JP4337607B2 - Method for analyzing Ni concentration in alkaline solution - Google Patents

Description

  The present invention relates to a method for simply analyzing the concentration of Ni contained in an alkaline solution containing impurity Ni having a concentration of 25 ppb or less which is difficult to directly analyze.

The alkaline solution used for alkaline etching contains a plurality of types of impurity metals. For this reason, in the alkaline etching process performed in the polishing wafer (Polished Wafer) process, the wafer bulk contamination of the metal impurities proceeds. Therefore, it is necessary to analyze metal impurities contained in the alkaline solution before etching. Examples of the quantitative analysis method include an AAS method and an ICP-MS method. However, when an alkaline solution is directly analyzed using these methods, it is necessary to take a procedure of collecting a predetermined amount of the alkaline solution, concentrating them, and then applying various analysis methods. For this reason, analysis cannot be performed unless a large amount is collected, and there is a problem that the lower limit of detection of the obtained result is high, and it is not possible to cope with an alkali solution having a metal impurity concentration of 25 ppb or less.
As a measure to solve the above-mentioned problems, the time-dependent change of the hydrogen peroxide concentration in the alkaline cleaning solution for silicon wafers composed of ammonia, hydrogen peroxide and water is measured, and the contamination degree of the cleaning solution is determined by the reduction rate of the hydrogen peroxide concentration. Is disclosed (for example, see Patent Document 1).
Japanese Patent No. 2789287 (Claim 1)

However, in the method disclosed in Patent Document 1, the degree of contamination of the cleaning liquid with metallic impurities such as Fe and Cu can be measured, but Ni is not particularly mentioned, and ammonia, hydrogen peroxide, and water are used. Since the degree of contamination can be evaluated only in the case of an alkaline cleaning solution, there is a problem that it cannot cope with an alkaline solution such as an alkaline etching solution. In an actual alkaline etching solution composed of ammonia, hydrogen peroxide solution, and water, there are many elements, so it is difficult and impractical to predict metal contamination only by the reduction rate of hydrogen peroxide solution.
In addition, as a method of analyzing the concentration of metal impurities contained in an alkaline solution, there is a method such as resin separation. However, this method is difficult to perform low-level analysis because it easily causes contamination of the resin and the system. There was also a difficult problem.
An object of the present invention is to provide a method for simply analyzing the Ni concentration contained in an alkaline solution whose Ni concentration is unknown.

As shown in FIG. 1, the invention according to claim 1 includes: (a) a step of producing a P-type silicon wafer group to which boron is added so that the resistivity of the entire wafer is 5 to 100 mΩcm; and (b) A step of taking a reference sample from the P-type wafer group in step (a), a step of contacting a reference sample with a Ni-containing alkaline standard solution having a known Ni concentration, and (d) contacting the alkaline standard solution. A step of measuring the Ni concentration in the entire reference sample, and (e) the Ni concentration contained in the alkali standard solution of the step (c) and the measurement result of the step (d). A step of creating a correlation line between the Ni concentration contained in the reference sample and the Ni concentration in the alkaline standard solution; (f) a step of collecting a measurement sample from the P-type wafer group in the step (a); (g) Ni-containing alkaline solution with unknown Ni concentration A step of contacting the regular sample, (h) a step of measuring the Ni concentration in the entire measurement sample brought into contact with the alkaline solution, and (i) by collating the measurement result of the step (h) with a correlation line And a step of estimating the Ni concentration contained in the alkaline solution. The method for analyzing the Ni concentration in the alkaline solution is characterized by comprising:
The invention according to claim 1 has been made based on the following findings. The present inventors paid attention to the correlation between the Ni concentration contained in an alkaline solution such as KOH and NaOH and the Ni adsorption concentration on the P-type wafer after etching using this alkaline solution. That is, the Ni concentration after the sample is brought into contact with an alkaline standard solution with a known concentration is measured, and a calibration curve is created from this correlation. The Ni concentration after contact with an alkali solution of unknown concentration is measured with a measurement sample, and the measured value is applied to a calibration curve to obtain Ni contained in the alkali solution. By performing the above steps (a) to (i), it is possible to easily analyze even the Ni concentration contained in an alkaline solution that contains impurities Ni having a concentration of 25 ppb or less, which is difficult to analyze directly. . Further, unlike the direct analysis method, since the Ni concentration in the entire sample is indirectly measured, there is no problem of contamination during the Ni concentration measurement, and there is an advantage that analysis with a low detection limit is possible.

