JP2012132779A - Method for analyzing silicon sample - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide means for the identification and quantification of a dopant species (B or P) in a desired region of a silicon sample such as a silicon wafer.SOLUTION: An embodiment of the present invention relates to a method for analyzing a silicon sample comprising: decomposing part or the whole of a silicon sample; heating and evaporating a solution containing silicon obtained after the decomposition, provided that the heating and evaporating is stopped in a state where one droplet remains to recover the droplet and a dopant element selected from the group consisting of boron and phosphorus contained in the droplet is analyzed by an inductively coupled plasma mass spectrometer.

Description

  The present invention relates to a silicon sample analysis method, and more particularly to a silicon sample analysis method capable of qualitative analysis and quantitative analysis of dopant elements (boron, phosphorus) in a silicon sample.

  Boron (B) and phosphorus (P) are dopant elements that determine the conductivity type and resistance value in the silicon wafer. If the silicon wafer or the ingot for producing the wafer is contaminated by B or P during the manufacturing process or storage, the resistance value may deviate from a desired value, so that unintended contamination should be avoided. Therefore, accurately grasping the B and P contents in the silicon sample is extremely important for process control.

  Until now, SIMS (secondary ion mass spectrometry) is used as a physical method, SR (Spreading resistance) method, four-probe method, PL (photoluminescence method) as an electrical method. ) Has been used. Furthermore, as an electrical technique, Patent Document 1 proposes a method for obtaining a dopant amount from a difference between a resistance value of a bulk portion measured by an eddy current method and a resistance value of a surface layer measured by a surface photovoltage method.

  However, in the SIMS, the measurement region is limited to a partial range on the analysis principle that measurement is performed using a cut-out chip. Therefore, in SIMS, it is possible to analyze a certain region (surface layer portion) from the outermost surface of the silicon wafer to a predetermined depth, but it is impossible to determine the amount of dopant in the entire wafer. On the other hand, since the B and P concentrations are calculated from the resistance value (Irbin curve) in the electrical measurement, it is impossible to simultaneously specify the dopant type (qualitative analysis) and quantitative analysis.

JP 2006-269962 A

  Accordingly, an object of the present invention is to provide means for specifying and quantifying dopant species (B, P) in a desired region of a silicon sample such as a silicon wafer.

  In order to achieve the above object, the present inventors have paid attention to a chemical analysis method used for metal contamination analysis of a silicon wafer. The chemical analysis method is a technique for analyzing the metal contamination of a silicon wafer by decomposing the surface layer portion and bulk of the silicon wafer and qualitatively and quantitatively analyzing the metal component contained therein. In the above chemical analysis method, analysis of metal components contained in silicon is usually performed using an inductively coupled plasma mass spectrometer (ICP-MS). According to ICP-MS, it is possible to identify the contamination species and measure the contamination amount.

  Therefore, the present inventors tried to perform qualitative and quantitative analysis of B and P in a silicon sample by a chemical analysis method by ICP-MS, but the detection sensitivity was low and a useful evaluation result could not be obtained. . The present inventors have repeatedly studied this point and have obtained the following knowledge.

