JP6136205B2 - Contamination assessment method in epitaxial growth apparatus furnace and test wafer for contamination assessment - Google Patents

Description

  The present invention relates to a contamination evaluation method and a contamination evaluation test wafer in an epitaxial growth apparatus furnace.

  Metal contamination is a factor that degrades the characteristics of semiconductor devices. For example, in a back-illuminated solid-state imaging device, metal mixed in an epitaxial silicon wafer serving as the substrate of the device causes a dark current of the solid-state imaging device to increase, and causes a defect called a white defect. The back-illuminated solid-state image sensor has a wiring layer, etc., placed below the sensor part, so that external light can be taken directly into the sensor and clearer images and videos can be taken even in dark places. In recent years, it has been widely used in mobile phones such as digital video cameras and smartphones. Therefore, it is desired to reduce white defect as much as possible.

  One cause of wafer metal contamination is contamination in the epitaxial growth apparatus furnace. That is, when a metal that becomes an impurity is contained in the furnace, the impurity metal in the furnace is taken into the silicon wafer or the growing epitaxial layer in the process of growing the epitaxial layer. Therefore, it is necessary to accurately grasp the presence and level of contamination in the epitaxial growth apparatus furnace, and to perform maintenance in the furnace when contamination is detected.

  As a contamination evaluation method in the epitaxial growth apparatus furnace, a contamination evaluation wafer such as a silicon wafer is placed in the epitaxial growth apparatus furnace and subjected to a predetermined heat treatment such as growing an epitaxial layer, and then an epitaxial layer and / or contamination evaluation is performed. There is a method for measuring the concentration of an impurity element incorporated on the surface of a manufacturing wafer.

  Methods for evaluating impurity metals in such epitaxial layers and silicon wafers include wafer lifetime measurement (WLT), surface photovoltaic method (SPV), deep level transient spectroscopy (DLTS), and chemistry. Analytical methods are included.

  However, although the WLT method is highly sensitive to impurities, it cannot identify and quantify impurity elements. The SPV method can quantitatively measure the Fe concentration in the p-type silicon single crystal substrate with high sensitivity, but cannot identify or quantify elements other than Fe, and when the silicon single crystal substrate is n-type. Is not applicable. The DLTS method is also suitable for identifying and quantifying some impurity metals, but it is difficult to identify and quantify all impurity metals.

  Here, in the chemical analysis method, almost all impurity elements can be identified and quantified by selecting a combination of an impurity recovery method and an analysis device. Patent Document 1 describes an impurity evaluation method for an epitaxial wafer in which an epitaxial layer is formed on a silicon wafer, cooled under predetermined conditions, and then the surface layer of the epitaxial layer is chemically analyzed to measure the impurity element concentration. . In Patent Document 2, a film made of a silicon oxide film, a polysilicon film, or an amorphous silicon film is formed on a silicon wafer, a polysilicon film is further grown thereon, and then the polysilicon film and the film are formed. A method for measuring the concentration of an impurity element by chemical analysis is described.

JP 2011-129570 A WO2008 / 149806

  However, the chemical analysis method has a problem that the impurity element cannot be measured with high accuracy. One of the causes can be diffusion of impurities. That is, the impurities in the epitaxial growth apparatus furnace are not only taken into the silicon wafer or the growing epitaxial layer by the high-temperature heat treatment during the epitaxial growth process, but also solid-phase diffused in the depth direction of the silicon wafer. For this reason, even if the surface layer portion of the grown epitaxial layer or the surface layer portion of the silicon wafer is chemically analyzed, only an amount smaller than the amount of impurity elements actually taken in from the furnace can be detected.

  Accordingly, in view of the above problems, the present invention provides a contamination evaluation method in an epitaxial growth apparatus furnace capable of highly accurately evaluating the degree of contamination in the epitaxial growth apparatus furnace, and a contamination evaluation test wafer used in the method. With the goal.

  According to the study by the present inventors, a silicon wafer having a gettering layer in the surface layer portion is used as a contamination evaluation wafer, which is placed in an epitaxial growth apparatus furnace to grow an epitaxial layer, or equivalent thereto. When heat treatment is performed, impurities taken into the contamination evaluation wafer from the furnace are trapped in the gettering layer and solid phase diffusion is suppressed, so that the gettering layer is chemically analyzed and taken into the contamination evaluation wafer. It has been found that the concentration of impurities can be measured with high accuracy. Based on the above findings, the present inventors have completed the present invention.

  That is, in the method for evaluating contamination in an epitaxial growth apparatus furnace according to the present invention, a first step in which a surface layer portion of a silicon wafer is a gettering layer, the silicon wafer is placed in an epitaxial growth apparatus furnace, and the silicon wafer is placed in a hydrogen atmosphere. And a third step of measuring the concentration of the impurity element present in the gettering layer after the heat treatment and evaluating the degree of contamination in the epitaxial growth apparatus furnace based on the measurement result. It is characterized by having.

The gettering layer is a solid solution layer of the constituent elements of the cluster ions obtained by irradiating the silicon wafer with cluster ions containing carbon as the constituent elements.

  The third step is preferably performed by dissolving at least the gettering layer of the silicon wafer and chemically analyzing the recovered solution.

  In one embodiment, the second step is preferably performed under heat treatment conditions that simulate a heat treatment sequence performed during the epitaxial growth process. In this case, the heat treatment sequence can be performed a plurality of times.

  In another embodiment, in the second step, a source gas may be supplied during the heat treatment to grow an epitaxial layer on the gettering layer. In this case, the third step can be performed by dissolving the epitaxial layer and the gettering layer and chemically analyzing the recovered solution. Or after removing the said epitaxial layer, the said gettering layer can be melt | dissolved and it can also carry out by chemically analyzing the collect | recovered solution.

  It is preferable that after the second step and before the third step, a second heat treatment is performed on the silicon wafer by a heat treatment apparatus different from the epitaxial growth apparatus.