The invention according to claim 2 is the invention according to claim 1, wherein the Ni concentration contained in the alkali standard solution and the alkali solution is 25 ppb or less.

  The method for analyzing the Ni concentration in the alkaline solution of the present invention is included in the alkaline solution containing impurities Ni having a concentration of 25 ppb or less, which is difficult to directly analyze, through the steps (a) to (i). Even if it is Ni concentration to be able to be analyzed, it can analyze simply. In addition, unlike the direct analysis method, the Ni concentration in the entire sample is measured by the indirect dissolution method, so that there is no problem of contamination during Ni concentration measurement, and there is an advantage that analysis with a low detection limit is possible. Contact with alkali concentrates Ni in the liquid several orders of magnitude in the wafer, so there is no problem of contamination, and analysis with a low detection limit becomes possible.

Next, the best mode for carrying out the present invention will be described with reference to the drawings.
The method for analyzing the Ni concentration in the alkaline solution of the present invention comprises the following steps.
First, as shown in FIG. 1, a P-type silicon wafer group to which boron is added so that the resistivity of the entire wafer is 5 to 100 mΩcm is manufactured (step (a)). The resistivity of the entire wafer in the P-type silicon wafer group is defined within a range of 5 to 100 mΩcm. When the resistivity exceeds 100 mΩcm, even if an alkali standard solution or an alkali solution is brought into contact with the wafer in the subsequent process, This is because Ni contamination does not progress into the P-type wafer bulk. Even if the resistivity is less than 5 mΩcm, the effect is not changed. A preferable resistivity of the P-type silicon wafer group is 10 to 20 mΩcm.

Next, a reference sample is collected from the P-type silicon wafer group in the step (a) (step (b)). In order to obtain a correlation line that can be estimated in a subsequent process, it is preferable to collect at least three reference samples.
Next, the reference sample collected in the step (b) is brought into contact with a Ni-containing alkaline standard solution having a known Ni concentration (step (c)). The Ni-containing alkaline standard solution having a known Ni concentration is preferably 51% by weight for NaOH and 48% by weight for KOH. The Ni concentration contained in the alkali standard solution is preferably 25 ppb or less. If the Ni concentration exceeds 25 ppb, the Ni contamination concentration in the P-type wafer bulk becomes saturated, so that a correlation line that can be estimated in the subsequent process cannot be obtained, and the Ni concentration can also be measured by direct analysis. . Note that alkaline standard solutions having different Ni concentrations corresponding to the number of samples collected in the step (b) are prepared. In the step (c), it is preferable to maintain a Ni-containing alkaline standard solution having a known Ni concentration at 40 to 90 ° C., and contact the reference sample with the alkaline standard solution for 3 to 20 minutes. If the maintenance temperature of the alkaline standard solution is less than 40 ° C., Ni contamination in the P-type wafer bulk does not proceed sufficiently. The effect does not change even if the maintenance temperature exceeds 90 ° C. If the contact time is less than 3 minutes, Ni contamination into the P-type wafer bulk does not proceed sufficiently. The effect does not change even if the contact time exceeds 20 minutes. A preferable maintenance temperature of the alkali standard solution is 60 to 80 ° C., and a preferable contact time is 5 to 10 minutes. As a method for contacting the alkali standard solution with the sample, a method in which the alkali standard solution is stored in the liquid phase and the sample is immersed in the stored alkali standard solution is suitable.
The reference sample brought into contact with the alkaline standard solution in the step (c) is cleaned with HF solution to remove Ni present on the surface, and then the sample washed with the HF solution is washed with pure water.

Next, the Ni concentration in the entire reference sample brought into contact with the alkaline standard solution in step (c) is measured (step (d)). Vapor Phase Decomposition (VPD method), which is an indirect dissolution method, is used as a method for measuring the Ni concentration in the entire reference sample. Is preferable because it is possible.
As shown in FIGS. 2A and 2B, the reaction container 10 includes a storage container 12 that stores the decomposition solution 11 of the reference sample 15 and a lid 13 that seals the storage container 12. The container 12 and the lid 13 are 100 to 400 mm in length, 100 to 400 mm in width, 100 to 200 mm in height, 2 mm in thickness, and polytetrafluoroethylene (trade name: Teflon, hereinafter referred to as PTFE), which is a kind of fluororesin. A plastic box such as) is preferred. A support base 14 is disposed in the storage container 12. The support base 14 is made of PTFE and has a stand portion 14a and a table 14b. The stand portion 14 a is placed on the bottom surface of the container 12, protrudes above the liquid surface of the decomposition solution 11, and has a height that is about half the depth of the container 12. The table 14b is formed integrally with the stand portion 14a on the upper portion of the stand portion 14a, and the reference sample 15 is placed on the upper surface. A flange 14c protrudes from most of the peripheral edge of the table 14b. The container 12, the lid 13 and the support base 14 need to be sufficiently cleaned before the reference sample 15 is disassembled.