The chemical analysis method by ICP-MS is usually performed by the following two methods.
(1) By bringing an etching solution into contact with a silicon sample, part or all of the silicon sample is decomposed (liquid phase etching), and silicon at a measurement target site is taken into the etching solution. The etching solution containing silicon is heated and evaporated to completely dry, and the silicon is vaporized. Then, the target element remaining in the container such as a beaker is recovered by the recovered liquid, and the recovered liquid is subjected to analysis by ICP-MS ( For example, refer to JP 2001-99766 A).
(2) An etching gas is brought into contact with the silicon sample, whereby a part or all of the silicon sample is decomposed (vapor phase etching). After the gas phase etching, silicon residue is taken into the solution, and the solution is heated and evaporated to dryness to vaporize the silicon. Thereafter, the target element remaining in the container such as a beaker is collected by the collected liquid, and the collected liquid is subjected to analysis by ICP-MS (for example, Japanese Patent No. 3487334 and Japanese Patent Application Laid-Open No. 2004-335554).
In both the methods (1) and (2), the solution containing silicon (Si) is completely dried, but this is because a large amount of Si is contained in the solution in which the silicon sample is dissolved. If this solution is introduced into ICP-MS as it is, the measurement will be hindered by the interference of the Si matrix, and the peak intensity will be desensitized, making accurate analysis difficult, so Si must be vaporized and removed. Because.
However, as a result of the study by the present inventors, it has been found that completely evaporating the solution incorporating silicon as described above causes a decrease in the detection sensitivity of B and P. This is presumably because the vaporization of B and P occurs along with the vaporization of Si by complete drying.
As a result of further investigation based on the above knowledge, the present inventors have suppressed the large loss of B and P by stopping heating and evaporation in a state where one droplet remains without being completely dried. Thus, it has been newly found that the dopant element species in the silicon sample can be identified and quantified by the chemical analysis method by ICP-MS. This is because Si is easier to vaporize than B and P, and almost all of it is removed before it completely evaporates. On the other hand, B and P evaporate mainly when it completely evaporates (the last drop evaporates). It is assumed that it is caused by
The present invention has been completed based on the above findings.

That is, the above object has been achieved by the following means.
[1] Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample, comprising:
[2] Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
After the droplets are taken into the acid solution, the acid solution is heated and evaporated; however, the heating and evaporation is stopped in a state where one droplet remains,
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample, comprising:
[3] The method for analyzing a silicon sample according to [1] or [2], wherein the silicon sample is decomposed by liquid phase etching.
[4] The method for analyzing a silicon sample according to [3], wherein the etching solution used for the liquid phase etching is a mixed acid of hydrofluoric acid and nitric acid.
[5] The method for analyzing a silicon sample according to any one of [1] to [4], wherein the dopant element contained in the surface layer portion is analyzed by decomposing the surface layer portion of the silicon sample.
[6] The method for analyzing a silicon sample according to any one of [1] to [4], wherein the dopant element contained in the silicon sample is analyzed by performing bulk decomposition on the silicon sample.

  According to the present invention, it is possible to specify and quantify the dopant element species in a silicon sample such as a silicon wafer. By performing process management based on the obtained analysis results, it is possible to provide a high-quality silicon product in which unintended dopant contamination is reduced or suppressed.

The analysis result in Example 1 is shown. The analysis result in Example 2 is shown. The analysis result in Example 3 is shown. The analysis result in Example 4 is shown. The confirmation result of Si removal by heating evaporation is shown.

The present invention
Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample (hereinafter referred to as “method 1”),
Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
After the droplets are taken into the acid solution, the acid solution is heated and evaporated; however, the heating and evaporation is stopped in a state where one droplet remains,
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample (hereinafter referred to as “method 2”),
About. In both method 1 and method 2, when the solution containing the dopant element contained in the decomposed silicon is evaporated by heating, the heating and evaporation is stopped in a state where one droplet remains without being completely dried. Therefore, as described above, it is possible to prevent B and P from being vaporized and detecting sensitivity from being lowered when Si is removed. The method 2 includes a step of collecting the droplets remaining after the heat evaporation in an acid solution and evaporating again by heating. When removal of Si is not sufficient by only one heating evaporation, it is preferable to apply method 2 and perform heating evaporation again to increase the Si removal rate.
Hereinafter, the present invention will be described in more detail. However, the matters to be described are common to the methods 1 and 2 unless otherwise specified.

  Examples of silicon samples to be analyzed in the present invention include silicon wafers, single crystal silicon ingots, polysilicon as a raw material thereof, polysilicon solar cells, polysilicon ingots, and various types of silicon or silicon samples cut from them, silicon powder Etc.

