DE102010037321A1 - Formation of doped surface of plate-shaped wafer or semiconductor device involves applying phosphorus dopant source, forming primary doped surface, removing dopant source and crystallized precipitate, and forming secondary doped surface - Google Patents

Abstract

A phosphorus dopant source is applied on a surface, and a primary doped surface is formed. The dopant source is removed, and a secondary doped surface with high depth with respect to primary doped surface is formed. After selectively removing primary dopant source, crystallized precipitates of silicon-phosphorus and silicon-phosphorus oxide are removed.

Description

The invention relates to a method of forming a phosphor dopant dopant profile from a surface of a plate or wafer shaped semiconductor device by applying a phosphorus dopant source to the surface, forming a first dopant profile with dopant source present on the surface, removing the dopant source, and forming a dopant source Compared to the first dopant profile, a second depth profile having a greater depth.

In order to produce a highly doped surface region of a semiconductor component consisting of p-silicon in order to form an emitter in a solar cell and at the same time allow good contact, a phosphorus source can be applied in the prior art in order to drive phosphorus into the semiconductor from this. With a high dopant concentration of phosphorus, precipitates usually form in the silicon surface. However, the high dopant concentrations are required to produce sufficiently deep emitters (<0.3 μm deep) in processes suitable for mass production of silicon solar cells which simultaneously have sufficient surface phosphorus concentration that allow the use of Ag pastes for contact formation. Usually, the dopant is driven in a temperature (T) time (t) treatment step.

Advantageously, a method can be used in which the dopant is driven in two Tt treatment steps, optionally after the first Tt step, the diffusion source is removed ( WO-A-2010/066626 ).

In the course of the typical diffusion process, a phase of the composition Si _x P _y or Si _{x '} P _y O _z , which in the form of orientation-dependent needles (see 1 ) can crystallize in the highly doped near-surface Si layer. These precipitates have a very low conductivity, can lead to recombination of charge carriers, generate crystal lattice defects in the Si crystal and lead to mechanical stresses in the Si surface, which can adversely affect the electrical properties of the semiconductor device.

In order to be able to produce solar cells with lower surface concentrations for highly efficient concepts on thin wafers, it is important to use an in-line process for the diffusion step which does not produce precipitates in the surface of the wafers.

The phase of the composition Si _x P _y or Si _{x '} P _y O _z or the precipitates can not be removed by HF solutions, as they are usually used for removing the Phosphorsilicatglasschicht, after diffusion.

The present invention has the object of developing a method of the type mentioned so that an optimized dopant profile on a silicon semi-finished material can be designed in particular intended for the production of a solar cell.

According to the invention, the object is essentially achieved by removing the dopant source after formation of the first dopant profile and selectively removing precipitates crystallized on or in the surface from the phase Si _x P _y and Si _x P _y O _z .

Optionally, at the same time a portion of the excess phosphorus in the semiconductor device such as wafers can be removed in the region of the wafer that is close to the surface.

Near means up to a maximum of 10 nm. In this respect, this etching of the near-surface region comprises the invention.

According to the invention, a per se known two-stage diffusion process for producing a phosphorus-doped layer of silicon wafers, which are required for the production of silicon wafer-based solar cells, further developed such that after the first diffusion step from the phase Si _x P _y and Si _x P _y O _z (SIP compounds) crystallized precipitates are removed. In this case, it is provided in particular that the selective removal of the precipitates is carried out by using an alkaline etching solution containing an oxidizing agent, the alkaline etching solution over a time t ₃ with 15 sec ≤ t ₃ ≤ 5 min, in particular 30 sec ≤ t ₃ ≤ 60 sec, at a temperature T ₃ with RT (room temperature) ≤ T ₃ ≤ 60 ° C, preferably 35 ° C ≤ T ₃ ≤ 45 ° C, contacted.

Regardless of this, the invention provides that for the formation of the first dopant profile, the semiconductor component is exposed to a temperature T ₁ with T ₁ ≥ 750 ° C over a time t ₁ , while in the surface layer of the semiconductor device in a thickness range D with 120 nm ≥ D ≥ 80 nm, a phosphorus concentration K with K ≥ 10 ²⁰ phosphorus atoms / cm ³ is formed, and that the formation of the second dopant profile, the semiconductor device is exposed to a temperature T ₂ over a time t ₂ , wherein t ₂ > t ₁ . It is provided in particular that the temperature T ₂ ≥ T ₁ , in particular T ₂ in about T ₁ , wherein 750 ° C ≤ T ₁ , T ₂ ≤ 1000 ° C, in particular 800 ° C ≤ T ₁ , T ₂ ≤ 900 ° C is.