The reference sample decomposition solution 11 is stored in the container 12 so as to have a liquid level slightly below the table 14b. This decomposition solution 11 is a mixture of HF and HNO ₃ and H ₂ SO ₄ . Specifically, HF: HNO ₃ : H ₂ SO ₄ = 0.76: 0.7 in a weight concentration ratio of HF having a concentration of 38% by weight, HNO ₃ having a concentration of 68% by weight, and H ₂ SO ₄ having a concentration of 98% by weight. Prepared by mixing uniformly at a ratio of 1.96.
As shown in FIG. 2 (a), when the decomposition solution 11 is stored in the storage container 12, the reference sample 15 is horizontally placed on the upper surface of the table 14b, and the reaction vessel 10 is sealed with a cover, the decomposition solution 11 H ₂ SO ₄ therein absorbs moisture of HF and HNO ₃ in the decomposition solution, absorbs moisture in the sealed air in the reaction vessel 10, and lowers the humidity of the sealed space 16. This facilitates vaporization of the cracked liquid without heating the cracked liquid and without specially pressurizing the sealed reaction vessel, and the vaporized high-concentration HF-HNO ₃ gas 17 is on the table 14b. The sample 15 is contacted and decomposed and sublimated in a relatively short time as shown in FIG. 2B.

The decomposition reaction of the reference sample is performed as follows. First, the oxidation of Si with HNO ₃ gas or NO ₂ gas and the removal of SiO ₂ with HF gas are simultaneously performed as shown in equations (1) and (2).
_{Si + 4HNO 3 ↑ → SiO 2} + 4NO ↑ + 2H 2 O ...... (1)
SiO ₂ + 4HF ↑ → SiF ₄ ↑ + 2H ₂ O (2)
There is no unstable gas in the reaction vessel, and NO gas reacts with oxygen in the reaction vessel immediately after the reaction as shown in the formula (3).
2NO ↑ + O ₂ ↑ → 2NO ₂ ↑ (3)
After the water vapor generated in the formulas (1) and (2) adheres to the inner surface of the container to form minute droplets, the SiF ₄ gas reacts with the droplets as shown in the formula (4) to form a gel. Of orthosilicic acid (H ₄ SiO ₄ ).
SiF ₄ ↑ + 4H ₂ O → H ₄ SiO ₄ ↓ + 4HF ↑ (4)
The equation (3) and NO ₂ gas and HF gas generated respectively by the formula (4), the above reaction formula (1) and (2) are repeated, recycled reaction vessel of the NO ₂ gas and HF gas Reduce the pressure inside. As shown in the above formulas (1) and (3), the oxidation of Si by HNO ₃ produces NO ₂ , while very little NH ₃ gas is also produced. With HF and HNO _3, 97% or more of Si decomposes to produce SiF ₄ , while 3% or less forms diammonium hexafluorosilicate ((NH ₄ ) ₂ SiF ₆ ). This (NH ₄ ) ₂ SiF ₆ is a white crystal and remains as a residue 18.

Next, as shown in FIG. 2 (c) and FIG. 2 (d), the support 14 is taken out of the reaction vessel 10, and a mixed acid solution 19 of HF and HNO ₃ is dropped onto the residue 18 of this sample on the table 14b. The residue 18 is dissolved by this, and the solution 21 is collected in a PTFE beaker 22 from a portion without a flange. The mixing ratio of HF and HNO ₃ is HF with a concentration of 38% by weight: HNO _{3 with a} concentration of 68% by weight = 2: 1. The amount of the mixed acid solution 19 dropped is 1 ml per 1 g of residue. By heating the solution 21 in the beaker 22 at a temperature of 150 to 220 ° C., (NH ₄ ) ₂ SiF ₆ contained in the residue becomes hydrofluoric acid (H ₂ SiF ₆ ), silicon tetrafluoride (SiF ₄ ) and then sublimate and decompose in a relatively short time. Residues other than (NH ₄ ) ₂ SiF ₆ remain in the beaker. In addition, instead of the mixed acid of HF and HNO _3, a mixed acid of HCl and HNO ₃ may be dropped to dissolve the residue. In this case, the heating temperature of the solution is preferably 60 to 90 ° C.

