TW201144224A - Quality control process for UMG-Si feedstock - Google Patents

Abstract

A quality control process for determining the concentrations of boron and phosphorous in a UMG-Si feedstock batch is provided. A silicon test ingot is formed by the directional solidification of molten UMG-Si from a UMG-Si feedstock batch. The resistivity of the silicon test ingot is measured from top to bottom. Then, the resistivity profile of the silicon test ingot is mapped. From the resistivity profile of the silicon test ingot, the concentrations of boron and phosphorous of the UMG-Si silicon feedstock batch are calculated. Additionally, multiple test ingots may be grown simultaneously, with each test ingot corresponding to a UMG-Si feedstock batch, in a multi-crucible crystal grower.

Description

201144224 VI. Description of the Invention: [Technical Fields of the Invention] Field of the Invention The present invention is broadly related to the field of the processing of the stone processing, and more particularly to the purification of the upgraded metallurgical grade. t Prior Art Background of the Invention Photovoltaic i industry (pv) is rapidly evolving, and is (4) an increase in consumption over the more traditional use as an integrated circuit (IC) application. To date, the demand for solar cells has begun to compete with the demand for the IC industry. Existing manufacturing technologies, both integrated circuits (ICs) and the solar cell industry, require repeated and purified germanium materials as starting materials. The material substitutes for solar cells range from single crystal, electronic grade (EG) to relatively dirty metallurgical grade (MG). EG矽 produces solar cells that are close to the theoretical limit of efficiency, but their price is too high. On the other hand, it is typical that a solar cell that does not work can be produced. Early use of polycrystalline silicon solar cells achieved a very low efficiency of about 6%. In this context, efficiency is a measure of the amount of energy incident on a battery that is collected and converted into a component of current. However, there are other semiconductor materials that can be used in the manufacture of solar cells. However, in practice, almost 9% of commercial solar cells are made from crystalline germanium. Bacies commercially available to date with 24% efficiency can be made from higher purity materials and improved processing techniques. These engineering developments have helped the industry approach the theoretical limit of single-junction solar cell efficiency by 31〇/〇. 201144224 Because of the high cost and complex processing requirements for obtaining and using high-purity tantalum raw materials and the competitive demand from the ic industry, the demand for solar cells cannot be used by EG, MG or other niobium manufacturers using known processing techniques. Satisfy. As long as this unsatisfactory condition persists, economical solar cells for large-scale power production cannot be achieved. Several factors determine the quality of the raw tantalum material that is useful for solar cell manufacturing. The quality of the raw materials often varies depending on the amount of impurities present in the material. The main elements that need to be controlled and removed to improve the quality of the tantalum material are boron (B), phosphorus (P) and aluminum (A1) because they significantly affect the resistivity of tantalum. Raw materials based on graded metallurgical grade (UM) crucibles often include similar boron and scales. Although chemical analysis can be used to measure the concentration of certain elements, this method requires too small a sample size (several grams) and often provides variable results, such as 'the amount of boron present can be from 5 parts per million parts by weight. (ppmw) changes to 1 ppmw. Furthermore, chemical analysis on different batches has provided consistent boron and phosphorus concentrations, but it has extreme variations in electrical parameters. These unreliable results can result from this relatively large amount of impurities. Resistivity is one of the most important qualities of ♦ (Si) used to make solar cells. This is because the solar cell efficiency is sensitive to resistivity. The current state of the art of solar cell technology typically requires a resistivity value in the range 〇 5 ohm centimeters to 5, 〇 ohm cm. The um矽-based raw materials currently manufactured often reach a minimum electrical conductivity of less than 〇·5 ohm centimeters (typically specified by the solar cell manufacturer) and the rate 8 has a simple reason for this: For UM-Si, expensive processes are primarily concerned with the removal of non-metals (including dopant atoms B and P). In order to reduce costs, there is a clear reduction in the trend of this addition. That is, UM-Si typically still contains high doping atomic concentrations. Segregation during directional solidification is often used in this process to purify to achieve upgraded metallurgical grades. The impurity removal method includes directional solidification which concentrates impurities (such as B, P, A, C, and