, and 2.24 g of 1,3-cyclohexanedione and 3.08 g of 3,5-dihydroxybenzoic acid were dissolved therein at room temperature. A reaction solution containing TEOS and TEOT was slowly added thereto by means of a dropping funnel, and then 9.8 g of ASB predissolved in 15 g of EGB was added by means of a dropping funnel. An additional 24 g of EGB was then added and the mixture was refluxed at 130 ° C for 1 h. Volatile by-products were stripped off in a rotary evaporator during which a mass loss of 16.58 g was observed. Ethylcellulose (2.59 g) pre-dissolved in ethanol was added to the gelatinized intermediate and homogenized. The ethanol is then removed in a rotary evaporator. The finished paste had a viscosity of 740 mPa*s (shear rate: 25 s ^-1 , T = 23 ° C).

Example 3:

7.5 g of EGB, 0.36 g of water, 2.04 g of TEOS and 1.8 g of acetic acid were mixed and refluxed at 130 ° C for 3 h. Then 4.56 g of TEOT diluted with 5 g of EGB was added and the mixture was refluxed for an additional hour. To the reaction mixture were added 30 g of Texanol, 1.54 g of 3,5-dihydroxybenzoic acid and 1.12 g of 1,3-cyclohexanedione, and the whole mixture was further refluxed until a clear solution was formed, followed by slowing to the clear solution. 4.93 g of ASB dissolved in 5 g of EGB was added dropwise. The reaction mixture was diluted by further addition of 12.5 g of EGB and refluxed for 1.5 h, and volatile by-products were distilled off by means of a water separator during the reaction. After the reaction, 1.47 g of ethyl cellulose dissolved in ethanol was homogenized in the paste. The still present ethanol is then removed by distillation on a rotary evaporator. This batch produced a printable paste having a viscosity between 8 Pa*s and 9 Pa*s (shear rate: 75 s ^-1 , T = 23 ° C). The screen printable paste was printed on the textured polysilicon wafer under the same printing conditions as described under Example 1. After printing, the wafers were dried directly on a standard laboratory hot plate at 400 ° C for 10 min.

Example 4:

19.90 g of ASB was dissolved in 55 g of EGB and 7.4 g of acetic acid, and 0.44 g of acetamidineacetone and 0.78 g of acetaldehyde oxime were added to the mixture. Next, a mixture of 1.45 g of water and 5 g of EGB was added to the reaction solution with stirring at room temperature. The reaction solution was then refluxed at 80 ° C for 19 h. After the reaction, the volatile reaction product was stripped off in a rotary evaporator, during which a mass loss of 21.2 g was observed. The resulting high viscosity gel was diluted with 144 g of alpha-terpineol. 14.07 g of dibutyltin diacetate, 4.51 g of PVP K30 and 0.55 g of SiC were sequentially stirred in a stirrer in a stirrer and homogenized in each case. The flow curve of the paste is determined using a special measurement program in which a simple flow curve depending on the shear rate is first determined, followed by a flow rate curve depending on the shear rate during an extended loading phase (usually longer than 2 minutes) Immediately during the test, the paste was again measured, during which the shear rate of 300 s ^-1 was maintained (measurement temperature = 23 ° C). The viscosity determined by this measurement program is 8.7 Pa*s (shear rate: 75 s ^-1 ), T = 23 ° C, 5.5 Pa * s (shear rate: 300 s ^-1 ) and 8.8 Pa * s (shear Rate: 75s ^-1 ). The paste was applied to a polished <100> CZ(R) wafer by means of a stainless steel hand applicator to produce a wet film thickness of 40 [mu]m and 50 [mu]m. The coated layers were dried on a standard laboratory hot plate at 400 ° C for 10 min, and the layer thicknesses of the films were measured using a surface leveler. In the case of applying a 50 μm wet film, a layer thickness of the dried stress-free crack of the 1260 nm glass film was achieved. In addition, the paste is printed on a CZ wafer that is etched smoothly and the CZ wafer side is polished. Four different screens (their properties are listed in Table 2) were used for screen printing.

The wafer printed using the screen 1 is shown in FIG.

Figure 6 shows an arrangement (center) printed on a wafer etched in a smooth manner using a screen 1 (according to Table 2). The line (right) and the gap between the surface or in the plurality of lines (left) were dried in 10 ° C for 10 min, in each case of light micrographs.

The polished CZ wafer is printed using screen 2. The film was dried on a standard laboratory hot plate at 400 ° C for 20 min. The arrangement used is shown in Figure 7.

Figure 7 shows the arrangement of printing on a polished wafer by means of a screen 2 (according to Table 2) on a hot plate at 400 ° C for 10 minutes after printing.