  The test wafer used for the contamination evaluation in the epitaxial growth apparatus furnace of the present invention is a silicon wafer having a gettering layer in the surface layer portion.

This gettering layer is a solid solution layer of the constituent elements of the cluster ions obtained by irradiating the silicon wafer with cluster ions containing carbon as the constituent elements.

  According to the contamination evaluation method in the epitaxial growth apparatus furnace or the test wafer for contamination evaluation of the present invention, the degree of contamination in the epitaxial growth apparatus furnace can be evaluated with high accuracy.

(A)-(E) are the schematic diagrams explaining each process of the contamination evaluation method in the epitaxial growth apparatus furnace by one Embodiment of this invention. (A)-(E) are the schematic diagrams explaining each process of the contamination evaluation method in the epitaxial growth apparatus furnace by other embodiment of this invention. (A)-(F) are the schematic diagrams explaining each process of the contamination evaluation method in the epitaxial growth apparatus furnace by further another embodiment of this invention. (A) And (B) is the schematic diagram which shows the test wafer 200 for pollution evaluation used in the pollution evaluation method in the epitaxial growth apparatus furnace by a reference example . (A) is a schematic diagram explaining the irradiation mechanism in the case of irradiating cluster ions, (B) is a schematic diagram explaining the injection mechanism in the case of injecting monomer ions. It is a graph which shows the result of the chemical analysis in Example 1-1, Comparative example 1-1, and Example 1-2, (A) shows Mo density | concentration, (B) shows Fe density | concentration. It is a graph which shows the result of the chemical analysis in Example 2-1, Comparative example 2-1, Example 2-2, and Example 2-3, (A) shows Mo concentration, (B) shows Fe concentration. . It is a carbon concentration profile obtained by SIMS measurement when monomer ions are implanted into a silicon wafer and when cluster ions are irradiated onto a silicon wafer. (A) is an epitaxial silicon wafer irradiated with cluster ions on a silicon wafer, (B) is a carbon concentration profile by SIMS measurement and Cu concentration profile after gettering ability evaluation of an epitaxial silicon wafer in which monomer ions are implanted into a silicon wafer. It is.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In principle, the same components are denoted by the same reference numerals, and description thereof is omitted. 2 and 3, the thickness of the epitaxial layer 20 is exaggerated with respect to the silicon wafer 10, unlike the actual thickness ratio, for convenience of explanation.

(First Embodiment: Contamination Evaluation Method in Epitaxial Growth Apparatus Furnace)
A contamination evaluation method in the epitaxial growth apparatus furnace according to the first embodiment of the present invention will be described with reference to FIG. In this method, first, the surface layer portion of the silicon wafer 10 is used as a gettering layer. Specifically, as shown in FIGS. 1A and 1B, the surface ions 10A of the silicon wafer 10 are irradiated with cluster ions 12, and the constituent elements of the cluster ions 12 are dissolved in the surface layer of the silicon wafer 10. A solid solution layer 16 is formed. In this embodiment, the solid solution layer 16 becomes a gettering layer, and the silicon wafer in this state is used as a contamination evaluation test wafer 100 in the subsequent steps.

  Next, as shown in FIG. 1C, the silicon wafer 10 is placed in the furnace of the epitaxial growth apparatus 30, and the silicon wafer 10 is heat-treated in a hydrogen atmosphere. In this embodiment, no source gas is supplied during the heat treatment, and no epitaxial layer is formed. This heat treatment is performed under the same heat treatment conditions as the heat treatment sequence performed during the epitaxial growth process. In this step, when the inside of the furnace of the epitaxial growth apparatus 30 is contaminated with the impurity element 22, the impurity element 22 is taken into the surface of the silicon wafer 10 according to the degree of the contamination. The incorporated impurity element 22 is captured in the solid solution layer 16 as a gettering layer, and solid phase diffusion in the depth direction of the silicon wafer 10 is suppressed.

  Finally, as shown in FIG. 1E, the concentration of the impurity element 22 present in the solid solution layer 16 is measured, and the degree of contamination in the furnace of the epitaxial growth apparatus 30 is evaluated based on the measurement result. That is, based on the measurement, the impurity element contaminating the furnace of the epitaxial growth apparatus 30 can be identified, and the degree of the contamination can be grasped.

(Second Embodiment: Contamination Evaluation Method in Epitaxial Growth Apparatus Furnace)
Next, a contamination evaluation method in the epitaxial growth apparatus furnace according to the second embodiment of the present invention will be described with reference to FIG. This method is the same as that shown in FIG. 1 until the step of producing the contamination evaluation test wafer 100 (FIGS. 2A and 2B).

  Next, as shown in FIG. 2C, the silicon wafer 10 is placed in the furnace of the epitaxial growth apparatus 30, and the silicon wafer 10 is heat-treated in a hydrogen atmosphere. In the present embodiment, a source gas is supplied during the heat treatment to grow the epitaxial layer 20 on the solid solution layer 16. In this step, when the inside of the furnace of the epitaxial growth apparatus 30 is contaminated with the impurity element 22, the impurity element 22 is taken into the solid solution layer 16 or the epitaxial layer 20 depending on the degree of the contamination. Among the impurity elements taken into the epitaxial layer 20, most of the elements having a high diffusion rate such as Fe, Ni, Cu, Cr, Zn, and Mg diffuse from the epitaxial layer 20 to the solid solution layer 16, and gettering layers are formed. And solid phase diffusion in the depth direction of the silicon wafer 10 is suppressed. Elements having a low diffusion rate such as Mo, Ti, W, and V are divided into those that are diffused and trapped in the solid solution layer 16 and those that remain in the epitaxial layer 20, but impurities trapped in the solid solution layer 16. Solid phase diffusion in the depth direction of the silicon wafer 10 is suppressed.

  Finally, as shown in FIG. 2E, in this embodiment, the concentration of the impurity element 22 present in the epitaxial layer 20 and the solid solution layer 16 is measured, and the inside of the furnace of the epitaxial growth apparatus 30 is measured based on the measurement result. Assess the degree of contamination.