transition metal) in the final portion of the resulting ingot ingot (often the top end of the ingot) to crystallize. In the perfect case, the crystallinity during the directional solidification process will be consistent from the top to the bottom and the solid-liquid interface will be planar throughout the monolith. This will result in a consistent impurity concentration profile throughout the ingot from top to bottom, allowing impurities in the ingot to be removed in accordance with a planar cut across the ingot (which removes the top portion of the ingot). However, it is difficult to control the thermal field during the directional solidification process and it is often formed and crystallized unevenly in the ingot ingot. This causes impurities from the top to the bottom. The concentration profile is uneven across the ingot (i.e., from one end of the ingot to the other). This effect is further amplified in the mass production of a large number of Shi Xi. Because different regions of the ingot have different impurity profiles, different resistivity plots 'planar cuts across the ingot cannot maximize the usable yield' while still removing most of the concentrated impurities. Furthermore, the variability in the quality of the feed UMG-Si feedstock requires a controlled method to test and analyze the UMG-Si material quality. Typically, elements such as boron (B) and dish (P) can reduce the quality of the Si material. If they are not controlled within certain concentration limits, they will produce considerable changes in the ingot resistivity. Its halogens (such as, but not limited to, carbon, oxygen, nitrogen), compounds with these elements (especially Sic) also degrade the quality of the cake. Due to the large effects of these and similar impurities, the raw materials 201144224 should be analyzed and tested to ensure proper quality. The variation in the feed rate of the feedstock between the feedstock affects the bottom to the top resistivity of the crucible and the electrical steepness. Type part). (The supplier of the n-type UMG-Si raw material does not establish quality control for transport to the buyer. Typical chemical analysis often produces fine fruit. This is due to the relatively small amount of impurities.) ^Lie responders often test too small sample sizes, which is related to the variability of the phosphorus supply concentration in the raw material. In addition, the superimposed measurement error: boron and the result are uncertain. These measurement errors occur. Signs, _ quantity..., chemical analysis on different peaches produces the same boron and phosphorus content, but there are variations in electrical parameters. Eight slabs of bismuth ingots are cast on a batch of UM G - S i raw materials. In the meantime, these changes in the batch are unacceptable. [In the Ming] Summary of the Invention Therefore, there is an increasing demand for quality control methods for UMG-Si raw material materials to provide reliable impurity data/measurement. Accurate and accurate provision of raw material batches from the sample test ingots. There is an increasing demand for more accurate identification of impurity concentration profiles in batches of raw materials' so that suppliers can be more convinced UMG-Si' and solar cell manufacturers that manufacture the desired impurity concentration threshold can improve the wafer yield. A simple method is used to measure UMG-based polysilicon materials (which produce good ingots) The impurity concentration of the yield and the improved mechanical and electrical properties of the material 'the latter is related to the quality of the solar cell' is further required by 201144224. This method should be easily transferable to a higher grade of non-1; Evening (partially or exclusively for crystalline single crystal germanium materials 'for example by applying cz technology or FZ technology) ° provides a measure for boron and fill concentration in batch UMG-Si materials according to the disclosed subject matter The method 'which substantially eliminates or reduces the disadvantages and problems associated with previously developed UMG-Si impurity concentration measurement methods. The present disclosure provides a method for measuring the concentration of shed and helium in a batch of UMG_Si material. The solidification is derived from the molten UMG-S i of the UMG_Sl raw material batch to form the niobium test ingot. The resistivity of the ingot is measured from the top to the bottom. Resistivity plot. Calculate the boron and phosphorus concentrations of the UMG-Si® raw material batch from the resistivity plot of the test block. According to one of the disclosed subject matter, 'from different UMG-Si raw materials Simultaneous growth of a plurality of niobium test ingots. The process advantages of the present disclosure include more accurate niobium impurity concentration data, which allows for higher usable Shih's yield, improved UMG-Si process control, and improved UMG-Si fabrication. Efficiency and cost. Further process advantages of calculating the impurity concentration of the UM G - S i raw material batch based on the resistivity plot