Furthermore, the first dopant profile during a time t ₁ should be 1 min ≦ t ₁ ≦ 10 min, in particular 2 min ≦ t ₁ ≦ 5 min, and the second dopant profile during a time t ₂ with 10 min ≦ t ₂ ≦ 120 min, in particular 20 min ≤ t ₂ ≤ 60 min, are formed.

In particular, it is also provided that after forming the first dopant profile and removing the dopant source, an etch-resistant mask is applied to the surface, which has a geometry which corresponds to a metallization to be applied to the surface.

The phosphorus diffusion can take place in different ways. It can, for. For example, POCl3 can be used as the P source and the process can run as a batch process. However, an inline method can also be used.

In these inline methods different P sources can be used, e.g. B. H3PO4 dissolved in solvent or in water, which can be applied with a spin-on, spray or mist system. After application of the phosphorus source, the silicon wafers are subjected to a T-t profile in a continuous furnace. The T-t profile is designed so that the phosphor partially diffuses into the silicon. Reaction with oxygen produces a phosphosilicate glass layer (PSG) on the surface.

In this first high-temperature step, a flat, ie not deeply driven, phosphor profile is generated. The depth of this profile is on the order of 100 nm 2 a corresponding first dopant profile in the form of a phosphor profile in silicon is shown after the first heat treatment, the substrate being exposed to a temperature of 870 ° C. over a period of 2 minutes.

In the case of this first diffusion step, at very high phosphorus concentrations, which are necessary in particular in the inline process, a phase of the composition Si _x P _y or Si _x P _y O _z arises at the interface of the silicon wafer with the PSG in the course of the diffusion in the form of can precipitate acicular precipitates. This is based on the 3 clear. These are taken from transmission electron micrographs of a cross section of a silicon sample, after the first temperature step. In the left picture the highly doped area with defects and precipitates is visible. In the right-hand illustration, the same sample and the same area are reproduced as dark-field recordings.

The acicular precipitates which crystallize out further in the course of the further diffusion are in the 1 clearly recognizable. This is an X-ray micrograph of a p-type monocrystalline silicon substrate after diffusion and removal of phosphosilicate glass.

The crystallized precipitates are orders of magnitude less conductive than regular diffused regions, can not be etched away with the usual PSG etch mixtures, and can cause additional perturbations and dislocations in underlying crystal zones. By different lattice constants, the SiP compounds generate dislocations and stresses in the Si lattice. These defects reduce the quality of semiconductor devices such as solar cells.

In the 4 Further transmission electron micrographs of a cross section of a silicon sample are reproduced, which illustrate defects according to the standard process. Right is according to the 3 reproduced the same sample and the same area in dark field recording.

After a long diffusion with a high P concentration on the surface of the wafers, these SiP compounds that have formed are clearly visible with the (X-ray electron microscope) SEM. They form rods on the surface with the dimensions: width 10-20 nm, length 100-200 nm). The size and frequency of these precipitates depends on the P concentration and the diffusion time.

After this first high-temperature step with a highly concentrated P source, the PSG is dissolved and the Si _x P _y or Si _x P _y O _z phase selectively removed, without substantially the normally diffused, about 100 deep emitter layer of the substrate Wafers is etched. The phosphorus concentration in this flat emitter can reach at most the value of the solubility of phosphorus in silicon, ie about 5 · 10 ²⁰ at / cm ^-3 ). In the course of further diffusion no precipitates can arise.

During the etching of the precipitates, if appropriate, a near-surface layer of the substrate, that is to say in particular the emitter layer, can be etched. This range usually amounts to a maximum of 20 nm. This etching of the near-surface region is thus to be subsumed under the teaching according to the invention.

It should also be pointed out that, in principle, the dopant source or the PSG is removed before the precipitates are etched. For this purpose, an HF solution is usually used.

The removal of the PSG and the SiP compounds and possibly excess phosphorus in the surface can be made only on certain areas of the emitter surface, ie selectively. For this purpose, after the first temperature step, an etch-resistant mask must be applied in the same form as the metallization grid. Thereafter, even with these wafers, the PSG and the SiP compounds and possibly a part of the highly doped layer is removed, but only in the intermediate areas in which no metal contacts are to be applied. Furthermore, it is possible to selectively diffuse these wafers with PSG on the surface with the laser and then perform only the etching step.

The phase of the composition Si _x P _y or Si _{x '} P _y O _z or the precipitates are wet-chemically removed after the first diffusion step, by immersion in an aqueous alkaline solution containing an oxidizing agent.

Preferably, a component from the group NaOH, KOH is used as the alkaline component for the aqueous alkaline solution. The oxidizing agent is preferably a component from the group of sodium hypochlorite, Natriamperoxidsulfate, hydrogen peroxide mentioned.