  The decomposition of the silicon sample may be performed by liquid phase etching or gas phase etching. When the sample to be analyzed is a powder sample, it is preferable to apply liquid phase etching to the powder sample because the sample may be scattered during etching in the gas phase etching. In addition, since P is an element that easily vaporizes during vapor phase etching, loss may occur in vapor phase etching due to vaporization of phosphorus during etching. Therefore, in order to analyze P more accurately, it is preferable to apply liquid phase etching.

  In the liquid phase etching, various acid solutions that are usually used for liquid phase etching of silicon can be used as an etching solution for bringing the sample into contact with the sample to be analyzed and decomposing (dissolving) the sample. Specifically, an HF acid solution, for example, a mixed acid of hydrofluoric acid and at least one acid aqueous solution selected from the group consisting of nitric acid, sulfuric acid and hydrochloric acid can be used. The composition of the mixed acid (acid concentration and mixing ratio) may be appropriately changed according to the desired etching amount of the silicon sample. In general, the mixing ratio (hereinafter, the mixing ratio is based on volume) is nitric acid: hydrogen fluoride. The acid is preferably 3: 2 or 1: 2. The acid concentration is preferably about nitric acid: 60 to 70% by mass and hydrofluoric acid: about 30 to 40% by mass.

According to one embodiment of the present invention, a silicon sample to be analyzed is subjected to bulk decomposition, and bulk analysis can be performed by performing bulk decomposition. The bulk decomposition in the present invention is not limited to decomposing the entire amount of the silicon sample, but may be cleaving and decomposing a part of the silicon sample. For example, when it is determined that B and P are uniformly diffused in the silicon sample by heat treatment, the time required for the decomposition and pretreatment can be shortened by cleaving and decomposing a part thereof. Alternatively, when it is determined by other measurements that B and P are localized in a certain part, B and P can be analyzed with high sensitivity by cleaving and decomposing the part.
Another aspect of the present invention is to decompose a portion of a silicon sample. Part is, for example, a surface layer part of a silicon sample, and by analyzing the surface layer part and analyzing by ICP-MS through the steps described later, the type of dopant contained in the surface layer part is specified, The content can be determined. Thereby, the analysis of the dopant element of a surface layer part can be performed. Furthermore, the information regarding the distribution of the dopant element in the depth direction of the silicon sample to be analyzed can be obtained by combining the surface layer analysis and the bulk analysis.

  In order to decompose a part of the silicon sample with the etching solution, a method can be used in which the silicon sample is tilted and rotated so that the etching solution dropped on the surface of the silicon sample is adapted to the portion to be etched. Alternatively, a method of etching by placing a silicon sample on the surface of a Teflon (registered trademark) plate to which an etching solution is dropped can also be used. This method is called a sandwich method in the analysis of a silicon wafer. In the sandwich method, after a silicon wafer surface layer is etched by placing a silicon wafer on the dropped etching solution, the etching solution is recovered by shifting the silicon wafer. On the other hand, when bulk analysis is performed on a silicon sample, the silicon sample may be decomposed by being immersed in an etching solution.

  On the other hand, gas phase etching is a method of decomposing and sublimating a sample to be analyzed by bringing it into contact with an etching gas, and is suitable for bulk analysis. For example, if a sample to be analyzed is placed in a sealed container together with a solution that serves as an etching gas generation source, and the etching gas is generated from the etching gas generation source, the etching gas is brought into contact with the sample to be analyzed and sublimated in the sealed container. Can do. As the etching gas generation source, a mixed acid of hydrofluoric acid, nitric acid and sulfuric acid is suitable. The mixed acid is preferable because it can generate an etching gas without being heated or pressurized. In order to generate the etching gas in a short time, it is preferable to add silicon pieces to the mixed acid. The gas phase etching using the mixed acid is described in detail in Japanese Patent No. 3487334 and Japanese Patent Application Laid-Open No. 2004-33595, and in the present invention, gas phase etching can be performed based on the description.