of the tested ingot include a more consistent and accurate impurity concentration measurement. It will be apparent from the description provided herein. Theme events and additional novel features. The purpose of this overview is not a comprehensive description of the claimed subject matter, but rather a brief overview of some of the functional aspects of the subject matter. Other systems, methods, features, and advantages will be apparent from the following figures and detailed descriptions. This additional system, method, features, and advantages are intended to be included within the scope of the accompanying claims. BRIEF DESCRIPTION OF THE DRAWINGS For a more complete understanding of the disclosed subject matter and its advantages, reference is now made to the accompanying description of the accompanying drawings, in which The process flow for reducing the boron, phosphorus and aluminum contents in the crucible; Figure 2 shows the actual measured impurity profile of the different UMG raw material batches; Figure 3 is a graph showing the 11] ^(3_The concentration curve of the impurity boron and phosphorus in the ingot; Fig. 4 is a graph showing the resistivity curve of the UMG-Si ingot measured in Fig. 3 (calculation pair) The measured resistivity); Figure 5 shows the cross-sectional image of the UMG-Si block after directional solidification; Figure 6 shows the cross-sectional image of the UMG-Si ingot after directional solidification and according to the disclosed subject matter The resulting cutting line; Figure 7 is a graphical illustration of the 3D solidification interface of the ingot casting block; Figure 8 is a graph showing the concentration profiles of the impurities boron, phosphorus and aluminum in the UMG-Si ingot; Figure 9 is a UMG-Si ingot as described in Figure 8. Concentration of cross sectional images; 8201144224 Figure 10 a cross-sectional image based UMg in FIG. 8 as described; Singles -Sl concentration of ingot

Figure 11 is a UIyIG cross-sectional image depicted in Figure 8; • The boron concentration of the Si ingot is shown in Figure 12 of the bidirectional solidification oven, a process flow showing the side of the solidified tantalum material Figure 13 is a top view of the process of solidification of the crucible material; it shows that in the bidirectional solidification furnace, Figure 14 is a diagram showing the solidification interface of the dumping block generated in the bidirectional solidification furnace. Explanation: Fig. 15 is a graph showing resistivity curves and cutting lines of various impurity concentrations; Figure 16-18 is a graph-curve graph illustrating the resistivity curve and impurity concentration curve in the block Figure 19 is a graph showing the resistivity curve of the Shixi ingot in Figure 16_18 (in ohm-cm to solidification); Figure 20 is shown in Figure 19. Corresponding impurity concentration graph of the resistivity graph; Figure 21 shows a schematic diagram of the process flow for reducing boron, phosphorus and aluminum in the prior art; Figures 22 and 23 are graphs showing different UMG_Si Actual measurement of raw material batch Resistivity; Figure 24 shows ICPMS data for B and P of an example test ingot for simultaneous directional solidification; 201144224 Figure 25 is a graph showing the measurement of Batch 1 in Figure 24 Resistivity data; Figure 26 is a graph showing the measured resistivity data for Batch 2 in Figure 24; Figure 27 is a graph showing the batch in Figure 24. 3 measured resistivity data; Fig. 28 is a graph showing the measured resistivity data of batch 4 in Fig. 24; Fig. 29 is a photograph of the cast ingot being cast; Fig. 30 A specific example of a crystallizer configuration based on the disclosed subject matter is depicted graphically; Figure 31 is a photo representation showing a modification, each configuration of a thermal configuration of a single thermal configuration to four (4) Figure 32 is a photographic embodiment showing impurities actually detected in an actual UMG-Si ingot; and Figure 33 is a process flow showing a specific example of the disclosed UMG-Si control method The main steps. I. Embodiment 3 Detailed Description of the Preferred Embodiments The following description is not intended to limit the concept, but is used for the purpose of describing the common principles of the present disclosure. The scope of the disclosure should be determined with reference to the scope of the patent application. While described with reference to the purification of aluminum-rich UMG(R), it is well known to those skilled in the art that the principles discussed herein can be applied to any upgraded metallurgical grade material. 