JP2007042950A - Quality evaluating method of epitaxial layer, quality evaluating method of soi (silicon on insulator) layer and manufacturing method of silicon wafer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To evaluate the quality of an epitaxial layer and an SOI layer with high reliability. <P>SOLUTION: In the silicon wafer having the epitaxial layer on a substrate, the quality of the epitaxial layer is evaluated by obtaining recombination life time. In the silicon wafer having an oxide film layer and the SOI layer on the substrate in this sequence, the quality of the SOI layer is evaluated by obtaining the recombination life time. The recombination life time of the epitaxial layer or the SOI layer is obtained by employing a first recombination life time measured after applying a first surface non-activation process with respect to the wafer, and a second recombination life time measured after applying a second surface non-activation process with respect to the wafer after measuring the first recombination life time. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for evaluating the quality of an epitaxial layer or SOI (Silicon On Insulator) layer of a silicon wafer by determining a recombination lifetime, and a method for manufacturing a silicon wafer using the method.

The recombination lifetime of silicon wafers is widely used for evaluating process contamination and metal contamination of bulk crystal parts. Further, in recent years, application to measurement of thin film portions such as n / n ⁺ and p / p ⁺ epitaxial layers and SOI layers has also been made.

As a method for measuring the recombination lifetime, a μ-PCD (μ-wave Photo Conductivity Decay) method for measuring the recombination lifetime by photoconductive decay using a reflected microwave is widely used (see Patent Document 1). ). In this method, excess carriers are generated in a wafer, and a decrease due to carrier recombination is measured and analyzed by reflected microwaves. When measuring the recombination lifetime of the epitaxial layer or the SOI layer by the μ-PCD method, the measurement is performed using a short wavelength laser beam that absorbs all the light in the layer.
Japanese Patent No. 3496058

However, when the recombination lifetime of the epitaxial layer is measured by the μ-PCD method, it is effective due to the recombination on the wafer surface and the excess carriers passing over the barrier with the high concentration substrate and flowing out to the substrate side. The influence of the recombination rate at the interface cannot be ignored, and it is difficult to accurately determine the intrinsic recombination lifetime of the epitaxial layer. The degree of excess carrier outflow is basically determined by the difference in resistivity between the epitaxial layer and the substrate, but it cannot be predicted from the resistivity uniformly due to interface stress, etc. It is necessary to obtain from the measured value.
Similarly, SOI wafers are also affected by recombination at the wafer surface and at the interface between the oxide film layer and the SOI layer, and it is difficult to accurately determine the recombination lifetime of the SOI layer itself.

  Under such circumstances, the present invention evaluates the quality of the epitaxial layer and the SOI layer with high reliability by obtaining the recombination lifetime of the epitaxial layer and the SOI layer separately from the recombination velocity at the surface and the interface. For the purpose.

  As a result of intensive studies to achieve the above object, the present inventor uses two types of recombination lifetimes measured after two types of surface deactivation treatment in the same wafer, It has been found that the recombination lifetime of the epitaxial layer and the SOI layer can be determined separately from the recombination velocity at the surface and interface, and the present invention has been completed.

