DE102018109571B4 - Method for doping semiconductors - Google Patents

Abstract

A method for doping semiconductors comprising the following steps: a) cleaning a substrate made of silicon; b) b1) bringing the substrate into contact with a dispersion containing at least one dopant and b2) applying a layer containing the at least one dopant; c) doping the semiconductor with at least one dopant by heating, characterized in that - a mineral containing at least one dopant is used as the dispersed phase, - one or more substances are selected as the source for the dopant from the group containing or consisting of calcium apatite, struvite, monazite, monetite, Xenotime, Fluorapatite, Chlorapatite, Pyromorfite, Welite, Chernovite, Mimetite, Hedyphane and Hydroxylapatite, - heating to 800-1000 ° C takes place in step c), - no hydrofluoric acid is used to work up the doped substrate

Description

The present invention relates to a method for doping semiconductors, in which a mineral layer containing at least one dopant is applied to a substrate, as well as intermediate products of this method.

Semiconductors are preferably used in photovoltaic cells, solar cells or computers. Computers are getting smaller and more powerful. This is made possible by a manufacturing technology that applies billions of electronic components with great precision to semiconductor wafers (chips) that are just a few square millimeters in size.

The basis for the production of these chips are usually millimeter-thin platelets made of monocrystalline silicon, so-called wafers. By doping certain areas of the silicon substrate with foreign atoms, among other things. Phosphorus, semiconducting transistors and other electronic components are created.

Nowadays, a layer of phosphorus-containing organic molecules is often applied to silicon wafers for doping. In order to only dope selected areas, the organic layer is structured by etching. To protect against evaporation, an inorganic layer is also applied to the organic layer. Brief heating causes the phosphorus atoms to diffuse from the organic layer into the silicon and the inorganic layer can be removed again. However, this technique, known as “monolayer doping” (MLD), has the disadvantage that the number of phosphorus atoms available for doping is limited and the process uses chemicals that are toxic and harmful to the environment.

In the DE 10 2015 110 367 describes a process which, in a few steps, leads to a doping of wafers via the formation of a so-called calcium silicate hydrate phase (CSH phase).

Giraudo et. al., investigate CSH phases and their reaction to phosphorus in the journal Langmuir 2016 . The synthesis of a CSH phase is also disclosed here as an essential step.

In the DE 2952116 A a liquid dopant for silicon conductors is disclosed, which consists of orthoethyl silicate ester, a solvent and an acidic, aqueous phase. The presence of a high concentration of phosphorus in the form of phosphoric acid and equal proportions of water and silica in the dopant are essential for this dopant. The dopant is applied in a spin-on process. The coated semiconductor is then heated in a pure oxygen atmosphere.

In the WO 2014/101990 A describes a process for the production of doping media in which an anhydrous sol-gel-based synthesis is carried out by condensation of alkoxysilanes and / or alkoxyalkylsilanes with symmetrical and asymmetrical carboxylic acid anhydrides in the presence of phosphorus-containing compounds. Controlled gelation creates low-viscosity doping media. These doping media are used to process silicon wafers. It is essential for this process that the alkoxysilanes or alkoxyalkylsilanes form a three-dimensional network in the sol-gel reaction. As a result, a thin layer is formed when it dries and compacts, which is converted into a dense layer of glass by heat treatment.

The following documents can be cited as further prior art: DE 11 56 384 A ; DE 10 2007 023 103 A1 ; US 2014/0120705 A1 ; US 2014/0370640 A1 ; US 2018/0218801 A1 ; US 9,306,087 B2 .

A problem with the doping of semiconductors and wafers, in particular those based on silicon, is the natively growing silicon dioxide layer that forms in situ and / or the formation of glass on the substrate. Doping through such a silicon dioxide layer is extremely difficult and expensive. Therefore, according to the state of the art, such layers are removed using hydrofluoric acid (HF).

The removal of glass that forms during the process also requires the use of toxic and / or environmentally harmful chemicals such as HF. Toxic and / or environmentally harmful chemicals such as phosphorus oxyl chloride POCl ₃ , phosphorus tribromide PBr ₃ or ammonium fluoride NH ₄ F are also used in other process steps according to the prior art.

The so-called silicon / silicon oxide barrier on the surface of the semiconductors, which makes doping difficult, is also problematic. Usually this is also eliminated with HF. If HF is not used, extremely long production times and high temperatures as well as poor distribution of the dopant in the substrate have to be accepted.