10 201144224 A preferred embodiment of the disclosed subject matter is illustrated in the figures, using similar numbers to indicate similar and corresponding portions of the different figures. Figure 1 shows the process flow used to reduce the shed, fill and inscriptions in Shi Xi. In step 2, pure raw materials such as 'quartz and coal are selected to produce MG-Si containing a low boron content. Then, step 4 further reduces the aluminum content by MG-Si refining. Additionally, the boron content can be further reduced (e.g., in a furnace containing an oxy-fuel burner), ultimately resulting in UMG-Si. Then, in order to further reduce impurities (such as boron, phosphorus and aluminum), UMG-Si is often processed through a directional solidification system until it is ready to release the tantalum material (typically when the boron concentration has been reduced below the specified threshold concentration). ). In both the first DSS pass 6 and the second DSS pass 8, the portion of the ingot (usually the tip portion) having the highest impurity concentration is cut to produce a relatively pure ruthenium. The first DSS can produce helium with impurities (eg, greater than 0.5 parts per million by weight required) and the second DSS pass 10 can produce less than 0.5 parts per million by weight of impurities. . More effective impurity control is required to provide a more pure enthalpy while reducing waste. Conductivity measurements on the tantalum ingot after the first DSS passes 8 and before cutting to remove impurities will substantially improve the tantalum yield. Similarly, resistivity measurements on the tantalum ingot after the second DSS passes 10 and before the second cut to remove impurities will substantially improve the tantalum yield of the final tantalum product. Figure 2 is a graph showing the actual measured concentration of 'selected elements' in different UMG stock batches, per million parts by weight. Note that there is a large change in elemental concentration throughout the different raw material batches. This variability is mainly made from the source material of UMG-Si raw materials (such as quartz and coal) 201144224. Small changes in impurity concentration can significantly affect the variability in the bottom to top resistivity of the ingot of each batch and the ingot yield (n part versus p part). Aluminum 4〇, Shipeng 42 and Lin 44 are the main elements to be controlled because they significantly affect the resistivity of the material. Fig. 3 is a graph showing the concentration profiles of the dopants Si Peng 50 and the can 52 in the UMG-Si ingot (in terms of the number of atoms per cubic centimeter of the solidification component). In Fig. 3, the initial concentration of boron 50 is 48 parts per million parts by weight and the initial concentration of filling 52 is 15 parts per million parts by weight. The change in concentration of boron and phosphorus along the solidification (or two degrees of ingot) reflects uneven segregation during directional solidification caused by element-specific segregation behavior. The uneven segregation of boron and phosphorus in the ingot causes a change in conductivity type at a height of about 80% of the ingot (from P type (boron 'aluminum) to !! type (phosphorus)). This is on the conductivity type. The change is shown by the B/P ratio 54 (in Figure 3, it is shown as the absolute value of the B_P difference). Therefore, the B/P ratio (such as B/P ratio 54) limits the yield of p-type materials. In the case where the UMG feedstock material has a relatively high aluminum concentration, the aluminum can also impact the yield by shifting the respective resistivity plots. Figure 4 is a graph 'which shows the 卩"^^ (the resistivity curve of the ingot (calculated resistivity 62 and measured resistance. 60).) The resistivity is in ohm-cm. To measure, and the height of the block is measured as a percentage of the bottom to the top (converted to a solidified component of grams). The resistivity is derived from the material, which is the absolute difference between the boron and phosphorus concentrations (in Figure 3 Abs^p) 54i shows)) measurement. Note that the resistivity graph reflects the conductivity pattern of boron and disc segregation in the ingot at about 80% of the ingot height. Figure 3 is 25 12 5 > 201144224 Figure 5 shows a cross-sectional image of the UMG-Si ingot after directional solidification. The impurity line 70 reflects the measurement in a typical ingot based on an A1-rich UMG-Si material. The conductivity pattern is changed. In the cross-sectional image of the ingot, a strong change in the ingot yield line (as indicated by the impurity line 7〇) is observed, which indicates that the ingot is produced on the left ingot side 72. The rate is close to 90% and the ingot yield on the right ingot side 74 is close to 60%. The large yield change across the ingot reflects solidification. During the non-uniform thermal state of the ingot, this results in a heterogeneous segregation of the doping elements B, P and A1. Directional solidification typically concentrates the impurities at the top of the ingot and then removes the top layer with most of the impurities. A relatively pure bottom layer is left for further processing. As shown in Figure 5, layer 78 has less impurities than layer 77. However, UMG-Si ingots have few flat, planar impurity curves after directional solidification. The flat cut line 76 shows a flat cut line that would typically be used to remove impurities concentrated in the top end of the ingot. However, the flat cut does not utilize the uneven and inconsistent