That is, the means for achieving the above object is as follows.
[1] A method for evaluating the quality of an epitaxial layer in a silicon wafer having an epitaxial layer on a substrate,
The quality evaluation is performed by determining a recombination lifetime of the epitaxial layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface passivation treatment on the wafer;
Epitaxial layer quality obtained by using a second recombination lifetime measured after performing a second surface deactivation treatment on the wafer after the first recombination lifetime measurement Evaluation methods.
[2] The method according to [1], wherein the surface inactivation treatment is performed using a chemical passivation solution.
[3] The method according to [1], wherein the surface inactivation treatment is performed by irradiating a charge on the wafer surface.
[4] The silicon wafer has an oxide film or an insulating film on the surface of the epitaxial layer,
The surface deactivation treatment is performed by performing a corona discharge treatment on the oxide film or the insulating film.
[5] A silicon wafer manufacturing method including a step of evaluating the quality of the epitaxial layer on a silicon wafer having an epitaxial layer on a substrate and selecting a silicon wafer having a quality higher than a target,
The quality evaluation is performed by the method according to any one of [1] to [4].
[6] A method for evaluating the quality of an SOI layer in a silicon wafer having an oxide film layer and an SOI layer in this order on a substrate,
The quality assessment is performed by determining a recombination lifetime of the SOI layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface passivation treatment on the wafer;
SOI layer quality evaluation characterized by using a second recombination lifetime obtained after performing a second surface deactivation treatment on the wafer after the first recombination lifetime measurement Method.
[7] The method according to [6], wherein the surface inactivation treatment is performed using a chemical passivation solution.
[8] The method according to [6], wherein the surface inactivation treatment is performed by irradiating a charge on the wafer surface.
[9] The silicon wafer has an oxide film or an insulating film on the surface of the SOI layer,
The method according to [6], wherein the surface deactivation treatment is performed by performing a corona discharge treatment on the oxide film or the insulating film.
[10] A silicon wafer manufacturing method including a step of evaluating the quality of an SOI layer on a silicon wafer having an oxide film layer and an SOI layer on a substrate in this order, and selecting a silicon wafer having a quality higher than a target. And
The quality evaluation is performed by the method according to any one of [6] to [9].

  According to the present invention, the quality of the epitaxial layer and the SOI layer can be evaluated with high reliability without being affected by the recombination rate at the surface and interface.

Hereinafter, the present invention will be described in more detail.

[First embodiment (epitaxial layer quality evaluation method)]
The first aspect of the present invention is:
In a silicon wafer having an epitaxial layer on a substrate, a method for evaluating the quality of the epitaxial layer,
The quality evaluation is performed by determining a recombination lifetime of the epitaxial layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface passivation treatment on the wafer;
Epitaxial layer quality obtained by using a second recombination lifetime measured after performing a second surface deactivation treatment on the wafer after the first recombination lifetime measurement Evaluation method (hereinafter referred to as “Method I”)
It is.

In Method I, the quality of an epitaxial layer is evaluated in a silicon wafer having an epitaxial layer on a substrate. The wafer, it can be an epitaxial wafer of p / p ⁺ structure having a p-type epitaxial layer having a desired resistivity to the low resistance of the p ⁺ -type silicon substrate, a low-resistance n ⁺ -type silicon substrate It is also possible to use an n / n ⁺ structure epitaxial wafer in which an n-type epitaxial layer having a desired resistivity is formed. The epitaxial wafer can be manufactured by a known method, for example, a method of epitaxially growing a high-purity low-defect layer on a silicon substrate cut out from a silicon single crystal grown by the Czochralski method.

Examples of the low resistance p ⁺ type silicon substrate include a silicon substrate to which boron or the like is added, and its resistivity can be, for example, 0.001 to 0.1 Ωcm, preferably 0.001 to 0.01 Ωcm. Examples of the low resistance n ⁺ type silicon substrate include a silicon substrate added with phosphorus, arsenic, antimony, etc., and its resistivity is, for example, 0.001 to 0.1 Ωcm, preferably 0.001 to 0.01 Ωcm. Can be.

As a method for forming the epitaxial layer on the substrate, a commonly used method can be used. For example, a desired thickness can be obtained by a CVD method using a source gas containing Si such as SiCl ₄ , SiHCl ₃ , SiH ₂ Cl ₂ or the like. A method of performing epitaxial growth so that the layer is formed can be used. The thickness of the epitaxial layer can be, for example, 1 to 150 μm, preferably 10 to 150 μm. The resistivity of the epitaxial layer can be, for example, 1 to 100 Ωcm, preferably 5 to 100 Ωcm.

  When determining the recombination lifetime of an epitaxial layer formed on a low-resistance substrate (high-concentration substrate) as described above, excess carriers cross the barrier with the high-concentration substrate and flow out to the substrate side. Under the influence of the effective recombination rate at the interface and the recombination rate at the wafer surface, it is difficult to obtain an accurate recombination lifetime of the epitaxial layer itself. Therefore, in Method I, the recombination lifetime is measured by performing a first surface deactivation treatment on the wafer, and the first recombination lifetime and the first recombination lifetime. It calculates | requires using the 2nd recombination lifetime measured after performing the 2nd surface deactivation process with respect to the wafer surface after a measurement. Thus, by using two types of recombination lifetime measured after two types of surface deactivation treatments on the same wafer, the recombination lifetime of the epitaxial layer can be reduced at the surface and interface. It can be determined separately from the binding speed. Hereinafter, a method for obtaining the recombination lifetime of the epitaxial layer will be specifically described.