ボロンは比較的短時間で分解昇華し、ビーカ２２内に不純物からなる残留物２４が残る。ボロン除去後の残留物２４にはＨ₂ＳＯ₄が含まれており、このＨ₂ＳＯ₄は後述する定量分析に影響（例えば、ＡＡＳ法のデータにブランクピークを発生させる。）を与えるので好ましくない。このためボロン除去後の残留物２４に熱処理を施す。この熱処理ではＨ₂ＳＯ₄は粘性が高いため２００〜２５０℃で５〜３０分間熱処理する必要がある。定量分析法にＡＡＳ法を用いる場合には、残留物中のＨ₂ＳＯ₄の残存量は１μｌ以下でよいが、ＩＣＰ-ＭＳ法を用いる場合には、残留物に含まれているほとんど全てのＨ₂ＳＯ₄を除去する必要がある。Ｈ₂ＳＯ₄除去後の残留物はＨＦ、ＨＮＯ₃の混酸で希釈され、この希釈液に含まれる微量不純物はＡＡＳ法、ＩＣＰ−ＭＳ法又はＩＣＰ−ＡＥＳ法で定量的に分析される。
なお、本実施の形態では分解液１１をＨＦとＨＮＯ₃の混酸にＨ₂ＳＯ₄を加えたものとしたが、ＨＦとＨＮＯ₃の混酸を用いてもよい。ＨＦとＨＮＯ₃の混酸を分解液１１とする場合では、分解液の気化を促進するために反応容器１０を加熱、或いは加圧する必要がある。 Subsequently, as shown in FIG. 2E, a mixed acid solution 23 of HF, HNO ₃ and H ₂ SO ₄ is dropped to dissolve the residue. The mixing ratio of HF, HNO ₃ and H ₂ SO ₄ is as follows: HF with a concentration of 38% by weight: HNO _{3 with a} concentration of 68% by weight: H ₂ SO _{4 with} a concentration of 98% by weight = 10-100: 10-100: 1 10. The drop amount of the mixed acid solution 23 is 100 to 200 μl per 1 g of the residue. By heating the solution in the beaker 22 at a temperature of 150 to 220 ° C., a reaction shown in the following formula (5) occurs, and the boron component contained in the residue is decomposed and sublimated.

Boron decomposes and sublimates in a relatively short time, and a residue 24 made of impurities remains in the beaker 22. The residue 24 after removing boron contains H ₂ SO ₄ , and this H ₂ SO ₄ is preferable because it affects the quantitative analysis described later (for example, a blank peak is generated in the data of the AAS method). Absent. For this reason, the residue 24 after the boron removal is subjected to heat treatment. In this heat treatment, H ₂ SO ₄ has a high viscosity, so it is necessary to perform heat treatment at 200 to 250 ° C. for 5 to 30 minutes. When the AAS method is used for the quantitative analysis method, the residual amount of H ₂ SO ₄ in the residue may be 1 μl or less. However, when the ICP-MS method is used, almost all of the residue contained in the residue is contained. It is necessary to remove H ₂ SO ₄ . The residue after removal of H ₂ SO ₄ is diluted with a mixed acid of HF and HNO ₃ , and trace impurities contained in this diluted solution are quantitatively analyzed by the AAS method, ICP-MS method or ICP-AES method.
Incidentally, in this embodiment it is assumed that the decomposition liquid 11 was added H ₂ SO ₄ in a mixed acid of HF and HNO _3, may be used mixed acid of HF and HNO _3. When the mixed acid of HF and HNO ₃ is used as the decomposition solution 11, the reaction vessel 10 needs to be heated or pressurized in order to promote vaporization of the decomposition solution.

  Next, referring back to FIG. 1, the Ni concentration contained in the reference sample brought into contact with the alkaline standard solution from the Ni concentration contained in the alkaline standard solution in the step (c) and the measurement result in the step (d). A correlation straight line is created between the Ni concentration in the alkaline standard solution (step (e)). It is inferred that the Ni concentration contained in the alkaline standard solution in the step (c) and the measurement result in the step (d) are correlated, and the Ni concentration in the step (c) and the measurement value in the step (d) Is plotted to obtain FIG. This FIG. 3 shows the preparation of 48 wt% KOH solutions containing Ni concentrations of 0.5 ppb, 1 ppb, 2.5 ppb, 10 ppb, 20 ppb and 100 ppb, respectively, and these KOH solutions are P-type silicon having a resistivity of 14 mΩcm. After contacting with the wafer, the Ni concentration contained in the P-type silicon wafer is measured, and the Ni concentration contained in the KOH solution and the Ni concentration contained in the P-type wafer after contacting with the KOH solution are plotted. FIG. It can be seen from FIG. 3 that both have a first-order correlation, and a calibration curve is created from a straight line based on this correlation.