distribution of the impurities in the material (by the impurity line 7〇 It shows that 'there is a UMG-Si process that results in poor performance and waste. Figure 6 shows a face image of a UMG-Si ingot after directional solidification with a cut line generated according to the disclosed subject matter. Impurity (such as boron, phosphorus and aluminum) have doping activity in the crucible and affect the resistivity of the ingot. Resistivity measurement provides an accurate measure of where to remove the contaminated portion of the ingot to reduce the overall ingot The absolute concentration of dopants and metal impurities. The lowest impurity concentration is found in the cold zone 80 (the zone where it first solidifies). The highest impurity concentration is found in the hot zone 82 (the final solidified zone). Segregation of impurities It is concentrated in the final state of the cast 13 201144224 block which is solidified during the directional solidification. This causes the impurity profile in the ingot to vary from region to region. ί's idea is in the scale bricks and the tracing 94 The degree of impurities is different. The ingot has been cut into a monument to control the removal of impurities by customizing the cutting line of each brick. The ortho-blocks 86, 88, 9〇, 92 have been cut after directional solidification. 94. Cutting _4 reflects the boundary line on the image (4). After the monument has been cut, the resistivity curve of the ingot is generated on the graph or 3D resistivity map by measuring the resistivity of the ingot from the bottom to the top and plotting (10) the calculated values. The measurement of the frequency of electrical rotation can also occur before the prayer block has been _ bricked. Furthermore, the brick cut is based on a number of factors, including (but not limited to, the size of the ingot, the impurity concentration of the block, the accuracy required to obtain (4) (4), and the manufacturing efficiency requirements. In Figure 6 The impurity line reflects the impurity concentration in the ingot at the threshold demand level. The standard cut dragon 4 flat cut line attempts to balance the impurity removal and the tantalum material yield. The controlled cut shows the resistivity curve according to the brick. The drawing defines a cutting line for each brick. The controlled cutting line defines the calculated cutting line according to the resistivity curve of the crucible (4), so only the part (4) containing the concentrated_quality is removed, and the tree is protected. Material yield. This allows for the most ideal removal of impurities without sacrificing the available stone eve. This cutting is calculated by measuring the resistivity of each brick from top to bottom. Conventional standard cutting is not performed according to the disclosure. The controlled cutting of the method, which leaves a lot of impurities in the ingot (such as in the ingot region of the brick 94), so that it needs to be subjected to another:: under-directional solidification for further purification originating from this casting 14 201144224 Figure 7 is a graphical illustration of the 3D solidification interface of the ingot casting block. Because it is difficult to control solidification, the solid-liquid interface during the ingot crystallization is not planar and produces a non-uniform segregation layer, as shown in the In Fig. 7, the impurities are concentrated in the top of the ingot after directional solidification. However, the solidified layers 90, 92 and 94 are obviously uneven. It is clear that the 5 kel solidified layer is not flat but vertically upward in the ingot. It varies downwards and has different thicknesses throughout the ingot. This causes the impurity profile to vary from region to region in the ingot, which results in an uneven grain profile. This uneven solidified layer makes it difficult to remove the concentration easily and efficiently. Impurities without sacrificing high yields or leaving too much impurities in the ingot. Figure 8 is a graph showing the concentration profiles of dopants boron 100, phosphorus 102 and aluminum 106 in UMG material ingots. (Percentage of ingot per cubic centimeter of ingot to mass converted to solidified component grams) In Figure 8, the initial concentration of boron is 0.411 ppmw, the initial concentration of phosphorus is 1.3 ppmw, and the initial concentration of #吕 is 23_08 ppmw. The different segregation coefficients of Si Peng, Phosphorus and Aluminum during directional solidification have a conductivity type change at an ingot height of about 87°. This change is based on the absolute concentration of boron and phosphorus plus the concentration of aluminum (abs). (B-P+Al), shown as 1〇4 on Fig. 8) reflects 'and its definition of the limit of P-type material yield. Figure 9 is the UMG-Si cast described in Figure 8. The cross-sectional image of the aluminum concentration curve of the block. Again, since it is difficult to control the thermal field during the directional solidification process, the crystal layer becomes uneven, resulting in a non-uniform impurity concentration profile. The aluminum concentration increases at the top of the ingot and is cast throughout Block cross-section variations, as indicated by impurity line 110. This makes it difficult to efficiently remove aluminum and other impurities through the entire ingot. 