（式中、ｔ_epiはエピタキシャル層の厚さ、D_epiはエピタキシャル層における拡散定数、Sr_surfaceは表面再結合速度、Sr_interfaceは界面での実効再結合速度、τr_eff.は実効再結合ライフタイム、τr_epiはエピタキシャル層の再結合ライフタイム、βは測定系に応じて定まる定数である。） The following formulas (1) and (2) hold among the surface recombination velocity, the effective recombination velocity at the interface, the recombination lifetime, the thickness of the epitaxial layer, and the effective recombination lifetime in the epitaxial wafer. (See J. Appl. Phys. 79 (9), 1 May 1996, IEEE TRANSITIONS ON ELECTRON DEVICES, VOL. 50, No. 4, APRIL 2003).

(Where t _epi is the thickness of the epitaxial layer, D _epi is the diffusion constant in the epitaxial layer, Sr_surface is the surface recombination velocity, Sr_interface is the effective recombination velocity at the interface, τr_eff. Is the effective recombination lifetime, and τr_epi is (The recombination lifetime of the epitaxial layer, β is a constant determined according to the measurement system.)

First, for the wafer to be evaluated, the recombination lifetime (first effective recombination lifetime) of the epitaxial layer after the first surface deactivation treatment with a known surface recombination rate is performed. Next, for the same wafer, the recombination lifetime (second effective recombination lifetime) of the epitaxial layer after performing the second surface deactivation treatment with a known surface recombination rate is measured. When the measured first effective recombination lifetime is set to τr_eff_1 and the second effective recombination lifetime is set to τr_eff_2, the following formulas (3) to (6) are substituted into the above formulas (1) and (2). Is obtained. Equations (3) and (4) are equations relating to the wafer after the first surface deactivation treatment, and equations (5) and (6) are relating to the wafer after the second surface deactivation treatment. It is a formula.

  A τr_epi.−Sr_interface curve satisfying the above equations (3) and (4) and τr_epi.−Sr_interface satisfying the equations (5) and (6) are drawn, and the point where these two curves intersect is the τr_epi. Layer recombination lifetime) and Sr_interface (effective recombination velocity at the interface). Thus, according to the present invention, the recombination lifetime of the epitaxial layer can be reduced by using two types of recombination lifetime measured after two types of surface deactivation treatments are performed on the same wafer. It can be determined separately from the recombination velocity at the surface and interface.

The surface recombination rate of the wafer subjected to the first and second surface deactivation treatments can be determined by the following method, for example.
First, a polished wafer having the same type and resistivity as the epitaxial layer to be measured is prepared, and two types of regions having different thicknesses are formed. For example, a wafer having an acid-resistant seal attached to a part of the surface is immersed in an acid solution (for example, HF / HNO ₃ / CH ₃ COOH solution). Thereby, the area | region which is not masked with a seal | sticker is melt | dissolved with an acid, and the wafer which has two area | regions from which thickness differs is obtained.
Next, a surface inactivation treatment is performed on the wafer having two types of regions having different thicknesses, and the recombination lifetimes of the two regions having different thicknesses are measured. The recombination lifetime can be measured by the μ-PCD method.
In the above formulas (1) and (2), the wafer thickness is substituted into the epitaxial layer thickness (t _epi ), and the recombination lifetime is substituted into the effective recombination lifetime (τr_eff.). From the two curves obtained for the two regions, the surface recombination rate can be determined.