The present invention is based on the object of eliminating the disadvantages of the prior art. In particular, a method is to be made available which leads to a high and homogeneous doping of semiconductors without the use of toxic, harmful and / or environmentally hazardous chemicals. The method should therefore be environmentally friendly and inexpensive. Furthermore, a simple method with a few method steps is to be made available, the individual steps without a complex device can be carried out. Furthermore, it should be possible to use the sources for the dopants in the method without complex preparation. In particular, it should be possible to use the method for the production of photovoltaic cells or solar cells.

Furthermore, there is also a need to improve the currently known inhomogeneous distribution of the dopant in the semiconductor.

This object is achieved by a method as defined in more detail by the claims and the description.

• a) Reinigen eines aus Silizium bestehenden Substrats;
• b)
  • b1) in Kontaktbringen des Substrats mit einer Dispersion (mindestens ein Dotierungsmittel enthaltend) und
  • b2) Aufbringen einer Schicht das mindestens ein Dotierungsmittel enthält;
• c) Dotieren des Halbleiters mit dem mindestens einem Dotierungsmittel durch Erhitzen.

One embodiment of the present invention relates to a method for doping semiconductors comprising or consisting of the following steps:

• a) cleaning a substrate made of silicon;
• b)
  • b1) bringing the substrate into contact with a dispersion (containing at least one dopant) and
  • b2) applying a layer which contains at least one dopant;
• c) doping the semiconductor with the at least one dopant by heating.

In an alternative, the substrate is a silicon wafer.

In one embodiment, the dopant is selected from the group containing or consisting of phosphorus, arsenic, boron, nitrogen, antimony, aluminum, indium and gallium, preferably phosphorus and / or arsenic.

In an alternative, the substrate is washed with deionized water in step a). In a further alternative, the washed substrate can be dried with nitrogen in an additional step a2).

An alternative relates to bringing the substrate into contact with a dispersion containing at least one dopant, the substrate being immersed in the dispersion of the dopant in step b1). The concentration of the dispersion of the source or of the doping agent is adjusted to approx. 0.1-10 mg / ml, preferably 0.5-5 mg / ml, in particular 1 mg / ml.

In one embodiment in step b2), the dopant can be applied to the substrate by removing the dispersion medium, for example by evaporating the dispersion medium by increasing the temperature and / or reducing the pressure, depending on the dopant selected.

The dispersion medium used is water, optionally with the addition of acid or base, or organic solvents such as alcohol, ether and / or ketones, preferably water, particularly preferably deionized water.

If necessary, the pH of the dispersion is adjusted by means of acid or base so that the source of the dopant disperses or goes into solution.

In the case of water, the dispersant is removed during heating to over 100 ° C., in particular to approx. 105 ° C. over a longer period of time.

According to the invention, a mineral containing at least one dopant is used as the dispersed phase.

In an alternative, the dispersed phase is the source of the dopant.

According to the invention, the source for the dopant is one or more substances selected from the group containing or consisting of calcium apatite, struvite, monazite, monetite, xenotime, fluorapatite, chlorapatite, pyromorfite, welite, chernovite, mimetite and hedyphane, preferably hydroxyapatite and / or struvite.

Based on the systematics of minerals according to Hugo Strunz, there can be a relatively large biodiversity within the mineral class of phosphates, to which arsenates also belong, due to the various diadochie possibilities. The most important structural unit, the tetrahedral anion complex P04 ³ "or AsO ₄ ^3- , can be found in every structure. The most important and most frequent representative of the phosphates is the mineral apatite Ca ₅ [(F, Cl, OH) / (PO)]. This mineral is a so-called "continuous mineral", ie it occurs in every type of rock. It can crystallize in the igneous early stage, it can be present biogenically in sediments (including guano) or as an accessory in metamorphic rocks the second anion fluoride [F], chloride [Cl] or hydroxide [OH] ions can occur, which can, however, represent each other.Depending on which anion predominates, fluorapatite is formed in the case of F and chlorapatite in the case of Cl and with OH to hydroxyapatite (HAp). The latter mineral, but not in its geogenically formed structure, is of particular importance in the human organism, as it is responsible for the structure of bones and teeth h is. In addition to the main representatives of phosphates, the previously mentioned apatites, there are a large number of other phosphates. Instead of Ca, Al ³⁺ or Fe ^{3+ can also} bond with [PO ₄ ] -. With which cation the Anion complex binds and finally crystallizes, depends not only on the supply but also on the environmental conditions such as the pH value. In an acidic pH range, aluminum phosphates (variscite) and iron phosphates (strictite) are mainly formed, while calcium phosphates, i.e. apatites, are formed under neutral and basic conditions. Another representative of this mineral class is struvite, an ammonium magnesium phosphate (NH ₄ ) Mg [PO ₄ ] * 6 H ₂ O.