15 201144224 Figure 10 depicts the concentration of UMG-Si ingots in Figure 8. A cross-sectional image of the graph. The scaly concentration increases at the top of the block and varies across the cross section of the block, as indicated by the miscellaneous line 112. The concentration of the lining is significantly souther in some parts of the prayer block, making it difficult for the larva to remove the phosphorus impurities from a flat cut line along the overall block. Fig. 11 is a cross-sectional view showing the boron concentration profile of the UMG-Si ingot in Fig. 8. The pen concentration increases at the top of the scale and varies throughout the section of the ingot as shown by the impurity line 114. The rot concentration is significantly more pronounced in some parts of the ingot than in the south. It is difficult to remove scale impurities from a flat cutting line along the overall prayer block. Figure 12 is a process flow showing a side view of the solidification of the tantalum material in a biaxial solidification furnace. A bidirectional solidification furnace, a type of solidification furnace comprising a top end and a side heater, is often arranged with a heater to heat the top of the ingot and a plurality of heaters to heat the sides of the ingot, which are at the top of the resulting ingot block and Concentrate impurities on one side. The bidirectional solidification system of Fig. 12 uses the top heater 122 and the side heaters 120 and 124 to concentrate the impurities at the top end of the ingot at the top end heater 122 and at the ingot side of the placement side heater 120. . Liquid helium contains concentrated impurities and is also known as a contaminated area. At a furnace temperature of 1500 ° C, it is a complete liquid. In step 126, the furnace temperature is lowered to 145 〇t, and the melted portion is solidified, and a solidified layer is formed at the bottom of the ingot below the Weirong. Recently, the crucible of the top heater 122 is still molten while the helium crystals and impurities away from the top end heater 122 are concentrated in the melting crucible. During the step of ^, the side heater 120 and the side heater 124 are set at the same temperature, and when solidified, a vertical gradient is formed which is the same as the horizontal solidification gradient. 16 201144224 In step 128, most of the crystallization is carried out at a furnace temperature of 1420 ° C, and only the regions of the top heater 122 and the side heater 120 are melted, and the remaining sand has crystallized. The side heater 124 and the top heater 122 are cooled, which allows the enthalpy of the nearmost side heater 124 and the top end heater 122 to crystallize, and the molten crucible moves closest to the side heater 120. The impurities are concentrated in the remaining liquid helium in the top corner of the side heater 120 which is hot recently. Therefore, the impurities are concentrated in the molten region closest to the tip heater 122 and the side heater 120. This is the area to be removed to purify the fully crystalline tantalum ingot. The bidirectional condensing furnace can be equipped with five holes (one at the center and four in the corner) on the top to control and measure the height of the solidified crotch portion (often using a simple stone rod). In step 130, at the furnace temperature 140 (TC, the cooling side heater 12 and the crucible ingot are completely solidified. The impurities are concentrated in the crystal region closest to the top heater 122 and the side heater 120. Now, ready The ingot is divided into monuments and impurities are removed. The bidirectional solidification furnace uses a hot zone close to the heater to concentrate the impurities so as to be efficiently removed after the crucible has completely crystallized. In the process, when the crucible is melted in the ingot When it begins to solidify, it produces a vertical stone eve. It is cut in the cast part of the towel (four), its solidification and miscellaneous (butter, fill and !g) move into the remaining cool *** towel ^ in solid / liquid The interface reaches an area that changes the conductivity pattern excessively (usually within the coagulation of 8〇% of the town block: the temperature of the M'j'S circumference of the full-side heater to produce a horizontal Shiga solidification gradient, which will leave the remaining dazzling The edge of the ingot (the side of the hotter edge heater). Figure 13 shows the process flow, which shows the top view of the solidification of the stone material in a bidirectional solidification furnace (without display top heater). Side plus 17 201144224 heater 132 and side heater 34 to produce a horizontal helium solidification gradient and the nearest side heater 132 to concentrate impurities. Initially, at a furnace temperature of 15 〇〇 β (:, all of the ruthenium in the crucible is melted. In step 丨 36, the furnace temperature is adjusted to 145 The crucible c and the molten crucible begin to solidify at the bottom of the crucible (see Figure 12, side view of crucible solidification in a bidirectional solidification furnace). 