  The surface deactivation process is a process for reducing carrier recombination on the wafer surface by deactivating the wafer surface. The surface inactivation treatment can be performed by using, for example, a chemical passivation solution. Specifically, the surface inactivation treatment can be performed by immersing the wafer to be measured in a chemical passivation solution. As the chemical passivation solution, it is preferable to use a solution that provides good passivation characteristics (surface deactivation characteristics) with little change over time. Specifically, iodine ethanol liquid (for example, concentration 0.005-0.5N), HF aqueous solution (for example, concentration 0.1-20%), quinhydrone methanol liquid (for example, concentration 0.005-0.5 N), quinhydrone ethanol liquid (for example, concentration 0.005) ˜0.5 N) or the like can be used. The concentration of each solution can be appropriately set so as to obtain a desired surface recombination rate. For details of the chemical passivation solution, reference can be made to, for example, Appl. Surf. Sci. 63 (1993) 306, Japanese Patent No. 3496058, and the like.

  When the epitaxial wafer to be measured has an oxide film or an insulating film on the surface of the epitaxial layer, as the first and second surface deactivation processes, the oxide film or the insulating film is subjected to two different conditions. Corona discharge treatment can be performed. As a result, charges generated by corona discharge can be deposited on the surface, and the amount of charge can be controlled, or the polarity can be changed to control the bending of the band and effectively control the surface recombination rate. . The conditions of the corona discharge treatment are, for example, 2 to 7 kV or −2 to −7 kV, and can be 1 to 10 minutes.

  In addition, for wafers that do not have an oxide film or insulating film, by continuing to irradiate the wafer surface with charges during recombination lifetime measurement, the bending of the band is controlled in the same manner as described above, and surface recombination is effectively performed. The speed can be controlled. For details of the irradiation conditions, refer to, for example, US Pat. No. 6,653,850 B2.

  The recombination lifetime can be measured by the μ-PCD method. Specifically, the wavelength of the laser beam to be irradiated is set according to the thickness of the epitaxial layer so that all light is absorbed in the layer. The set wavelength can be about 300 to 900 nm, for example. When an optical pulse is generated at this set wavelength and irradiated to the wafer, the conductivity of the epitaxial layer increases due to excess carriers generated on the wafer surface and inside by the optical pulse, and the conductivity increases as the excess carriers disappear due to recombination. Decrease. The recombination lifetime can be obtained by detecting and analyzing the time change of the power of the reflected microwave at this time.

[Second embodiment (SOI layer quality evaluation method)]
The second aspect of the present invention is:
In a silicon wafer having an oxide film layer and an SOI layer in this order on a substrate, a method for evaluating the quality of the SOI layer,
The quality assessment is performed by determining a recombination lifetime of the SOI layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface deactivation treatment on the SOI layer surface;
An SOI layer obtained by using a second recombination lifetime obtained after performing a second surface deactivation treatment on the surface of the SOI layer after the first recombination lifetime measurement Quality evaluation method (hereinafter referred to as “Method II”)
It is.

  In Method II, the quality of the SOI layer is evaluated in a silicon wafer (SOI wafer) having an oxide film layer and an SOI layer in this order on the substrate. The wafer may be a bonded wafer in which two polished wafers are bonded together via an oxide film. A SIMOX (Separation by) is formed by injecting oxygen ions into the polished wafer to form an oxide film layer inside the wafer. Implanted Oxygen) wafers.

  The SOI wafer is obtained by thermally oxidizing a silicon substrate cut out from a silicon single crystal grown by a known method, for example, the Czochralski method, and stacking the silicon substrates cut out from the silicon single crystal in the same manner, and performing heat treatment. After bonding, the surface of one side can be manufactured to a desired thickness by thinning it by polishing or the like to form an SOI layer. Examples of the silicon substrate include a silicon substrate added with boron or the like in the case of p-type, and a silicon substrate added with phosphorus, arsenic, antimony, or the like in the case of n-type. The resistivity can be, for example, 0.1 to 1000 Ωcm. The thickness of the SOI layer can be, for example, 1 to 150 μm, preferably 10 to 150 μm. The resistivity of the SOI layer can be, for example, 0.1 to 1000 Ωcm, preferably 1 to 100 Ωcm.