  The etching solution obtained by decomposing the sample by contact with the silicon sample by the above method contains various elements (B, P, and other contaminating elements) contained in the decomposed silicon. Similarly, the silicon residue after the vapor phase etching contains various elements contained in the decomposed silicon. Therefore, by identifying the element contained in the etching solution after the liquid phase etching or the element contained in the silicon residue after the vapor phase etching, the element type contained in the decomposed silicon is known. It is possible to know the content or the amount of contamination by quantifying the above elements. However, the etching solution after the liquid phase etching and the solution into which the silicon after the gas phase etching are incorporated contain a large amount of Si together with various elements. If the solution is introduced into the ICP-MS as it is, the solution in the ICP-MS It will be affected by the Si matrix that hinders analysis. Therefore, in the present invention, in order to remove Si, in the liquid phase etching, the etching solution brought into contact with the silicon sample is heated and evaporated in the vapor phase etching. Thereby, Si can be vaporized and the influence of the Si matrix can be removed. However, as described above, in the present invention, the heating evaporation is stopped in a state where one droplet remains. As a result, it is possible to prevent the loss of B and P from occurring due to the vaporization of B and P when the last drop evaporates, thereby reducing the detection sensitivity. In addition, as a solution used in order to take in the silicon residue after a gaseous-phase etching, the acid solution illustrated as an etching solution for liquid phase etching can be used, for example. In order to decompose and remove the silicon residue, a mixed solution of hydrofluoric acid, hydrochloric acid and nitric acid is preferable, and the mixing ratio is preferably 1: 4: 2. For example, when a silicon sample is bulk-decomposed by vapor phase etching, the silicon residue is taken into the acid solution by adding the acid solution into the beaker in which the silicon sample to be analyzed is placed. Can do. In addition, when a part of the silicon sample is decomposed by gas phase etching, the acid solution is scanned on the surface of the silicon sample after decomposition, or the acid solution is brought into contact with the portion subjected to the gas phase etching by the sandwich method described above. The decomposition residue can be taken into the acid solution.

  Heat evaporation for removing Si is preferably performed using a container whose contact surface with an etching solution such as a Teflon (registered trademark) beaker is a hydrophobic surface. Specifically, a vessel containing a part or all of an etching solution after contact with a silicon sample in liquid phase etching or an acid solution containing a silicon residue after vapor phase etching is placed on a hot plate, for example, 150 to By heating to a temperature of about 250 ° C., the solution can be evaporated. During heating and evaporation, visually check the inside of the container and remove the container from the hot plate before the liquid droplets are completely dried to stop heating and evaporation with one liquid droplet remaining. Can do. The volume of the remaining droplets is preferably 50 μl or less from the viewpoint of increasing the Si removal rate, more preferably 30 μm or less, and even more preferably 20 μl or less, and the loss of B and P is reduced. From the viewpoint of the above, it is preferably 10 μl or more.

  In some cases, such as when the amount of silicon in the solution subjected to heat evaporation is large, removal of Si may be insufficient with one heat evaporation. In such a case, it is preferable to apply method 2 and perform heat evaporation (Si removal) again. For example, by adding an acid solution to the container after the heat evaporation, the acid solution in which one droplet remaining after the heat evaporation of the solution is taken in is heated and evaporated. However, also in this case, in order to prevent the loss of B and P, the heating evaporation stops with one droplet remaining. As the acid solution, a mixed solution of hydrofluoric acid, hydrochloric acid and nitric acid effective for decomposing and removing Si is preferable, and the mixing ratio is preferably 1: 4: 2. In order to further increase the Si removal rate, the same operation can be performed twice, three times, or four times or more.

  Finally, droplets remaining after the evaporation of the solution by heating are collected, and dopant elements (B and / or P) contained in the droplets are analyzed by ICP-MS. Since ICP-MS can identify and quantify the element species contained in the sample, the type and amount of the dopant element contained in the droplet, that is, the dopant element contained in the decomposed silicon is determined. Can be sought. For example, when a larger amount of B than the intended amount is detected in a boron-doped p-type silicon wafer, it can be determined that B contamination has occurred, and an amount of P that is expected to undergo pn conversion is detected. It can be determined that P contamination has occurred. On the contrary, if a larger amount of P is detected than is intended in the phosphorus-doped n-type silicon wafer, it can be determined that P contamination has occurred, and an amount of B expected to cause pn conversion. Is detected, it can be determined that B contamination has occurred.