'The simultaneously molten crucible moves to the nearest top heater. In step 138 At a furnace temperature of 1420 ° C, the side heater 132 is heated and the side heater 134 is cooled, which produces a horizontal helium solidification gradient. When the crucible of the heater 134 is cooled and solidified recently, the molten crucible moves to the nearest side. The impurities are collected in the molten stone of the nearest side heater 132. When the furnace temperature is lowered to 1400 ° C in step 140, the remaining molten helium (with the degree of concentrated impurities) solidifies 'and the impurities are trapped in the nearest In the ingot region of the side heater 132. Figure 14 is a graphical illustration of the 3d solidification interface of the Shixi ingot produced in the bidirectional solidification furnace. It is shown that the solid-liquid interface is substantially during the ingot crystallization. Keep flat on Thus, a substantially flat and planar solidified layer is formed. Therefore, the impurity profile from the top to the bottom is substantially the same for any region of the tantalum ingot. The solidified layers 150, 152, and 154 are throughout the ingot plane, unlike in the seventh Layers 90, 92 and 94 in the Figure. Further, as shown from the top plan view, the contaminated solidified layer has been further concentrated on side 156 by use of a bidirectional solidification oven (such as shown in Figure 13). The formation permits the concentration of impurities to be concentrated in an area that can be easily cut according to the disclosed method. The bidirectional solidification furnace is preferably operated using a crucible having a rectangular, non-secondary cross section, thereby facing a smaller financial disaster. The side heater ° 18 201144224 Figure 15 is a graph showing resistivity curves of various impurity concentrations (patterned in ohm-cm to solidified component grams) and cutting lines. The resistivity plot is strongly dependent on the impurity concentration. This allows the determination of the impurity concentration at each point on the resistivity graph. The cutting lines 166, 168 and 170 are dependent on the resistivity profile of the ingot. The cutting line can be determined based on the impurity concentration of the stone that is allowed in the final product. The graph of resistivity of the prayer block 160 has a shed concentration of ppm·45 ppmw, a dish yield of 1.59 ppmw, and an aluminum concentration of 0.087 ppmw. The cut line 166 corresponds to the resistivity curve 160 and the resistive curve 160 produces a controlled cut line that corrects the impurity concentration. The ingot resistivity curve 162 has a shed concentration of 0.45 ppmw, a dish concentration of 1.45 ppmw, and an aluminum concentration of 〇_〇79 ppmw. The cut line 168 corresponds to the resistivity curve 162 and the resistive rate graph 162 produces a controlled cut line that corrects the amount of impurity interposed. The ingot resistivity curve 164 has a boron concentration of 0.45 ppmw, a dish concentration of 1.59 ppmw, and an aluminum concentration of 119.119 ppmw. The cutting line 17〇 corresponds to the resistivity curve 164 and the resistivity curve 164 produces a controlled cutting line that produces a large amount of impurity concentration. Figures 16-18 are a graph ' which shows the relationship between the resistivity curve of the scale and the impurity concentration profile of the same ingot. The controlled cutting line can be calculated based on the desired threshold concentration of the particular impurity. Figures 16-18 show a cut line based on an aluminum concentration of 0.5 ppmw' but the cut line can be based on a different amount of impurities (such as 'deleted or filled) at any concentration. Figure 16 illustrates the calculation of the resistivity curve from the same Shixi ingot and the cutting line of the impurity concentration 19 201144224 degree graph. The upper graph shows the resistivity curve of the Shixi ingot with boron concentration 〇_45 Ppmw, squama concentration I·45 ppmw, and aluminum concentration 〇.〇79 ppmw (in ohm-cm to solidification component percentage). The lower graph shows the concentration profiles of boron 186, phosphorus 184, and aluminum 188 for the same ingot (percentage of solids per cubic centimeter). The cutting line 180 has been calculated for an aluminum concentration of 5.5 ppmw at a scale height of 84.5%. It means that the radon block has a concentration below 0.5 ppmw below the cutting line 180, and the ingot has an aluminum concentration above 0.5 ppmw on the cut line 180. Figure 17 illustrates the resistivity from the same tantalum ingot. The calculation of the cutting line of the graph and the impurity concentration graph. The upper graph shows the resistivity curve of the tantalum ingot with boron concentration 〇·45 PPmw, phosphorus concentration 1.45 ppmw, and aluminum concentration 0.117 