  When determining the recombination lifetime of the SOI layer provided on the oxide film layer in the SOI wafer, the recombination on the wafer surface and the recombination at the interface between the oxide film layer and the SOI layer are performed as in the case of the epitaxial wafer described above. Due to speed effects, it is difficult to accurately determine the recombination lifetime of the SOI layer itself. Therefore, in the method II, the recombination lifetime of the SOI layer is determined by the first recombination lifetime measured after the first surface deactivation treatment is performed on the wafer, and the first recombination lifetime. It calculates | requires using the 2nd recombination lifetime measured after performing the 2nd surface deactivation process with respect to the wafer after a measurement. Thus, by using two types of recombination lifetime measured after two types of surface deactivation treatment on the same wafer, the recombination lifetime of the SOI layer can be reduced at the surface and interface. It can be determined separately from the binding rate. The details are as described for Method I above. The details of the surface deactivation treatment and recombination lifetime measurement are also as described for Method I above.

[Manufacturing method of silicon wafer]
The present invention further includes
For a silicon wafer having an epitaxial layer on a substrate, the quality of the epitaxial layer is evaluated, and a method for producing a silicon wafer including a step of selecting a silicon wafer having a quality higher than a target,
The quality evaluation is performed by Method I, a silicon wafer manufacturing method (hereinafter referred to as “Production Method I”), and
For a silicon wafer having an oxide film layer and an SOI layer on a substrate in this order, the quality of the SOI layer is evaluated, and a silicon wafer manufacturing method including a step of selecting a silicon wafer having a quality higher than a target,
The quality evaluation is performed by Method II, which is a method for manufacturing a silicon wafer (hereinafter referred to as “Process II”).
About.

  The manufacturing method I is a method for manufacturing an epitaxial wafer, and includes a step of performing quality evaluation by the method I and selecting a silicon wafer having a quality higher than a target. Production method II is a method for producing an SOI wafer, and includes a step of selecting a silicon wafer having a quality higher than a target by performing quality evaluation by method II. As described above, according to Method I and Method II, the recombination lifetime of the epitaxial layer and the SOI layer can be obtained with high reliability by separating from the recombination velocity at the surface and interface. Therefore, a quality silicon wafer can be obtained by performing quality evaluation by Method I and Method II and selecting a silicon wafer that has been confirmed to have a quality higher than the target. The target value of the recombination lifetime can be set in consideration of physical properties required for the wafer according to the use of the wafer.

Hereinafter, the present invention will be further described based on examples. However, this invention is not limited to the aspect shown in the Example.

(1) Determination of surface recombination rate An iodine ethanol solution (0.05 N) was used as the first chemical passivation solution, and a quinhydrone methanol solution (0.01 N) was used as the second chemical passivation solution. First, the surface recombination rate of each chemical passivation solution was determined by the following method.
A polished wafer having the same n-type and resistivity (7 Ωcm) as the epitaxial layer to be measured was prepared, and an acid-resistant seal was attached to a region of the wafer surface. The wafer was immersed in an HF: HNO ₃ : CH ₃ COOH solution (1:10:10) for 30 minutes to produce wafers having regions with different thicknesses. When the thickness of each region was measured, the wafer thickness of the region where the seal was applied (region A) was 735 μm, and the region where the seal was not applied (region B) was 721 μm. It was immersed in the two types of chemical passivation solutions described above, and the recombination lifetime of regions A and B was measured by the μ-PCD method. Prior to the measurement, cleaning at the product shipment level was performed to make the surface state the same. The measured recombination lifetime is shown below.
(i) Iodine ethanol solution A region: 3213.6 μs, B region: 3164.6 μs
(ii) Quinhydrone methanol solution A region: 5321.2 μs, B region: 5254.6 μs.

In the above formulas (1) and (2), the wafer thickness is substituted into the epitaxial layer thickness (t _epi ), and the recombination lifetime is substituted into the effective recombination lifetime (τr_eff.). When the surface recombination velocity was determined from the two curves obtained for the two regions, the iodine ethanol solution was 9.11 cm / s and the quinhydrone methanol solution was 4.52 cm / s.