(Third Embodiment: Contamination Evaluation Method in Epitaxial Growth Apparatus Furnace)
Next, a contamination evaluation method in an epitaxial growth apparatus furnace according to a third embodiment of the present invention will be described with reference to FIG. This method is the same as FIG. 2 until the step of growing the epitaxial layer 20 (FIG. 3C).

  In this embodiment, as shown in FIG. 3E, after removing the epitaxial layer 20, the concentration of the impurity element 22 existing in the solid solution layer 16 is measured as shown in FIG. The degree of contamination in the furnace of the epitaxial growth apparatus 30 is evaluated based on the measurement result.

( Reference example : Method for evaluating contamination in epitaxial growth furnace)
Next, a contamination evaluation method in the epitaxial growth apparatus furnace according to the reference example will be described with reference to FIG. In this method, as shown in FIGS. 4A and 4B, monomer ions (single ions) 14 are implanted into the surface 10A of the silicon wafer 10, and the constituent elements of the monomer ions 14 are formed on the surface layer portion of the silicon wafer 10. The injection layer 18 into which is injected is formed. In this reference example , the implantation layer 18 becomes a gettering layer, and the silicon wafer in this state is used as a contamination evaluation test wafer 200 in the subsequent steps.

  Subsequent steps may take any of the steps described in the first to third steps, and measure the concentration of the impurity element present in the implantation layer 18 as a gettering layer or in addition to the epitaxial layer 20, and the measurement is performed. Based on the results, the degree of contamination in the furnace of the epitaxial growth apparatus is evaluated.

(Detailed explanation of contamination assessment method)
Here, the features of the present invention common to these embodiments are as follows. As shown in FIG. 1B, FIG. 2B, FIG. 3B, and FIG. A silicon wafer on which a gettering layer such as the layer 16 or the injection layer 18 is formed is used as a contamination evaluation test wafer to evaluate the contamination in the epitaxial growth apparatus furnace.

  The technical significance of this feature will be described together with the operational effects of the present invention. When the above evaluation process is performed using a silicon wafer without a gettering layer, solid phase diffusion of impurities taken into the silicon wafer or the epitaxial layer from the furnace of the epitaxial growth apparatus becomes a problem. That is, an element having a high diffusion rate is solid-phase diffused in the depth direction of the silicon wafer, and the ratio of elements remaining in the surface layer portion or epitaxial layer of the silicon wafer is considerably low. Also, some elements having a low diffusion rate are diffused, and the concentration distribution is high on the surface layer portion or epitaxial layer side of the silicon wafer, and becomes lower in the depth direction of the silicon wafer. For this reason, when the surface layer part of a silicon wafer is chemically analyzed, the contamination evaluation in the epitaxial growth apparatus furnace cannot be performed with high accuracy.

  On the other hand, in the present invention, as described above, gettering layers such as the solid solution layer 16 and the injection layer 18 capture impurities and suppress solid-phase diffusion in the silicon wafer. Therefore, by chemically analyzing the gettering layer, the concentration of impurities taken into the contamination evaluation wafer can be measured with high accuracy, and as a result, the degree of contamination in the epitaxial growth apparatus furnace can be evaluated with high accuracy.

  Here, in the present invention, the impurity element concentrations shown in FIGS. 1E, 2E, and 3E are measured by, for example, inductively coupled plasma mass spectrometry (ICP-MS). ) And atomic absorption spectrometry (AAS: Atomic Absorption Spectrometry). Specifically, the gettering layers 16 and 18 or the epitaxial layer 20 in addition thereto are dissolved, and the collected solution is chemically analyzed. A mixed acid of hydrofluoric acid and nitric acid can be used for dissolving the gettering layer and the epitaxial layer, but vapor phase etching using the same mixed acid gas is preferred for high sensitivity analysis. The reaction product after etching can be recovered by a method of scanning the surface of the contamination evaluation wafer with a recovery solution in which an oxidizing agent is added to hydrofluoric acid or the like. The impurity element in the collected liquid thus obtained is analyzed by ICP-MS or the like.

  As the silicon wafer 10, one obtained by slicing a single crystal silicon ingot grown by the Czochralski method (CZ method) or the floating zone melting method (FZ method) with a wire saw or the like can be used. Also, carbon and / or nitrogen may be added to obtain higher gettering ability. Further, an arbitrary impurity may be added to obtain n-type or p-type. Further, in order to evaluate the contamination in the epitaxial growth apparatus furnace with high sensitivity, it is preferable to use a high-purity wafer having a lifetime value of 1000 μsec or more when the wafer lifetime measurement is performed.

  Next, the gettering layer important in the present invention will be described in detail. As the gettering layer, as shown in FIGS. 1 to 3, the solid solution layer 16 of the constituent elements of the cluster ions obtained by irradiating the silicon wafer 10 with the cluster ions 12, or as shown in FIG. Examples thereof include an injection layer 18 of constituent elements of the monomer ions obtained by injecting the monomer ions 14 into the silicon wafer 10.

  For example, when carbon monomer ions 14 are implanted into the silicon wafer 10, as shown in FIG. 5B, the monomer ions 14 repel silicon atoms constituting the silicon wafer, and a predetermined depth position in the silicon wafer. Injected into. The implantation depth depends on the type of constituent elements of the implanted ions and the acceleration voltage of the ions. In this case, the carbon concentration profile in the depth direction of the silicon wafer is relatively broad, and the thickness of the region where the implanted carbon is locally present (that is, the implanted layer 18) is approximately 0.5 to 1 μm. (See FIG. 8 described later). In the process of forming the epitaxial layer 20 after the monomer ion implantation, the implanted carbon is diffused by heat, so the carbon concentration profile after the formation of the epitaxial layer 20 is on both sides of the peak where carbon is locally present. A broad diffusion region is formed. However, the thickness of the injection layer 18 does not change greatly (see FIG. 9B described later).