The layer thickness was determined to be 612 nm on one of the two squares and the other was determined to be 550 nm by means of a surface leveler. The layer thickness of the thin line (the right side in the figure) is 910 nm. The formed glass film is continuous and free of stress cracks. After heating at 900 ° C for an additional 1 minute, the films were shrunk to 490 nm, 360 nm and 675 nm in a prescribed order. Even after heating at high temperatures, the glass films are free of stress cracks.

The arrangement printed on the polished wafer surface by means of the screen 3 (according to Table 2) consists only of lines. The thickness of the glass layer measured by the surface leveler at 400 ° C for 10 minutes by means of a surface leveler was 418 nm.

A suitable paste formulation is also achieved by means of the following additives: 0.45 g α-SiN _x , 0.55 g aluminum titanate, 2.47 g TiN, 0.55 g ZrSiO ₄ , 0.55 g BN and in each case 0.55 g Al ₂ O ₃ replaces SiC. The remaining paste system was applied to the polished wafer by means of a stainless steel hand applicator and then dried at 400 ° C for 10 min. Table 3 gives an overview of the layer thickness and flow properties obtained for a paste.

Example 5:

20.5 g of ASB was dissolved in 55 g of DEGMEE (diethylene glycol monoethyl ether) and 7.25 g of acetic acid, 1.3 g of acetamidine acetone and 2 g of 3,5-dihydroxybenzoic acid were added to the solution, and the mixture was stirred until completeness was obtained. Clarify the solution. Next, a mixture of 1.44 g of water and 5 g of DEGMEE was slowly added dropwise to the reaction solution at room temperature. The fully reacted mixture was refluxed for 15 h in an oil bath maintained at a temperature of 80 °C. The volatile by-product was then removed from the reaction mixture by distillation at 70 ° C to a final pressure of 10 mbar. The resulting gel was diluted with 28.5 g of Texanol and 84.5 g of TEG (triethylene glycol). 5.2 g of Al ₂ O ₃ , 41.8 g of water and 0.9 g of acetic acid were stirred into the mixture and homogenized. Excess water and excess acetic acid were removed by stripping in a rotary evaporator. The viscosity of the paste was 2.5 Pa*s (shear rate: 25 s ^-1 , T = 23 ° C). The paste was applied to the polished CZ wafer by means of a stainless steel hand applicator to achieve a 30 μm wet film. After the coated wafer was subsequently dried at 300 ° C for 10 min, the layer thickness measured by means of a surface leveler was 1.1 μm. The wafer was subsequently heated at 900 ° C for 1 minute to form a 530 nm layer thickness of glass.

Claims (10)