  The droplets can be recovered by adding a recovery liquid into a container in which one droplet after the heat evaporation remains. The recovered liquid includes pure water, a mixed solution of hydrofluoric acid and hydrogen peroxide; pure water, a mixed solution of hydrogen peroxide and hydrochloric acid; pure water, hydrofluoric acid, hydrogen peroxide and hydrochloric acid A mixed solution of, etc. can be used. By introducing the recovered liquid into ICP-MS, the dopant in the droplet can be analyzed. The droplets can be analyzed without being adversely affected by interference or desensitization by the Si matrix because Si has been removed by the heat evaporation. Further, since heating and evaporation is stopped in a state where one droplet remains, the dopant element can be analyzed without causing a decrease in sensitivity due to loss (vaporization) of B and P.

ICP-MS is roughly classified into a quadrupole ICP-MS and a double convergence ICP-MS. The quadrupole ICP-MS performs mass separation by energy convergence by an electric field, is easily affected by molecular ion interference, and sensitivity is reduced when Si is introduced.
In the measurement of P, molecular ions such as NO, NOH, and SiH are present in the vicinity of the mass number 31, and there is a concern that the quadrupole ICP-MS is affected by these molecular ion interferences.
On the other hand, since the double-focusing ICP-MS can perform mass separation by a magnetic field and energy convergence by an electric field, analysis can be performed without being affected by the above-described molecular ion interference. Moreover, if it is about 2000 ppm Si, it can analyze without desensitization. Therefore, in order to further reduce the decrease in sensitivity due to the Si matrix, it is preferable to use a double convergence ICP-MS. Further, since the double convergence ICP-MS has a higher resolution than the quadrupole ICP-MS, it is preferable to use the double convergence ICP-MS from the viewpoint of resolution.

  In addition to the dopant element, the droplets also contain a contaminating element contained in silicon dissolved by contact with the etching solution. According to ICP-MS, together with B and P, Li, Na, Mg, Al, K, Ca, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Ge, As, Rb, Mo, Since analysis of Ag, Sb, W, and Pb is also possible, the qualitative / quantitative analysis of the above element group is performed together with the qualitative / quantitative analysis of the dopant element, thereby identifying the contamination species of the silicon sample to be analyzed and Measurements can also be made.

The silicon sample analysis method of the present invention is used, for example, as a wafer evaluation method for grasping process contamination such as wafer surface contamination and bulk contamination in the manufacturing process such as epitaxial layer contamination and heat treatment process in the epi process step. Can do. Furthermore, the present invention can also be applied to the dopant contamination management of silicon wafers for high resistance devices that have been attracting attention in recent years to reduce power consumption.
In addition, the present invention is not limited to application to silicon wafers, and can also be used for analysis of various types of silicon such as single crystal ingots, polysilicon ingots, solar battery ingots, and solar battery cells.