  The element implanted by the monomer ions may be any element that can cause the implanted layer 18 to function as a gettering layer, and examples thereof include carbon, boron, phosphorus, and arsenic. However, from the viewpoint of obtaining higher gettering ability, the constituent element of the monomer ion is preferably carbon. Since the carbon atom at the lattice position has a smaller covalent bond radius than that of the silicon single crystal, a contraction field of the silicon crystal lattice is formed, so that the gettering ability to attract impurities between the lattices is high.

  Further, a plurality of types of elements may be implanted by injecting monomer ions a plurality of times. In particular, it is preferable to irradiate one or more dopant elements selected from the group consisting of boron, phosphorus, arsenic and antimony in addition to carbon. Since the types of impurity metals that can be efficiently gettered differ depending on the types of elements, a wider range of impurity metals can be efficiently gettered by injecting two or more elements. For example, in the case of carbon, nickel can be efficiently gettered, and in the case of boron, copper and iron can be efficiently gettered.

The acceleration voltage of the monomer ions 14 is generally 150 to 2000 keV / atom, and may be set as appropriate within that range. Further, the dose amount of the monomer ions 14 is not particularly limited, but may be, for example, 1 × 10 ¹³ to 1 × 10 ¹⁶ atoms / cm ² .

  Here, the inventors show higher gettering ability in the solid solution layer 16 by the cluster ion irradiation shown in FIGS. 1 to 3 than in the injection layer 18 by the monomer ion implantation shown in FIG. I found out.

  The solid solution layer 16 formed as a result of the irradiation of the cluster ions 12 is a region where the constituent elements of the cluster ions 12 are locally present as a solid solution at the interstitial positions or substitution positions of crystals on the silicon wafer surface. Work as a ring site. The reason is presumed as follows. That is, elements such as carbon and boron irradiated in the form of cluster ions are localized at a high density in the substitution position / interstitial position of the silicon single crystal. It was experimentally confirmed that the solid solubility of heavy metals (saturation solubility of transition metals) greatly increases when carbon or boron is dissolved to an equilibrium concentration or higher of the silicon single crystal. That is, it is considered that the solid solubility of heavy metals is increased by carbon and boron dissolved to an equilibrium concentration or higher, and the capture rate for heavy metals is thereby remarkably increased.

  In the present specification, the “cluster ion” means an ionized product in which a plurality of atoms or molecules are aggregated to give a cluster having a lump to give a positive charge or a negative charge. A cluster is a massive group in which a plurality (usually about 2 to 2000) of atoms or molecules are bonded to each other.

  When irradiating the silicon wafer 10 with cluster ions 12 made of, for example, carbon and boron, as shown in FIG. 5A, the cluster ions 12 are instantaneously 1350-1400 with the energy when irradiated to the silicon wafer. It becomes a high temperature of about ℃ and silicon melts. Thereafter, the silicon is rapidly cooled, and carbon and boron are dissolved in the surface layer portion in the silicon wafer. That is, the “solid solution layer” in this specification means a layer in which constituent elements of ions to be irradiated are dissolved in the interstitial positions or substitution positions of crystals on the surface layer portion of the silicon wafer. In this case, the concentration profile of carbon and boron in the depth direction of the silicon wafer depends on the acceleration voltage and cluster size of the cluster ions, but is sharper than that of the monomer ions, and the irradiated carbon and boron are localized. The thickness of the existing region (that is, the solid solution layer 16) is approximately 500 nm or less (for example, about 50 to 400 nm) (see FIG. 8 described later). Note that the elements irradiated in the form of cluster ions undergo some thermal diffusion during the formation process of the epitaxial layer 20. For this reason, in the concentration profile of carbon and boron after the formation of the epitaxial layer 20, broad diffusion regions are formed on both sides of the peak where these elements exist locally. However, the thickness of the solid solution layer 16 does not change greatly (see FIG. 9A described later).

  Thus, by irradiating the cluster ions 12, the carbon and boron precipitation regions can be locally and highly concentrated. Moreover, since the solid solution layer 16 is formed in the vicinity of the surface of the silicon wafer, closer gettering is possible. As a result, it is considered that higher gettering ability can be obtained.

  Therefore, by using the gettering layer as the solid solution layer 16 in the present invention, the degree of contamination in the epitaxial growth apparatus furnace can be evaluated with higher accuracy. In addition, since the solid solution layer 16 is thinner than the implantation layer 18 formed by monomer ion implantation, the amount of etching during chemical analysis can be reduced. For this reason, the analysis can be performed in a shorter time, and the analysis accuracy becomes lower as the etching amount at the time of chemical analysis is larger. Therefore, the analysis can be performed with higher accuracy.

  The cluster ion 12 has various clusters depending on the binding mode, and can be generated by a known method as described in the following document, for example. As a method for generating a gas cluster beam, (1) JP-A-9-41138, (2) JP-A-4-354865, and as an ion beam generating method, (1) charged particle beam engineering: Junzo Ishikawa: ISBN978 -4-339-00734-3: Corona, (2) Electron and ion beam engineering: The Institute of Electrical Engineers of Japan: ISBN4-88686-217-9: Ohm, (3) Cluster ion beam basics and applications: ISBN4-526-05765 -7: Nikkan Kogyo Shimbun. In general, a Nielsen ion source or a Kaufman ion source is used to generate positively charged cluster ions, and a large current negative ion source using a volume generation method is used to generate negatively charged cluster ions. It is done.

  Hereinafter, irradiation conditions of cluster ions will be described. First, the irradiation element is not particularly limited, and examples thereof include carbon, boron, phosphorus, and arsenic. However, from the viewpoint of obtaining higher gettering ability, the cluster ions preferably contain carbon as a constituent element. Since the carbon atom at the lattice position has a smaller covalent bond radius than that of the silicon single crystal, a contraction field of the silicon crystal lattice is formed, so that the gettering ability to attract impurities between the lattices is high.

  Further, from the viewpoint of efficiently gettering a wider range of impurity metals, it is more preferable to include two or more elements including carbon as a constituent element. In particular, it is preferable to irradiate one or more dopant elements selected from the group consisting of boron, phosphorus, arsenic and antimony in addition to carbon.