(2) Determination of Recombination Lifetime of Epitaxial Layer An n / n ⁺ epitaxial wafer having an n-type 7 Ωcm epitaxial layer was used as a measurement target. The epitaxial wafer was immersed in iodine ethanol solution, and the recombination lifetime measured by the μ-PCD method was 220.6 μs, and that of quinhydrone methanol solution was 242.2 μs. Before the measurement, cleaning at the product shipment level was performed to make the surface condition the same. It was 114.3 micrometers when the thickness of the epitaxial layer was measured by FTIR method.
By substituting the measured recombination lifetime (effective recombination lifetime) and the surface recombination rate obtained in the above (1) into the above formulas (3) to (6), the above two types of surface inactivation Two τr_epi-Sr_surface curves corresponding to the treatment were drawn, and the effective recombination velocity at the n / n ⁺ interface was determined from the intersection of the two curves. The recombination lifetime of the epitaxial layer was determined to be 36.7 cm / s. However, it was 1806 μs. Thus, according to the present invention, the recombination lifetime of the epitaxial layer can be determined separately from the recombination velocity at the surface and interface.

(3) Determination of recombination lifetime of SOI layer An SOI wafer having a p-type 10 Ωcm, 12.4 μm thick SOI layer was used as a measurement target. This SOI wafer was immersed in iodine ethanol solution, and the recombination lifetime measured by μ-PCD method was 19.84 μs, and that of quinhydrone methanol solution was 21.42 μs. Before the measurement, cleaning at the product shipment level was performed to make the surface condition the same.
By substituting the measured recombination lifetime (effective recombination lifetime) and the surface recombination rate obtained in the above (1) into the above formulas (3) to (6), the above two types of surface inactivation When two τr_epi-Sr_surface curves corresponding to the treatment were drawn, the recombination velocity at the oxide layer interface was calculated from the intersection of the two curves, 51.43 cm / s, and the recombination lifetime of the SOI layer was determined. It was 626.7 μs. Thus, according to the present invention, the recombination lifetime of the SOI layer can be obtained separately from the recombination rate at the interface between the surface and the oxide film layer.

According to the present invention, the recombination lifetime of the epitaxial layer and the SOI layer can be obtained with high reliability.

Claims (10)

In a silicon wafer having an epitaxial layer on a substrate, a method for evaluating the quality of the epitaxial layer,
The quality evaluation is performed by determining a recombination lifetime of the epitaxial layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface passivation treatment on the wafer;
Epitaxial layer quality obtained by using a second recombination lifetime measured after performing a second surface deactivation treatment on the wafer after the first recombination lifetime measurement Evaluation methods. The method according to claim 1, wherein the surface deactivation treatment is performed using a chemical passivation solution. The method according to claim 1, wherein the surface deactivation treatment is performed by irradiating a charge on the wafer surface. The silicon wafer has an oxide film or an insulating film on the surface of the epitaxial layer,
The method according to claim 1, wherein the surface deactivation treatment is performed by performing a corona discharge treatment on the oxide film or the insulating film. For a silicon wafer having an epitaxial layer on a substrate, the quality of the epitaxial layer is evaluated, and a method for producing a silicon wafer including a step of selecting a silicon wafer having a quality higher than a target,
The method for producing a silicon wafer, wherein the quality is evaluated by the method according to claim 1. In a silicon wafer having an oxide film layer and an SOI layer in this order on a substrate, a method for evaluating the quality of the SOI layer,
The quality assessment is performed by determining a recombination lifetime of the SOI layer,
Said recombination lifetime,
A first recombination lifetime measured after performing a first surface passivation treatment on the wafer;
SOI layer quality evaluation characterized by using a second recombination lifetime obtained after performing a second surface deactivation treatment on the wafer after the first recombination lifetime measurement Method. The method according to claim 6, wherein the surface deactivation treatment is performed using a chemical passivation solution. The method according to claim 6, wherein the surface deactivation treatment is performed by irradiating a charge on the wafer surface. The silicon wafer has an oxide film or an insulating film on the SOI layer surface,
The method according to claim 6, wherein the surface inactivation treatment is performed by performing a corona discharge treatment on the oxide film or the insulating film. For a silicon wafer having an oxide film layer and an SOI layer on a substrate in this order, the quality of the SOI layer is evaluated, and a silicon wafer manufacturing method including a step of selecting a silicon wafer having a quality higher than a target,
The method for producing a silicon wafer, wherein the quality is evaluated by the method according to any one of claims 6 to 9.