The method according to the invention is carried out in one embodiment with a dispersion of the dopant in which this is present in molecularly disperse or colloidal disperse form, that is to say has no sedimentation during step b), in particular b1).

In one embodiment, the substrate is doped with at least two different dopants at the same time.

In an alternative, doping with two different dopants can be carried out in that the mineral used contains at least two different dopants. It is also possible to use minerals that contain 2 to 10 or more different dopants.

In a further alternative, different minerals are used, so that doping in at least 2-9 or 10 dopants can thereby take place.

Simultaneous doping with at least two different dopants within the meaning of the invention means that the respective method steps are run through exactly once, and at the end of the method there is a semiconductor doped with at least two different dopants.

A further embodiment relates to a method in which steps b1) and b2) are carried out in one step, for example by printing the dispersion of the dopant.

In step c) there is heating to 800-1000 ° C, preferably 850-950 ° C, particularly preferably 880-920 ° C, in particular 900 ° C, with a respective deviation of + - 10 ° C, preferably + - 5 ° C, especially + - 2 ° C.

As an alternative, the heating can also be carried out only locally on certain sections of the substrate, e.g. by using a laser.

In an alternative, a phase transformation of the outer native SiO ₂ layer of the substrate to a thermal oxide (for For definition see Ruhai Tian et.al. Langmuir Letter 2010, 4563-4566 , in particular 1 ) carried out. In one alternative, this oxide is more stable than the native SiO ₂ layer. The phase transformation takes place by increasing the temperature from 100 to 500 ° C.

In the planned publication there is a phase transformation into a more stable thermal oxide. This term could not be found in any lexicon.

In addition, in a step c2) the native SiO ₂ layer is opened by the metal ions of the source of the dopant and a so-called calcium silicate (CS) phase is created. In other words, the crystalline structure of the thermal SiO ₂ layer is destroyed by the metal ions from the doping source.

The substeps c1) and c2) take place simultaneously or one after the other, preferably simultaneously. In an alternative, in a further sub-step c3), the native SiO2 layer is opened by the metal of the source of the dopant, in which a silicate is formed with the respective metal, that is to say the calcium from the source of the dopant.

In one embodiment, in step c) the dopant is penetrated into the oxide-free region of the silicon substrate. This results in a controlled doping which leads to a change in the electrical properties and / or activity of the substrate. The penetration is carried out by increasing the temperature from 800 to 1000.degree. C., preferably from 850 to 950.degree. C., in particular 900.degree.

• d) Austreiben von organischen Verunreinigungen, bevorzugt durch Temperaturerhöhung von 100 auf 500 °C;

Another embodiment relates to the method with at least one of the following further steps:

• d) driving out organic impurities, preferably by increasing the temperature from 100 to 500 ° C .;

• e) Entfernen der CS-Phase mittels Salzsäure (HCl). In einer Alternative werden Reste der CS-Phase nach der Dotierung mittels HCl entfernt.

In an alternative, step d) is carried out simultaneously with step c), preferably with c1).

• e) Removal of the CS phase using hydrochloric acid (HCl). In an alternative, residues of the CS phase are removed by means of HCl after the doping.

• f) Transport des Dotierungsmittels durch die CS-Phase.

In contrast to the prior art, no HF is necessary for this, since there is no Si / SiO ₂ barrier. This was converted into a CS phase by penetration of the metal ions from the doping source.

• f) Transport of the dopant through the CS phase.

This transport is preferably carried out by increasing the temperature from 600 to 800.degree. C., particularly preferably from 650 to 750.degree. C., in particular to 700.degree.

It is characteristic of the process according to the invention that no calcium hydroxide Ca (OH) _{2 is added} during the entire process. In an alternative, no base is added during the entire process.

An alternative or additional characteristic of the present invention is the absence of a C-S-H phase during the entire process.

The process according to the invention is further characterized by the absence of silica, i.e. there was no addition of silica, nor is silica present in one of the process steps according to the invention.

In a further alternative, a glass is not formed in any of the method steps according to the invention or a glass is present. For the purposes of the invention, glass contains or consists of silicon dioxide, is an amorphous substance and is thermodynamically referred to as frozen, supercooled liquid. That is, a substance that is melted and cooled down quickly so that the solidifying glass solidifies too quickly to allow the building blocks to be arranged to form a crystal.