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

[Example 1]
Three types of B-doped p-type silicon wafers (φ8 inches) having different resistivity were prepared, and the boron content of the surface layer portion was determined for each wafer by the following method.
(1) 1.8 ml of a mixed acid aqueous solution (etching solution) in which 38% by mass of hydrofluoric acid and 68% by mass of nitric acid were mixed at a volume ratio of 3: 2 was dropped onto a clean Teflon (registered trademark) plate.
(2) The etching solution was sandwiched between a plate and an 8-inch silicon wafer so that the wafer and the etching solution were brought into contact with each other to decompose (etch) the surface layer portion having a depth of 1 μm from the wafer surface. NOx gas is generated during the etching.
(3) After etching (after completion of NOx gas generation), the wafer is removed from the plate with Teflon tweezers, and the etching solution remaining on the plate (hereinafter referred to as “sample solution”) is placed in a Teflon (registered trademark) beaker. 1 ml was added and the whole beaker was heated on a hot plate (about 150 to 250 ° C.). At that time, the sample solution was not completely dried and the beaker was removed from the hot plate while only one drop (20 μl) remained on the bottom surface of the beaker, and the evaporation by heating was stopped.
(4) Add 300 μl of a 38% by mass hydrofluoric acid / 68% by mass nitric acid / 20% by mass hydrochloric acid mixed acid aqueous solution (mixing ratio 1: 2: 4) to the beaker and heat on a hot plate again (150 to About 250 ° C.). Again, the solution was not completely dried and the evaporation was stopped by removing the beaker from the hot plate while only one drop (20 μl) remained on the bottom of the beaker.
(5) 1 ml of a mixed acid aqueous solution of 2% by mass hydrofluoric acid / 2% by mass hydrogen peroxide is added to the above beaker, and 1 drop (20 μl) remaining in the beaker is dissolved. Qualitative and quantitative analysis was performed.

  The resistance value of the wafer after removing the sample solution in (3) above was determined by the four-probe method, and the B concentration was calculated from the resistance value. The measurement is performed at 49 points on the line up to both ends of the wafer passing through the center of the wafer, and the average value is set as the resistance value (in the measurement by the four-probe method, the surface of the contact portion of the electrode is contaminated. Measurements were made after removal of the part). FIG. 1 shows the relationship between the B quantitative result by ICP-MS and the B concentration calculated from the resistance value by the four-probe method.

[Example 2]
Three types of P-doped n-type silicon wafers (φ8 inches) with different resistivity were prepared, and each wafer was analyzed by ICP-MS in the same manner as in Example 1, measured by a four-point probe method, and P concentration. Was calculated. FIG. 2 shows the relationship between the P quantitative result by ICP-MS and the P concentration calculated from the resistance value by the four-probe method.

Evaluation Results As shown in FIGS. 1 and 2, the B and P concentrations determined from the ICP-MS B and P quantitative values and the resistance values obtained by the four-probe method were almost the same (error less than 10%). Since the silicon wafer subjected to the above analysis was manufactured and stored in a high clean atmosphere without concern about contamination by B and P, it can be considered that there is no B contamination and P contamination. This sample can be regarded as having no bias in the dopant amount in the depth direction. Therefore, the fact that the dopant concentration determined from the resistance value by the four-probe method and the quantitative value by ICP-MS match, enables quantitative analysis of B and P with good detection sensitivity by the analysis method according to the present invention. It means that. In addition, qualitative analysis is impossible with an electrical method such as the four-probe method, but as described above, qualitative analysis of B and P can be performed according to ICP-MS.
From the above results, it was shown that according to the present invention, qualitative analysis and quantitative analysis of the dopant element in the silicon sample can be performed.

[Example 3]
For three types of B-doped p-type silicon wafers having different resistivity, the B concentration was determined from the resistance values measured by the four-probe method in the same manner as in Example 1. Thereafter, the wafer was cleaved to 0.1 g to produce a wafer piece, and the surface was washed with a hydrofluoric acid / hydrochloric acid / hydrogen peroxide mixed solution. About the wafer piece after washing | cleaning, boron content was calculated | required with the following method.
(1) Using 2 ml of a mixed acid aqueous solution in which 38% by mass of hydrofluoric acid and 68% by mass of nitric acid were mixed at a volume ratio of 2: 1, the wafer piece was dissolved in a Teflon (registered trademark) beaker.
(2) The solution after dissolution (hereinafter referred to as “sample solution”) was heated on a hot plate together with the beaker (about 150 to 250 ° C.). At that time, the sample solution was not completely dried and the beaker was removed from the hot plate while only one drop (20 μl) remained on the bottom surface of the beaker, and the evaporation by heating was stopped.
(3) Add 300 μl of 38% by mass hydrofluoric acid / 68% by mass nitric acid / 20% by mass hydrochloric acid mixed acid aqueous solution (mixing ratio 1: 2: 4) to the beaker and heat again on a hot plate (150 to About 250 ° C.). Again, the solution was not completely dried and the evaporation was stopped by removing the beaker from the hot plate while only one drop (20 μl) remained on the bottom of the beaker.
(4) 1 ml of a mixed acid aqueous solution of 2% by mass hydrofluoric acid / 2% by mass hydrogen peroxide is added to the above beaker, and 1 drop (20 μl) remaining in the beaker is dissolved. When qualitative / quantitative analysis was performed, the quantitative result of B was as shown in FIG.