Applying a corresponding aqueous alkaline solution with an oxidizing agent over a contact time of 15 sec to 5 min, preferably 30 sec to 60 sec, at a temperature between 20 ° C and 60 ° C, in particular between 35 ° C and 45 ° C. , so the precipitates are selectively removed. This manifests itself through trenches, like these 5 which shows an X-ray electron micrograph of a silicon substrate, that of the 1 equivalent.

After selectively removing the precipitates without etching the highly doped surface layer, the semiconductor device is subjected to another T-t treatment step without refeeding a source of phosphorus. In this case, the time within which the second diffusion profile is formed is significantly longer than that for the formation of the first dopant profile.

During the second treatment step, the phosphor already diffused in the first process step diffuses deeper into the semiconductor substrate, so that the second phosphorus doping profile is formed, but without the interfering Si _x P _y compounds or Si _{x '} P _y O _z compounds and these crystallized precipitates occur.

This results from the 6 , which shows an X-ray electron micrograph of a silicon sample after optimized diffusion process. Precipitates are absent.

Depending on the time duration and the temperature, the second dopant profile may have a low surface concentration and a very large penetration depth.

this is the 7 can be seen, which shows a second dopant profile in the form of a phosphorus concentration after the second temperature step. The temperature used was 870 ° C. The amount of time that the temperature applied to the wafer was 60 minutes.

For the selective removal of the precipitates, an alkaline etching solution with at least one oxide from the group of peroxydisulfates, peroxomonosulfates, hypochlorite, hydrogen peroxide can be used, particular mention being made of sodium hypochlorite and sodium peroxodisulfate. The alkaline component of the alkaline etching solution used is at least one component from the group consisting of NaOH, KOH, ammonia, ammonia derivatives, alkylamines, alkanolamines, hydroxyalkyl-alkylamines, polyalkylamines, cyclic N-substituted amines.

Furthermore, the alkaline etching solution may contain at least one component from the group of complexing agents, surfactants, stabilizers. Complexing agents from the group of hydroxyphenols, amines, hydroxycarboxylic acids, polyalcohols, phosphoric acids and polyphosphates are particularly suitable as complexing agents. Citric acid or tartaric acid are used as hydroxycarboxylic acids, and polyhydric alcohols include glycerol, sorbitol and other sugars and sugar alcohols.

The contact time of the surface of the semiconductor device with the alkaline etching solution should be 15 seconds to 5 minutes, preferably 30 seconds to 60 seconds, at a temperature between room temperature and 60 ° C, preferably between 35 ° C and 45 ° C.

When using a corresponding etching solution and in particular taking into account the aforementioned parameters, the precipitates, which can be seen after the second temperature step without optimized parameters, are etched away, so that empty trenches result.

After the precipitates have been etched away, the semiconductor component, that is to say in particular the wafer without the addition of a dopant source, is subjected to a renewed Tt treatment. In this case, the time compared to the formation of the first dopant profile is to be chosen considerably larger. In particular, for the formation of the second dopant profile, a temperature between 750 ° C. and 1000 ° C. should act for a period of between 10 minutes and 120 minutes, in particular between 20 minutes and 60 minutes. In contrast, the time for forming the first dopant profile amounts to between 1 min and 10 min, preferably between 2 min and 5 min. The temperature is usually chosen in the same order of magnitude.

The diffused phosphor after the first T-t profile can then be driven very deep into the semiconductor substrate during the second T-t profile. The depth depends on the temperature and time (2nd T-t profile).

If the process is selective, standard P-profiles can be achieved in the area below the metallization, which can be screen-printed with any commercial paste. In the intermediate finger region, however, the emitter is not driven in so deeply, has a lower surface concentration and is low-defect in the surface, because here the precipitates are avoided and therefore no defects limit the solar cell performance.

If a good emitter has been produced, but which promotes the formation of precipitates, it is also possible to remove the precipitates onto the wafer after the diffusion step with the etching solutions indicated.

The attached figures show:

1 a photograph of a monocrystalline silicon substrate (p-type) after phosphorosilicate glass removal,

2 a representation of the measured with SIMS (secondary ion mass spectrometer) phosphor profile after the 1st temperature step (1st dopant profile),

3 a transmission electron microscope (TEM) recording of a cross section of a sample (monocrystalline wafer, mirror-polished surface) after the first temperature step,

4 a TEM image of a cross section of a sample (monocrystalline wafer, mirror-polished surface) after the 2nd temperature step,

5 an X-ray electron micrograph of the 1 silicon substrate to be removed after removing the precipitates in alkaline sodium hypochlorite solution at a temperature of 40 ° C and a contact time of 1 minute,

6 a SEM micrograph of the surface of a multicrystalline textured silicon wafer after formation of an optimized dopant profile without precipitates and

7 a representation of a measured with SIMS phosphor profile, which was prepared with the optimized process for the prevention of precipitates, after the 2nd temperature step (2nd dopant profile).