A compound to be ionized is not particularly limited, and ethane, methane, carbon dioxide (CO ₂ ), or the like can be used as a carbon source compound that can be ionized, and diborane, decaborane ( B ₁₀ H ₁₄ ) or the like can be used. For example, when a gas obtained by mixing dibenzyl and decaborane is used as a material gas, a hydrogen compound cluster in which carbon, boron and hydrogen are aggregated can be generated. If cyclohexane (C ₆ H ₁₂ ) is used as a material gas, cluster ions composed of carbon and hydrogen can be generated. As the carbon source compound, it is particularly preferable to use a cluster C _n H _m (3 ≦ n ≦ 16, 3 ≦ m ≦ 10) formed from pyrene (C ₁₆ H ₁₀ ), dibenzyl (C ₁₄ H ₁₄ ) or the like. This is because it is easy to control a small-sized cluster ion beam.

  Next, the acceleration voltage and cluster size of cluster ions affect the position of the peak of the concentration profile of each constituent element in the solid solution layer 16 in the depth direction. In the present embodiment, from the viewpoint of obtaining a higher gettering capability, the concentration profile peak in the depth direction of each constituent element in the solid solution layer 16 is within a range where the depth from the surface 10A of the silicon wafer 10 is 150 nm or less. The cluster ions 12 are irradiated so that is positioned.

When C _n H _m (3 ≦ n ≦ 16, 3 ≦ m ≦ 10) is used as a cluster ion as a necessary condition for setting the peak position within the depth range, the acceleration voltage per carbon atom is , 0 keV / atom and 50 keV / atom or less, preferably 40 keV / atom or less. The cluster size is 2 to 100, preferably 60 or less, more preferably 50 or less.

  For adjusting the acceleration voltage, two methods of (1) electrostatic acceleration and (2) high frequency acceleration are generally used. As the former method, there is a method in which a plurality of electrodes are arranged at equal intervals and an equal voltage is applied between them to create an equal acceleration electric field in the axial direction. As the latter method, there is a linear linac method in which ions are accelerated using a high frequency while running linearly. In the present specification, “cluster size” means the number of atoms or molecules constituting one cluster. The cluster size can be adjusted by adjusting the gas pressure of the gas ejected from the nozzle, the pressure of the vacuum vessel, the voltage applied to the filament during ionization, and the like. The cluster size can be obtained by obtaining a cluster number distribution by mass spectrometry using a quadrupole high-frequency electric field or time-of-flight mass spectrometry and taking an average value of the number of clusters.

Moreover, the dose amount of cluster ions can be adjusted by controlling the ion irradiation time. In this embodiment, the carbon dose of the cluster ions 12 is 1 × 10 ¹³ to 1 × 10 ¹⁶ atoms / cm ² , preferably 5 × 10 ¹⁵ atoms / cm ² or less. If it is less than 1 × 10 ¹³ atoms / cm ² , the gettering ability may not be sufficiently obtained. If it exceeds 1 × 10 ¹⁶ atoms / cm ² , the surface of the epitaxial layer 20 may be seriously damaged. Because there is.

In the solid solution layer 16 by the cluster ion irradiation, the following is obtained as the concentration profile in the depth direction of the constituent elements of the cluster ions measured by SIMS (secondary ion mass spectrometer). First, compared to monomer ion implantation, in the case of cluster ion irradiation, the irradiation element precipitation region can be locally and highly concentrated, so that the half width W of the concentration profile of the constituent elements in the solid solution layer 16 is 100 nm. Or less than 85 nm. The lower limit can be set to 10 nm. Further, the peak concentration of the concentration profile is in the range of 1 × 10 ¹⁷ to 1 × 10 ²² atoms / cm ³ .

  As the epitaxial layer 20 shown in FIGS. 2C and 3C, a silicon epitaxial layer can be cited, and can be formed under general conditions. For example, a source gas such as dichlorosilane or trichlorosilane is introduced into the chamber using hydrogen as a carrier gas, and the growth temperature differs depending on the source gas used, but silicon is formed by CVD at a temperature in the range of about 1000 to 1200 ° C. An epitaxial layer 20 can be grown on the wafer 10. Also, an optional dopant gas may be introduced to form a p-type or n-type epitaxial layer.

  The standard heat treatment sequence during epitaxial growth is as follows. First, the inside of the furnace is a hydrogen atmosphere, the semiconductor wafer 10 is put into the furnace at a furnace temperature of 600 to 900 ° C., and a first predetermined range in the range of 1000 to 1200 ° C. at a temperature rising rate of 1 to 15 ° C./second. The temperature is raised to a temperature, and hydrogen baking is performed at that temperature for a predetermined time, for example, 30 seconds to 1 minute. Thereafter, similarly, the temperature is raised / fallen to a predetermined second temperature in the range of 1000 to 1200 ° C., and epitaxial growth is performed at that temperature for a predetermined time. Thereafter, the temperature is lowered to 650 to 800 ° C. The first temperature and the second temperature may be the same. This heat treatment sequence may be performed a plurality of times, but the thickness of the epitaxial layer 20 is preferably 10 μm or less, and particularly preferably 5 μm or less. If it exceeds 10 μm, diffusion of slowly diffusing impurities into the gettering layer may be reduced, and if it exceeds 5 μm, it may be difficult to analyze the epitaxial layer and the gettering layer at the same time. is there.

  Moreover, if the heat treatment similar to the heat treatment sequence at the time of epitaxial growth treatment is performed except that the source gas is not supplied, the embodiment of FIG. In this case, it is preferable to perform this heat treatment sequence a plurality of times because the amount of impurities taken in can be increased and the analysis can be performed with higher accuracy without increasing the etching amount of the gettering layer.