[Example 4]
Three types of P-doped n-type silicon wafers (φ8 inches) having different resistivity were prepared, and each wafer was analyzed by ICP-MS in the same manner as in Example 3, measurement of resistance value by four-probe method, and P concentration. Was calculated. FIG. 4 shows the relationship between the P quantitative result by ICP-MS and the P concentration calculated from the resistance value by the four probe method.

Evaluation Results As shown in FIGS. 3 and 4, the B and P concentrations obtained from the ICP-MS B and P quantitative values and the resistance values obtained by the four-probe method were almost the same (error less than 10%). Since the silicon wafer subjected to the above analysis was manufactured and stored in a high clean atmosphere without concern about contamination by B and P, it can be considered that there is no B contamination and P contamination. Therefore, the fact that the dopant concentration determined from the resistance value by the four-probe method and the quantitative value by ICP-MS match, enables quantitative analysis of B and P with good detection sensitivity by the analysis method according to the present invention. It means that. In addition, qualitative analysis is impossible with an electrical method such as the four-probe method, but as described above, qualitative analysis of B and P can be performed according to ICP-MS.
From the above results, it was shown that according to the present invention, qualitative analysis and quantitative analysis of the dopant element in the silicon sample can be performed.

Confirmation of B vaporization by complete drying Four sample solutions were prepared by dropping a B standard sample into a mixed acid aqueous solution having the same composition as the etching solution used in Examples 1 and 2 to 10 ppb. (Registered trademark) Placed in a beaker and heated on a hot plate together with the beaker. The two beakers were heated and evaporated until the droplets disappeared and the solution was completely dry. The remaining two beakers were removed from the hot plate with only one droplet (20 μl) remaining on the bottom, and heated and evaporated. Was stopped.
A sample solution for ICP-MS analysis was prepared by adding 1 ml of a mixed acid aqueous solution of 2% by mass hydrofluoric acid / 2% by mass hydrogen peroxide to each beaker. The prepared sample solution was subjected to the same ICP-MS analysis as in the example. The B quantitative results are shown in Table 1 below.

  As shown in Table 1, when heating and evaporation was stopped with one drop remaining, B could be concentrated and recovered in the remaining one drop. On the other hand, from the results shown in Table 1, it can be confirmed that when completely dried, B is vaporized and hardly remains in the residue. Therefore, even if the residue is recovered after complete drying, B cannot be detected.

Confirmation of Si Removal by Heating Evaporation In Examples 1 and 2 in which an 8-inch wafer was etched by 1 μm, the Si concentration of the sample solution was 42 mg / ml.
In addition, the upper limit of the Si concentration that can be introduced into the double convergence ICP-MS used in the above embodiment is about 2000 ppm, and the lower limit of Si detection of this apparatus is about 10 ppb.
Therefore, two types of sample solutions were prepared by dissolving Si in a mixed acid aqueous solution having the same composition as the etching solution used in Examples 1 and 2 so as to be 2000 ppm and 42 mg / ml. Prepare 3 samples at each concentration, put each in a Teflon (registered trademark) beaker and heat on the hot plate with the beaker, remove the beaker from the hot plate and heat with only one drop (20 μl) remaining on the bottom. Evaporation was stopped. A sample solution for ICP-MS analysis was prepared by adding 1 ml of a mixed acid aqueous solution of 2% by mass hydrofluoric acid / 2% by mass hydrogen peroxide to each beaker. The prepared sample solution was subjected to the same ICP-MS analysis as in the example. The results of Si quantification are shown in FIG.
From the results shown in FIG. 5, both the sample solution with Si concentration of 2000 ppm and the sample solution with 42 mg / ml showed that Si was almost completely removed even when heating evaporation was stopped with one drop remaining. It can be confirmed. When the Si removal rate is calculated from the measurement results shown in FIG. 5 and the Si initial concentration for both sample solutions of Si concentration 2000 ppm and 42 mg / ml, the removal rate is calculated to be 100%. In the sample solution having a Si concentration of 42 mg / ml, Si of about 150 to 250 ppb was detected after the Si removal treatment, which is a level that does not lower the analytical sensitivity of B and P. Moreover, in Examples 1-4, since dry evaporation is performed again, Si is further removed from the value shown in FIG.