With the method according to the invention, a silicon substrate with a silicon layer containing metal ions as a doping source with a phosphor layer arranged above can be obtained as an intermediate product.

One advantage of the present method, which is also abbreviated to Mineral Interface Doping (MID), is first of all chemical-free doping of the substrate. In particular, the direct attachment of the dopant to the substrate can save work steps. Another advantage is the low price of the dopants.

Particularly advantageous for doping with two different dopants is the use of naturally occurring minerals in which several dopants are already present that diffuse into the substrate by means of the method according to the invention. Minerals as a doping source are neither poisonous nor corrosive, and therefore environmentally friendly. In the method according to the invention, no organic compounds are used; in the broadest sense, doping is chemical-free, since only minerals are used.

The position of each individual doping atom in the substrate is of great importance for high activity and trouble-free operation of the doped substrates. The method according to the invention results in a homogeneous, controlled and uniform doping of the substrate. This is through the methods of the prior art, such as using ink and printing paste (such as in the WO 2014/101990 A1 described) or by using pure phosphoric acid (such as in the DE 2952116 A described) is not guaranteed, in particular no controlled local dissolution of the doping in the nano range.

According to the method according to the invention, no hydrofluoric acid is used to work up the doped substrate. Since only a metal-silicate layer can arise, the metal being the counterion, i.e. the cation from the mineral used as a source for the dopant, the use of HCl is sufficient to remove this silicate layer if necessary.

Another advantage of the process according to the invention is that it can be carried out in an oxygen-containing atmosphere. In an alternative, however, step c) takes place in an oxygen-free atmosphere, preferably under protective gas, such as nitrogen (N2) or argon (Ar), preferably N2.

The invention is explained in more detail below with reference to examples.

Examples:

Doping the substrate.

A silicon wafer with a thickness of 500 μm and a size of 3 × 1 cm was used as the substrate. The wafer was washed with deionized water and dried _{under N 2.}

The wafer was then immersed in a dispersion of Ca ₅ (PO ₄ ) ₃ OH (HAp) or MgNH ₄ PO ₄ × 6H ₂ O (MAP). The minerals HAp and MAp were 99% pure. The dispersed phase was made up at a concentration of 1 mg / ml. The dispersion was heated to 105 ° C. so that the solvent water evaporated completely and both sides of the silicon wafer were completely covered by the dispersed phase or by the mineral particles.

The silicon wafer coated in this way was attached to a holder made of titanium and analyzed by means of FITR spectroscopy under a nitrogen atmosphere. The temperature was gradually increased from 60 ° C to 900 ° C.

FTIR

The FTIR spectra were recorded using a BRUKER-VERTEX V 70 spectrometer equipped with a deuterated triglycine sulfate detector. 1,024 scans between 7,500 and 400 cm ^-1 were recorded with a resolution of 4 cm ^-1 .

Electrochemical Impedance Spectroscopy

The coated wafer was examined before and after heating to 900 ° C. by means of impedance spectroscopy (impedance analyzer Hioki IM 3570). The measurement was carried out at room temperature in the frequency range between 5 Hz and 500 MHz, titanium clamping contacts with a diameter of 1 mm being used.

The results are in the - summarized.

, and show differential IR spectra of HAp deposited on a silicon wafer as a function of temperature.

At 100 ° C a broad negative band is ^{visible between 3,670 and 2,570 cm -1} and the release of CO2 can be observed at 2,364 and 2,336 cm ^-1 .

At 200 ° C a positive broad band is visible at 2,345 cm and another broad band at 3,700 cm ^-1 and 2,550 cm ^-1 . Both bands correspond to the stretching oscillation of OH groups. The sharp band of 2,345 cm ^-1 can be assigned to a metastable CO ₂ species.

The deformation vibration of water is in at .670 cm ^-1 to observe.

Also in oscillations of NH3 at 1455 cm ^-1 can be observed, which show a shift to 1426 cm ^-1 when heated to 400 ° C.

At 300 ° C, a negative band ^{can be observed at 3,572 cm -1} , which corresponds to the framework OH ions that are bound to Ca ions.

The sharp band at 2,375 cm ^-1 becomes negative at 400 ° C. and corresponds to the decomposition of metastable carbonate species.

Furthermore, a shift of the PO stretching vibration at 1,145 and 1,184 cm ^-1 to lower wavelengths can be observed ( .