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• WO 2010/066626 A [0003]

Claims (13)

A method of forming a dopant profile with phosphorus dopant from a surface of a plate or wafer shaped semiconductor device by applying a phosphorus dopant source to the surface, forming a first dopant profile with dopant source present on the surface, removing the dopant source, and forming one compared to the first dopant profile a second depth dopant profile having a greater depth, characterized in that after forming the first dopant profile, the dopant source is removed and precipitates selectively precipitated on or in the surface from the phase Si _x P _y and Si _x P _y O _z are removed. A method according to claim 1, characterized in that after forming the first dopant profile and removing the dopant source, an etch-resistant mask is applied to the surface having a geometry corresponding to that of a metallization to be applied to the surface. A method according to claim 1 or 2, characterized in that for the formation of the first dopant profile, the semiconductor device is exposed to a temperature T ₁ with T ₁ ≥ 750 ° C for a time t ₁ , while in the surface layer of the semiconductor device in a thickness range D with 120 nm ≥ D ≥ 80 nm, a phosphorus concentration K with 10 ¹⁶ phosphorus atoms / cm ³ ≤ K ≤ 10 ²¹ phosphorus atoms / cm ³ , in the near-surface region to a thickness D _o with D _o equal to or approximately equal to 20 nm, a phosphorus concentration K _o with K _o ≥ 10 ²⁰ phosphorus atoms / cm ³ is formed, and that the formation of the second dopant profile, the semiconductor device is exposed to a temperature T ₂ over a time t ₂ , wherein t ₂ > t ₁ . Method according to at least one of the preceding claims, characterized in that the temperature T ₂ ≥ T ₁ , in particular T _{2 is} approximately T ₁ , where 750 ° C ≤ T ₁ , T ₂ ≤ 1000 ° C, in particular 800 ° C ≤ T ₁ , T ₂ ≤ 900 ° C. Method according to at least one of the preceding claims, characterized in that the first dopant profile during a time t ₁ with 1 min ≤ t ₁ ≤ 10 min, in particular 2 min ≤ t ₁ ≤ 5 min, and the second dopant profile during a time t ₂ with 10 min ≤ t ₂ ≤ 120 min, in particular 20 min ≤ t ₂ ≤ 60 min, is formed. Method according to at least one of the preceding claims, characterized in that the selective removal of the precipitates is carried out by using an alkaline etching solution containing an oxidizing agent, wherein the alkaline etching solution over a time t ₃ with 15 sec ≤ t ₃ ≤ 5 min, in particular 30 sec ≤ t ₃ ≤ 60 sec, at a temperature T ₃ with RT (room temperature) ≤ T ₃ ≤ 60 ° C, preferably 35 ° C ≤ T ₃ ≤ 45 ° C, contacted. Method according to at least one of the preceding claims, characterized in that the surface for removing the precipitates is contacted with an alkaline etching solution with at least one oxide from the group of peroxydisulfates, peroxomonosulfates, hypochlorite, hydrogen peroxide. Method according to at least one of the preceding claims, characterized in that the alkaline component of the alkaline etching solution at least one component from the group NaOH, KOH, ammonia, ammonia derivatives, alkylamines, alkanolamines, hydroxyalkyl-alkylamines, polyalkyleneamines, cyclic N-substituted amines is used. Method according to at least one of the preceding claims, characterized in that the alkaline etching solution further contains at least one component from the group of complexing agents, surfactants, stabilizers. Method according to at least one of the preceding claims, characterized in that a complexing agent from the group hydroxyphenols, amines, hydroxycarboxylic acids, polyalcohols, phosphoric acids, polyphosphates is used as the complexing agent. Method according to at least one of the preceding claims, characterized in that the alkaline etching solution is a dilute hypochlorite solution of the composition NaOH: 1 to 50 g / L, preferably 5 to 10 g / L, - Sodium hypochlorite: with 6 to 14% active chlorine, 150-750 mL / L, preferably 250 to 300 mL / L, and - as another possible component KOH is used. Method according to at least one of the preceding claims, characterized in that an alkaline etching solution with ammonium peroxide disulfate is used as the oxidizing agent with a composition: NaOH: 1 to 50 g / L, preferably 5 to 10 g / L, - ammonium peroxodisulfate 5 to 350 g / L , preferably 50 to 250 g / L, and - at least one component from the group KOH, ammonia or ammonia derivatives, amines as an alkaline component, other peroxodisulfate salts such. For example, sodium or potassium peroxodisulfates, Peroxomonosulfate such. B. potassium peroxomonosulfates. Method according to at least one of the preceding claims, characterized in that when removing the precipitates optionally a near-surface region is etched to a thickness d with d ≤ 10 nm.

Images