  Here, as shown in FIG. 1D, FIG. 2D, and FIG. 3D, after the heat treatment in the epitaxial growth apparatus 30 and before the impurity element concentration measurement, it is different from the epitaxial growth apparatus 30. The second heat treatment may be performed on the silicon wafer 10 by the heat treatment apparatus 40. Of the impurity elements taken in from the back surface opposite to the front surface 10A of the silicon wafer by this heat treatment, elements having a high diffusion rate such as Fe, Ni, Cu, Cr, and Mg are transferred to the gettering layers 16 and 18. Can be diffused, and analysis with higher accuracy becomes possible. The conditions for the second heat treatment are not particularly limited, but the atmosphere temperature is 800 to 1100 ° C. in a non-oxidizing, oxidizing or reducing atmosphere, and the holding time at the atmosphere temperature is, for example, 10 seconds to 100 hours. Under these conditions, it is preferable to carry out at a temperature and time at which the impurity to be analyzed is sufficiently diffused to the gettering layer.

  Finally, the difference in effect between the first to third embodiments will be described. In the first embodiment, since the impurity element 22 is directly taken into the gettering layers 16 and 18 and only the gettering layer needs to be dissolved during the chemical analysis, the etching amount is small. For this reason, it is most preferable because an element having a high diffusion rate and an element having a low diffusion rate can be analyzed in a short time and with the highest accuracy.

  The second embodiment is preferable because an impurity element is taken into the epitaxial layer 20 and the gettering layers 16 and 18 and both of them are chemically analyzed, so that both a fast diffusion element and a slow element can be analyzed with high accuracy. .

In the third embodiment, since the impurity element is taken into the epitaxial layer 20 and the gettering layers 16 and 18 and only the gettering layers 16 and 18 are chemically analyzed, only the elements having a high diffusion rate are limited to a short time and high accuracy. It is suitable for a simple analysis. In this case, if the second heat treatment shown in FIG. 3D is performed, an impurity element having a low diffusion rate in the epitaxial layer 20 is also diffused and trapped in the gettering layers 16 and 18. Even slower elements are preferable because they can be analyzed with high accuracy. The conditions for the second heat treatment can be appropriately set in consideration of the thickness of the epitaxial layer 20 and the diffusion coefficient of the impurity element. When the diffusion coefficient of typical impurity metals (diffusion distance (m) in heat treatment for 1 hour at 1000 ° C.) is shown, as an example of a metal having a high diffusion rate, Fe: 6.89 × 10 ⁻⁴ , Cr: 4 .45 × 10 ⁻⁴ , Cu: 4.10 × 10 ⁻³ , Ni: 2.67 × 10 ⁻³ , W: 5.69 × 10 ⁻⁸ , Mo: 6. 57 × 10 ⁻⁶ , Ti: 2.04 × 10 ⁻⁶ .

(Experimental example 1)
(Example 1-1)
An n-type silicon wafer (diameter: 300 mm, thickness: 725 μm, dopant: phosphorus, specific resistance 10 Ωcm) obtained from a CZ single crystal silicon ingot was prepared. Next, C ₅ H ₅ cluster ions are generated from dibenzyl (C ₁₄ H ₁₄ ) using a cluster ion generator (manufactured by Nissin Ion Equipment Co., Ltd., model number: CLARIS), and the dose amount is 9.0 × 10 ^13. The surface of the silicon wafer is irradiated on the surface of the silicon wafer under conditions of Clusters / cm ² (4.5 × 10 ¹⁴ atoms / cm ^{2 in} terms of carbon) and an acceleration voltage of 14.8 keV / atom per carbon atom. A solid solution layer was obtained.

  Next, the silicon wafer was transferred into a furnace of a single wafer epitaxial growth apparatus (Applied Materials Co., Ltd.), supplied with only hydrogen gas, and subjected to heat treatment at 1100 ° C. for 2 minutes. In order to investigate the effect of the present invention, the heat treatment was performed in a state where the metal impurities were relatively large at the time of start-up immediately after maintenance of the epitaxial growth apparatus.

  When the carbon concentration profile in the depth direction of the silicon wafer was obtained by SIMS measurement, the thickness of the solid solution layer (carbon peak width) was 350 nm. Therefore, 1 μm of the surface layer of the silicon wafer is etched in the gas phase with a mixed acid gas using hydrofluoric acid and nitric acid, and the reaction product after the etching is recovered with a recovery liquid in which an oxidizing agent is added to hydrofluoric acid. The metal impurity concentration was analyzed by ICP-MS.

(Comparative Example 1-1)
The experiment was performed in the same procedure as in Example 1-1 except that the cluster ion irradiation was not performed.

(Example 1-2)
After the heat treatment in the epitaxial growth apparatus, the silicon wafer was transferred to a vertical heat treatment apparatus (manufactured by Hitachi Kokusai Electric), and the silicon wafer was subjected to a heat treatment at 1000 ° C. for 30 minutes in a nitrogen gas atmosphere. The experiment was performed in the same procedure as in Example 1-1.

<Evaluation results>
As a measurement result of ICP-MS, an analysis result of Mo is shown in FIG. 6A as a representative of a slow diffusion metal, and an analysis result of Fe is shown in FIG. 6B as a representative of a fast diffusion metal. As is clear from FIG. 6A, the Mo concentration detected in Examples 1-1 and 1-2 was higher than that in Comparative Example 1-1. This is because in Comparative Example 1-1, although Mo diffuses slowly, a part of Mo diffuses to a depth of 1 μm or more from the surface of the silicon wafer by the heat treatment in the epitaxial growth apparatus. This is because in 1-2, Mo was trapped in the solid solution layer and its diffusion was suppressed. In Comparative Example 1-1, even if the etching depth is increased to detect more Mo, not only does the etching time take, but also the recovery of the solution becomes difficult. Analysis of accuracy is difficult.