A printable hybrid gel based on a group of precursors selected from the group consisting of cerium oxide, aluminum oxide, tin oxide, tin dioxide, and titanium dioxide; based on a sol-gel technique; The purpose of the (preferably so-called PERC solar cell) is printed in a structured manner on the surface of the crucible or on the surface of the electronically passivated crucible by means of a screen printing process or another application process and then dried, and then by means of a filament The screen printed aluminum paste is optionally coated and treated in a subsequent (co)fibrous process wherein the dried hybrid gel inhibits the aluminum paste at the site where the mixed gel has been printed. And diffusing into the germanium wafer, and the electronic passivation layer is additionally obtained in a functionalized form at exactly these equipotential points. A printable hybrid gel based on the following oxide material precursors: a. cerium oxide : symmetrical and asymmetrical mono-substituted to tetra-substituted carboxy-, alkoxy- and alkoxyalkyl decane, And the like specifically includes an alkyl alkoxy decane wherein the central ruthenium atom may have at least one [void] substitution degree of a hydrogen atom directly bonded to the ruthenium atom, such as, for example, triethoxy decane, and wherein the degree of substitution is further Corresponding to the number of possible carboxyl groups and/or alkoxy groups present, which are equal to the alkyl group and/or the alkoxy group and/or the carboxyl group, each containing a different or different saturated, unsaturated branched chain, unbranched chain Aliphatic, alicyclic, and aromatic groups, which may be further functionalized at any desired position of the alkyl, alkoxide or carboxyl group via a hetero atom selected from the group consisting of O, N, S, Cl, and Br. And a mixture of the foregoing precursors, b. alumina: a symmetrically and asymmetrically substituted aluminum alkoxide (alkoxide) such as triethoxyaluminum, triisopropoxyaluminum, tri-second butanol, Tributoxy aluminum, aluminum trispentoxide and aluminum triisolylate, ginseng (β-diketone) aluminum, Such as acetonitrile aluminum pyruvate or ginseng (1,3-cyclohexanedicarboxylate) aluminum, ginseng (β-ketoester) aluminum, aluminum acetoacetate monoalcoholate, ginseng (hydroxyquinolinate) aluminum, aluminum Soap, such as monobasic and dibasic aluminum stearate and aluminum tristearate, aluminum carboxylate, such as alkaline aluminum acetate, aluminum triacetate, basic aluminum formate, aluminum triacetate and aluminum trioctate, hydroxide Aluminum, aluminum partial hydroxide and aluminum trichloride and the like, and mixtures thereof, c. tin oxide (II, IV): tin alkoxides, such as tetraisopropoxy tin and tetra-tert-butoxide a tin carboxylate such as tin diacetate, tin oxalate, tin tetraethoxide, alkyl tin carboxylate such as dibutyltin diacetate, tin hydroxide and the like, and mixtures thereof, d. titanium dioxide: titanium alkoxide, such as Titanium ethoxide, titanium tetraisopropoxide, titanium tetrabutoxide, titanium (triethanolamine), titanium triisopropoxide, bis(triethanolamine) titanium diisopropoxide and titanium tetraoctyloxide, hydrogen Titanium oxide, β-diketone titanium, such as titanyl acetylacetonate, monoethyl acetonide, titanium triisopropoxide, bis(acetonitrile) titanium diisopropoxide, titanium hydroxide a titanium carboxylate, such as di-titanium dihexyloxydiacetic acid and the like, and mixtures thereof, wherein the precursors and mixtures thereof can be subjected to sol-gel technology under water-containing conditions or under anhydrous conditions Simultaneous or sequential partial or complete intra- and/or inter-species condensation, and the degree of gelation of the resulting hybrid gel may be due to the set condensation conditions (such as precursor concentration, water content, catalyst content, reaction). Temperature and time, the addition of condensation control agents (such as, for example, various combinations of the above-mentioned various chelating agents and chelating agents, various solvents) and their individual volume fractions, as well as the specific elimination of volatile reaction auxiliaries and adverse by-products, are specifically controlled and The desired mode of influence results in a formulation that is stable in storage, extremely easy to screen printable, and pressure stable and therefore sufficiently shear stable. A printable hybrid gel according to claim 1 or 2 which, in terms of its degree of condensation, can be influenced by the selection of suitable reaction conditions such that a high viscosity mixture is formed which can be used in the manner claimed a printing process (preferably a screen printing process) of the mixture (usually expressed as a paste or paste) and Apply to the substrate. The printable hybrid gel of any one of claims 1 to 3, which may additionally comprise a polymeric thickener and a particulate additive such as, for example, SnO ₂ , SiC, BN, Al ₂ O ₃ , SiO ₂ , Al ₂ TiO ₅ , TiO ₂ , TiC, Si ₃ N ₄ , TiN, and Ti _x C _y N _z ), which are used to positively affect rheological properties and properties of layer resistance, layer thickness adjustment, and scratch resistance. A printable hybrid gel according to any one of claims 1 to 4, which is preferably printed in a structured manner by means of a screen printing process for the purpose of producing a solar cell, preferably a so-called PERC solar cell. a surface of the crucible or an electron passivated crucible surface, and then dried, also applied to the entire area by a thin layer of aluminum by means of a PVD process and then subjecting the known structure to a local laser melting process, wherein the aluminum is During the laser melting process, it is intentionally alloyed into an unprotected crucible (laser-fired contact cell). The printable hybrid gel of any one of claims 1 to 5, which is based on an oxide precursor of tin dioxide and aluminum oxide and is printed on the surface of the crucible and dried, and has electronic surface passivation The action also acts as a barrier to aluminum alloying and diffusion into the crucible below the layer. The printable hybrid gel of any one of claims 1 to 5, which is based on an oxide precursor of ceria, alumina and titania and is printed on the surface of the crucible and dried, and has both electrons Surface passivation also acts as a barrier to aluminum alloying and diffusion into the crucible below the layer. The printable hybrid gel of any one of claims 1 to 7, which improves and increases the internal back surface reflectance in a solar cell, preferably a so-called PERC solar cell, wherein the reflectance can be used to manufacture The concentration ratio of the main oxide precursor is specifically adjusted over a wide range. A printable hybrid gel according to any one of claims 1 to 7 which, after drying, forms an electrical insulation barrier between two electrical contacts on the surface of the tantalum wafer, or generally on the surface. Barrier layer. The printable hybrid gel of any one of claims 1 to 4, which can be printed and dried in the manufacture of microelectronics and microelectromechanical (MEMS) components, thin film solar cells, thin film solar modules, and organic solar energy. In the manufacture of batteries, in the manufacture of printed circuits and organic electronic devices, based on thin film transistors (TFTs), liquid crystals (LCDs), organic light-emitting diodes (OLEDs), and contact-sensitive capacitive and resistive sensors. The use of the display element of the technology is used as a scratch-resistant layer and an anti-corrosion layer and a reflection-reducing layer.