[Example 5, Comparative Example 1]
Confirmation of P vaporization by complete dryness P-doped n-type having a P concentration of 2.77E + 15 atoms / cm ³ obtained from a resistance value (1.7 Ω · cm) by a four-deep needle method in the same manner as in Example 4. Six wafer pieces prepared by cleaving the silicon wafer to 0.1 g were prepared, and three of them were subjected to P quantitative analysis by ICP-MS in the same manner as in Example 4.
The remaining three samples were processed and analyzed in the same manner as the above three samples except that they were completely dried without leaving any droplets during heating and evaporation.
Table 2 below shows the P quantification results of the six samples.

As shown in Table 2, in Example 5, compared to Comparative Example 1, a P quantitative value close to the B concentration obtained by the four-probe method was obtained.
Since the silicon wafer subjected to the analysis in Example 5 and Comparative Example 1 was manufactured and stored in an atmosphere of high cleanliness without concern about contamination by B and P, there was no B contamination and P contamination. Can be considered. Therefore, it can be determined that the closer the P concentration obtained from the resistance value obtained by the four-probe method and the P quantitative value obtained by ICP-MS are, the higher the detection sensitivity and the higher the reliability of the analysis result.
For comparison, when the average value of the P quantitative value in Comparative Example 1 is calculated as a percentage with respect to the average value of the P quantitative value in Example 5, it is about 86%. That is, a loss of P of about 15% is generated as compared with the case where heating and evaporation are stopped in a state where one droplet remains by drying completely. Since P, which is a dopant element that determines the resistance value, is required to be analyzed accurately without loss, a measurement error of about 15% is not allowed in terms of detection sensitivity.
From the above results, it was shown that the present invention enables accurate analysis of dopant elements.

  The present invention has been described with reference to the embodiment relating to the method 2. As shown in FIG. 5, it is possible to remove Si to a level that does not affect the ICP-MS analysis by one heating evaporation, and it goes without saying that high sensitivity analysis is possible even by Method 1. Absent.

  The present invention is useful in the field of manufacturing semiconductor substrates and solar cells.

Claims (6)

Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample, comprising: Decomposing part or all of a silicon sample;
Heating and evaporating the silicon-containing solution obtained after the decomposition, provided that one droplet remains, and heating evaporation is stopped;
After the droplets are taken into the acid solution, the acid solution is heated and evaporated; however, the heating and evaporation is stopped in a state where one droplet remains,
Collecting the droplet, and analyzing the dopant element selected from the group consisting of boron and phosphorus contained in the droplet by an inductively coupled plasma mass spectrometer;
A method for analyzing a silicon sample, comprising: The method for analyzing a silicon sample according to claim 1 or 2, wherein the silicon sample is decomposed by liquid phase etching. 4. The method for analyzing a silicon sample according to claim 3, wherein the etching solution used for the liquid phase etching is a mixed acid of hydrofluoric acid and nitric acid. The analysis method of the silicon sample of any one of Claims 1-4 which analyzes the said dopant element contained in this surface layer part by decomposing | disassembling the surface layer part of the said silicon sample. The method for analyzing a silicon sample according to claim 1, wherein the dopant element contained in the silicon sample is analyzed by performing bulk decomposition on the silicon sample.