  Further, as is clear from FIG. 6B, in the case of Fe, it was below the detection limit in Comparative Example 1-1. This is because Fe diffuses very rapidly and diffuses deep into the silicon wafer by the heat treatment in the epitaxial growth apparatus, so that the Fe concentration in the surface layer portion is greatly reduced. On the other hand, in Examples 1-1 and 1-2, Fe could be detected. In particular, in Example 1-2 in which an additional heat treatment was performed after the heat treatment in the epitaxial growth apparatus, the detected Fe concentration was higher than in Example 1-1. This is presumably because Fe from the back surface of the silicon wafer was also captured in the solid solution layer by the additional heat treatment.

(Experimental example 2)
(Example 2-1)
An n-type silicon wafer (diameter: 300 mm, thickness: 725 μm, dopant: phosphorus, specific resistance 10 Ωcm) obtained from a CZ single crystal silicon ingot was prepared. Next, C ₅ H ₅ cluster ions are generated from dibenzyl (C ₁₄ H ₁₄ ) using a cluster ion generator (manufactured by Nissin Ion Equipment Co., Ltd., model number: CLARIS), and the dose amount is 9.0 × 10 ^13. The surface of the silicon wafer is irradiated on the surface of the silicon wafer under conditions of Clusters / cm ² (4.5 × 10 ¹⁴ atoms / cm ^{2 in} terms of carbon) and an acceleration voltage of 14.8 keV / atom per carbon atom. A solid solution layer was obtained.

  Next, the silicon wafer is transferred into a furnace of a single wafer epitaxial growth apparatus (Applied Materials), and after hydrogen baking at 1200 ° C. for 30 seconds, hydrogen is a carrier gas, trichlorosilane is a source gas, and diborane gas is a dopant gas. Then, epitaxial growth was performed at 1100 ° C. for 2 minutes to form a p-type silicon epitaxial layer (thickness: 4 μm, specific resistance: 20 Ωcm) on the silicon wafer, and an epitaxial silicon wafer was produced. In order to investigate the effect of the present invention, the epitaxial growth was performed in a state where the metal impurities were relatively large at the time of startup immediately after maintenance of the epitaxial growth apparatus.

  When the carbon concentration profile in the depth direction of the silicon wafer was obtained by SIMS measurement, the thickness of the solid solution layer (carbon peak width) was 350 nm. Therefore, a total of 5 μm of the epitaxial layer and 1 μm of the surface layer of the silicon wafer is etched in a gas phase with a mixed acid gas using hydrofluoric acid and nitric acid, and the reaction product after etching is a recovered liquid obtained by adding an oxidizing agent to hydrofluoric acid. And the metal impurity concentration in the recovered solution was analyzed by ICP-MS.

(Comparative Example 2-1)
An experiment was performed in the same procedure as in Example 2-1, except that no cluster ion irradiation was performed.

(Example 2-2)
After the heat treatment in the epitaxial growth apparatus, the epitaxial silicon wafer was transferred to a vertical heat treatment apparatus (manufactured by Hitachi Kokusai Electric), and the epitaxial silicon wafer was heat-treated at 1000 ° C. for 30 minutes in a nitrogen gas atmosphere. Except that, the experiment was performed in the same procedure as in Example 2-1.

(Example 2-3)
After etching and removing the epitaxial layer with hydrofluoric acid and nitric acid, only the 1 μm surface layer of the silicon wafer is etched in a gas phase with a mixed acid gas using hydrofluoric acid and nitric acid, and the reaction product after etching is treated with hydrogen fluoride. The experiment was performed in the same procedure as in Example 2-2, except that the acid was recovered in a recovery solution obtained by adding an oxidizing agent to the acid, and the metal impurity concentration in the recovery solution was analyzed by ICP-MS.

<Evaluation results>
As a measurement result of ICP-MS, an analysis result of Mo representing a slow diffusion metal is shown in FIG. 7A, and an analysis result of Fe representing a fast diffusion metal is shown in FIG. 7B. As is clear from FIG. 7A, the Mo concentration detected in Examples 2-1 to 2-3 was higher than that in Comparative Example 2-1. In Comparative Example 2-1, although Mo diffuses slowly, a part of Mo diffuses deeper than 1 μm from the surface of the silicon wafer during epitaxial growth, while Examples 2-1 to 2- This is because Mo was trapped in the solid solution layer 3 and its diffusion was suppressed. In Comparative Example 2-1, even if the etching depth is increased to detect more Mo, not only etching time is required but also recovery of the solution becomes difficult. Analysis of accuracy is difficult. In Example 2-3, the detected Mo concentration was almost the same as in Examples 2-1 and 2-2. However, since the etching amount was small, stable analysis was possible.

  Further, as apparent from FIG. 7B, in the case of Fe, it was below the detection limit in Comparative Example 2-1. This is because Fe diffuses very quickly and diffuses deep into the silicon wafer during epitaxial growth, and the Fe concentration in the surface layer portion is greatly reduced. On the other hand, in Examples 2-1 to 2-3, Fe was detected. In particular, in Examples 2-2 and 2-3 in which additional heat treatment was performed after epitaxial growth, the detected Fe concentration was higher than that in Example 2-1. This is presumably because Fe from the back surface of the silicon wafer was also captured in the solid solution layer by the additional heat treatment.

  As is clear from these experimental examples 1 and 2, by using the present invention, not only the detection accuracy is improved for the slowly diffusing impurities, but also the impurities that have been diffused quickly and could not be detected conventionally. It was found that it could be detected. Although Mo and Fe are shown in this experimental example, other impurities such as Ti, W, V, etc., which are slow diffusion, and Ni, Cu, Cr, Zn, Mg, etc., which are fast diffusion can be similarly analyzed with high sensitivity.

(Reference experiment example)
Next, the following experimental example shows that the solid solution layer by cluster ion irradiation exhibits higher gettering ability than the injection layer by monomer ion implantation.

(Reference Experiment Example 1)
A silicon epitaxial wafer having a solid solution layer by cluster ion irradiation was produced in the same procedure as that produced in Example 2-1 described above except that the film thickness of the epitaxial layer was 3 μm.