Next, returning to FIG. 1, a measurement sample is collected from the P-type silicon wafer group in the step (a) (step (f)).
Next, the measurement sample collected in the step (f) is brought into contact with a Ni-containing alkaline solution with an unknown Ni concentration (step (g)). The concentration of the alkaline solution is preferably the same as the concentration of the alkaline standard solution in the step (c). The Ni concentration contained in the alkaline solution is preferably 25 ppb or less. If the Ni concentration exceeds 25 ppb, the Ni contamination concentration in the P-type wafer bulk will be saturated, so that the Ni concentration contained in the Ni-containing alkaline solution cannot be estimated from the correlation line in the subsequent process. Ni concentration can also be measured by direct analysis. In this step (g), it is preferable to maintain a Ni-containing alkaline solution with an unknown Ni concentration at 40 to 90 ° C., and bring the measurement sample into contact with the alkaline solution for 3 to 20 minutes. If the maintenance temperature of the alkaline solution is lower than 40 ° C., Ni contamination into the P-type wafer bulk does not proceed sufficiently. The effect does not change even if the maintenance temperature exceeds 90 ° C. If the contact time is less than 3 minutes, Ni contamination into the P-type wafer bulk does not proceed sufficiently. The effect does not change even if the contact time exceeds 20 minutes. A preferable maintenance temperature of the alkaline solution is 60 to 80 ° C., and a preferable contact time is 5 to 10 minutes.

Subsequently, the Ni concentration in the entire measurement sample brought into contact with the alkaline solution in the step (g) is measured (step (h)). This step (h) is preferably performed by an indirect dissolution method such as the VPD method of step (d) described above.
Furthermore, the Ni concentration contained in the alkaline solution is estimated by collating the measurement result of the step (h) with a correlation line (step (i)). By applying the measured value of the Ni concentration in the entire measurement sample obtained in the step (h) to the calibration curve shown in FIG. 3, the Ni concentration contained in the alkaline solution with the unknown Ni concentration is estimated.

  Thus, even if it is Ni density | concentration contained in the alkaline solution in which impurity Ni of the density | concentration of 25 ppb or less which is difficult to analyze directly is contained by passing through the said process (a)-process (i), it analyzes simply. be able to. In addition, unlike the direct analysis method, the Ni concentration in the entire sample is measured by the indirect dissolution method, so that there is no problem of contamination during Ni concentration measurement, and there is an advantage that analysis with a low detection limit is possible.

The figure which shows the method of analyzing Ni density | concentration in the alkaline solution of this invention. The block diagram which shows the method of measuring Ni density | concentration in the whole sample. The figure which plotted Ni density | concentration contained in a KOH solution, and Ni density | concentration contained in the wafer after making it contact with KOH solution.

Claims (2)

(a) producing a P-type silicon wafer group to which boron is added so that the resistivity of the entire wafer is 5 to 100 mΩcm;
(b) collecting a reference sample from the P-type wafer group in the step (a);
(c) contacting the reference sample with a Ni-containing alkaline standard solution having a known Ni concentration;
(d) measuring the Ni concentration in the entire reference sample in contact with the alkaline standard solution;
(e) From the Ni concentration contained in the alkali standard solution of the step (c) and the measurement result of the step (d), the Ni concentration contained in the reference sample brought into contact with the alkali standard solution and the alkali standard solution Creating a correlation line with the Ni concentration of
(f) collecting a measurement sample from the P-type wafer group in the step (a);
(g) contacting the measurement sample with a Ni-containing alkaline solution having an unknown Ni concentration;
(h) measuring the Ni concentration in the entire measurement sample brought into contact with the alkaline solution;
(i) analyzing the Ni concentration in the alkaline solution, comprising the step of estimating the Ni concentration contained in the alkaline solution by collating the measurement result of the step (h) with the correlation line. Method. The method of claim 1 Symbol placement Ni concentration in the alkaline standard solution and alkaline solution is less than 25 ppb.

Images