After the loss of hydrate, the decomposition of carbonate is at 2,345 cm ^-1 in observed by a sharp band and by two weaker bands at 1,420 and 1,480 cm ^-1 ( 2 B) . The decarboxylation is accompanied by the loss of the Ca-O vibration at 3572 cm ^-1 , which is evidence of the reorganization of the hydroxyapatite structure.

shows IR spectra as a function of the temperature of the MAp film deposited on the silicon wafer as substrate. All organic components are released from 100 ° C. This is confirmed by a strong negative band of OH at 3,225 cm ^-1 , NH at 3,050 cm ^-1 , H ₂ O at 1,645 cm ^-1 and NH ₃ at 1,473 cm ^-1 .

A strong positive band at 3,685 cm ^-1 from a temperature of 400 ° C shows the protonation of the phosphate group. This is also confirmed by the POH bond band at 950 cm ^-1 .

At a temperature of 600 ° C the sample is free of NH3 and H ₂ O. As a result of the release of these compounds, the phosphate is transported through the native SiO ₂ layer, this phase being transformed into a thermal SiO phase.

also shows a differential IR spectrum of MAp deposited on a silicone substrate as a function of temperature between 600 and 700 ° C.

Figure 10 shows a differential IR spectrum of MAp deposited on a silicon substrate as a function of temperature between 800 ° C and 900 ° C.

shows the individual sub-steps of doping when the temperature increases.

List of reference symbols

1

adsorption

2

Wave number (cm ^-1 )

3

Organics; Mineral; native layer; oxide-free silicon

4th

Release of the organic matter and phase transition

5

Transport of phosphorus into the Si boundary layer

6th

Penetration of phosphorus into the oxide-free silicon

Claims (15)

A method for doping semiconductors comprising the following steps: a) cleaning a substrate made of silicon; b) b1) bringing the substrate into contact with a dispersion containing at least one dopant and b2) applying a layer containing the at least one dopant; c) Doping the semiconductor with at least one dopant by heating, characterized in that - a mineral containing at least one dopant is used as the dispersed phase, - one or more substances are selected as the source for the dopant from the group containing or consisting of calcium apatite, Struvite, Monazite, Monetite, Xenotime, Fluorapatite, Chlorapatite, Pyromorfite, Welite, Chernovite, Mimetite, Hedyphane and Hydroxylapatite, - in step c) heating to 800-1000 ° C, - no hydrofluoric acid is used to work up the doped substrate Procedure according to Claim 1 characterized in that the substrate is a silicon wafer. Method according to one of the preceding claims, characterized in that the dopant is selected from the group containing or consisting of phosphorus, arsenic, boron, nitrogen, antimony, aluminum, indium and gallium, preferably phosphorus and / or arsenic. Method according to one of the preceding claims, characterized in that in step b1) the substrate is immersed in the dispersion of the dopant. Method according to one of the preceding claims, characterized in that in step b2) the dopant is applied to the substrate by removing the dispersion medium. Method according to one of the preceding claims, characterized in that water is used as the dispersion medium. Method according to one of the preceding claims, characterized in that the dispersion of the dopant is molecularly disperse or colloidally disperse. Method according to one of the preceding claims, characterized in that the substrate is doped with at least two different dopants at the same time. Method according to one of the preceding claims, characterized in that steps b1) and b2) are carried out in one step. Method according to one of the preceding claims, characterized in that in a first sub-step of step c) c1) a phase transformation of the outer native SiO ₂ layer of the substrate to a thermal SiO ₂ oxide is carried out by increasing the temperature from 100 to 500 ° C; and c2) the structure of the thermal SiO ₂ layer is destroyed by metal ions from the source for the dopant. Procedure according to Claim 10 characterized in that the structure of the thermal SiO ₂ layer is destroyed by metal ions from the source for the dopant with the formation of a metal-silicate phase. Procedure according to Claim 11 characterized in that the metal-silicate phase is a calcium-silicate phase. Method according to one of the preceding claims, characterized in that in step c) a penetration of the dopant into the oxide-free region of the silicon substrate is carried out. Procedure according to Claim 11 with at least one of the following further steps: d) driving out organic impurities, preferably by increasing the temperature to 100-500 ° C; e) removing a metal-silicate phase by means of hydrochloric acid or removing the residues of a metal-silicate phase after doping by means of hydrochloric acid and / or f) transporting the dopant through a metal-silicate phase. Procedure according to Claim 3 characterized in that a silicon substrate with a silicon layer containing metal ions is obtained as an intermediate product from the source for the dopant with a phosphor layer arranged above it.

Images