(Reference Experiment Example 2)
In place of irradiation with cluster ions, carbon monomer ions are generated from CO ₂ gas, and implanted on the silicon wafer surface under the conditions of a dose of 9.0 × 10 ¹³ atoms / cm ² and an acceleration voltage of 150 keV / atom. Produced a silicon epitaxial wafer in the same manner as in Reference Experimental Example 1.

<Evaluation method and evaluation results>
The following evaluation was performed for each sample prepared in Reference Experimental Examples 1 and 2.

(1) SIMS measurement First, in order to clarify the difference in carbon distribution immediately after irradiation of cluster ions and immediately after injection of monomer ions, with respect to Reference Experimental Examples 1 and 2, the silicon wafer before the epitaxial layer formation, SIMS measurement was performed. The obtained carbon concentration profile is shown in FIG. 8 for reference. Here, the depth of the horizontal axis in FIG. 8 is zero on the surface of the silicon wafer.

  Next, SIMS measurement was performed on the epitaxial silicon wafers of Reference Experimental Examples 1 and 2. The obtained carbon concentration profiles are shown in FIGS. 9 (A) and 9 (B), respectively. The depth of the horizontal axis in FIGS. 9A and 9B is zero on the surface of the epitaxial silicon wafer.

(2) Evaluation of gettering ability The gettering ability was evaluated using Cu as an example. First, the epitaxial layer surface of each sample produced was intentionally contaminated by a spin coating contamination method in Cu contamination solution ^{(1.0 × 10 12 / cm 2} ), subsequently 900 ° C., was subjected to heat treatment for 30 minutes . Thereafter, SIMS measurement was performed. The obtained Cu concentration profiles are shown in FIGS. 9 (A) and 9 (B).

<Consideration of evaluation results>
As shown in FIG. 8, when comparing the carbon concentration profile of the silicon wafer before the epitaxial layer formation, the carbon concentration profile is sharp in the case of cluster ion irradiation, and the carbon concentration profile is broad in the case of monomer ion implantation. From this, it is presumed that the tendency of the carbon concentration profile is the same after the formation of the epitaxial layer.

Actually, as can be seen from the carbon concentration profile when the epitaxial layer is formed (FIGS. 9A and 9B), the cluster ion irradiation causes a local and higher concentration solid solution layer than the monomer ion implantation. Is formed. 9A, the half width of the carbon concentration profile is 95 nm, the peak concentration is 1.5 × 10 ¹⁹ atoms / cm ³ , and the peak position is 100 nm from the surface of the silicon wafer (3.1 μm from the epitaxial layer surface). Met. On the other hand, in FIG. 9B, the half width of the carbon concentration profile is 250 nm, the peak concentration is 1.0 × 10 ¹⁹ atoms / cm ³ , and the peak position is 300 nm from the surface of the silicon wafer (3.3 μm from the surface of the epitaxial layer). It was the position of. In the Cu concentration profile, the peak concentration is higher in FIG. 9A when cluster ions are irradiated than in the case of FIG. 9B, so that more Cu is captured and higher. It turns out that the gettering ability is demonstrated.

  ADVANTAGE OF THE INVENTION According to this invention, the contamination evaluation method in the epitaxial growth apparatus furnace and the test wafer for contamination evaluation which can evaluate the contamination degree in an epitaxial growth apparatus furnace with high precision can be provided.

DESCRIPTION OF SYMBOLS 10 Silicon wafer 10A Silicon wafer surface 12 Cluster ion 14 Monomer ion 16 Solid solution layer (gettering layer)
18 Injection layer (gettering layer)
DESCRIPTION OF SYMBOLS 20 Epitaxial layer 22 Impurity element 30 Epitaxial growth apparatus 40 Heat processing apparatus 100,200 Contamination evaluation test wafer

Claims (9)

A first step in which the surface layer of the silicon wafer is a gettering layer;
Placing the silicon wafer in an epitaxial growth apparatus furnace, and performing a heat treatment on the silicon wafer in a hydrogen atmosphere;
A third step of measuring a concentration of the impurity element present in the gettering layer after the heat treatment, and evaluating a degree of contamination in the epitaxial growth apparatus furnace based on the measurement result;
Have
The contamination evaluation method in an epitaxial growth apparatus furnace, wherein the gettering layer is a solid solution layer of constituent elements of the cluster ions obtained by irradiating the silicon wafer with cluster ions containing carbon as a constituent element.   The contamination evaluation method in an epitaxial growth apparatus furnace according to claim 1, wherein the third step is performed by dissolving at least the gettering layer of the silicon wafer and chemically analyzing the recovered solution.   The contamination evaluation method for an epitaxial growth apparatus furnace according to claim 1 or 2, wherein the second step is performed under heat treatment conditions simulating a heat treatment sequence performed during the epitaxial growth processing.   The contamination evaluation method in the epitaxial growth apparatus furnace according to claim 3, wherein the heat treatment sequence is performed a plurality of times.   The contamination evaluation method in an epitaxial growth apparatus furnace according to claim 1 or 2, wherein in the second step, a source gas is supplied during the heat treatment to grow an epitaxial layer on the gettering layer.   The contamination evaluation method in an epitaxial growth apparatus furnace according to claim 5, wherein the third step is performed by dissolving the epitaxial layer and the gettering layer and chemically analyzing the recovered solution.   6. The method for evaluating contamination in an epitaxial growth apparatus furnace according to claim 5, wherein the third step is performed by dissolving the gettering layer after removing the epitaxial layer and chemically analyzing the recovered solution.   8. The method according to claim 1, further comprising performing a second heat treatment on the silicon wafer by a heat treatment apparatus different from the epitaxial growth apparatus after the second process and before the third process. 2. A method for evaluating contamination in an epitaxial growth apparatus furnace according to item 1. A test wafer used for contamination evaluation in an epitaxial growth apparatus furnace, a silicon wafer having a gettering layer in a surface layer portion,
The gettering layer is a test wafer for contamination evaluation, which is a solid solution layer of constituent elements of the cluster ions obtained by irradiating the silicon wafer with cluster ions containing carbon as the constituent elements.

Images