ppmw. Figure 2〇2 (in ohm-cm to solidification component percentage). The lower graph shows the concentration profiles of boron 2〇8, phosphorus 204, and aluminum 206 for the same ingot (percentage of solids per cubic centimeter). The cutting line 2〇〇 has been calculated for the intrinsic concentration 〇.5 ppmw at 77% of the ingot height. It means that the 3 Xuanzhen block has a concentration lower than 0.5 ppmw when it is lower than the cutting line 200 and the casting has a concentration higher than 〇.5 ppmw on the cutting line 200. Figure 18 illustrates the calculation of the resistivity plot from the same Shixi ingot and the cut line of the impurity concentration plot. The upper graph shows the resistivity plot 224 (in ohm-cm versus solids fraction) with a boron concentration of 0.45 ppmw, a fill concentration of 1.8 ppmw, and an aluminum concentration of 0_079 ppmw. The lower graph shows the concentration profiles of boron 228, phosphorus 226, and aluminum 230 for the same ingot (percentage of solids per cubic centimeter). The cutting line 222 has been calculated for a concentration of ppm·5 ppmw at a height of 84.5% of the block. It means that the 20 201144224 ingot has an aluminum concentration below 0.5 ppmw below the cutting line 222 and the ingot has a concentration above the 〇·5 ppmw on the cutting line 222. The cutting line 220 has also been calculated at a height of 83% of the ingot from the resistivity curve at p/N (where the ingot moves from P to n). This cut line reflects the most ideal cut line' to preserve the highest p-type tantalum material yield from the ingot. Figure 19 is a graph showing the resistivity of the tantalum ingot in Figures 16-18 (in ohm-cm versus solids fraction). The resistivity graph 182 shows the resistivity of the ingot in Fig. 16, and the cut line 180 is calculated at an ingot height of 0.8.5% at the ingot height of 84.5%. The resistivity graph 102 shows the resistivity of the ingot in Fig. 17, and the cut line 2 is calculated at the ingot height of 0.5 ppmw at the ingot height of 77% (the resistivity curve 224 is shown at the 18th). The resistivity of the ingot in the figure and the cut line 220 calculated at the P/N transition at the ingot height of 83.5%. Fig. 20 shows the boron of the resistivity curves 182, 202 and 224 in Fig. 19. Filled with the corresponding concentration curve of aluminum. Figures 21 to 33 are for a control process and a method for evaluating the quality of UMG-Si raw materials. By analyzing the ingots obtained from the batch of multiple UMG-Si raw materials. The resistivity plot of the test sample determines the boron and phosphorus content of those batches (thus determining the quality of the UMG-Si material). Furthermore, other impurities such as, but not limited to, SiC can be detected. Figure 21 shows a prior art process flow 'Description of boron, fill and aluminum content in a crucible according to an inductively coupled plasma mass spectrometer (ICPMS) method. In step 210, a pure feedstock is selected (such as Quartz and coal) to make MG-Si with low boron content. Then, step 21 2 Reducing the aluminum content by refining the gas into a 21 201144224. Additionally, for example, further boron content can be reduced in a furnace containing an oxy-fuel burner, ultimately resulting in UMG-Si. Then, in order to further reduce impurities (such as boron) , phosphorus and aluminum), applying ICPMS for chemical analysis of UMG-Si (shown as step 214). If the analysis provides a boron concentration less than the specified threshold concentration (shown as 1 ppmw), the material is considered Ready for crystallization and shipping for casting ingots' shows as the final UMG-Si product 216. However, if the boron concentration is measured to be greater than the specified threshold concentration (shown as 1 ppmw), then The refining process is repeated until the material is suitable for a UMG-Si product with a minimum boron threshold concentration. It is important that other threshold concentrations of other impurities, such as phosphorus, be used for boron. More efficient impurity control is required to provide The purer Shi Xi, while reducing waste. The disclosed subject matter provides an alternative to the above chemical analysis (ICPMS), and can alternatively introduce another to control 1;] ^ (3_ by raw material quality Process and method. Analysis of the disclosed control method 11] ^ (3_& test ingot before the release of the material resistivity curve. This control method uses a reasonably large test 51 ingot (from the raw material batch to be controlled) The electrical data obtained. In particular, the measurement of the resistivity curve from the bottom of the casting ghost to the top releases the UMG_Si raw material batch as a criterion for the product. As for the method of the uncovered method, Provides a method for simultaneously testing a plurality of ingots, and estimates that the alpha heating and the joyful have a pure domain (which can & top and bottom)

The second s has a top heater such as the bidirectional solidified TT J crystal grower shown in Figures 12 and 13. ,,,. Crystal grower. Therefore, the modified Xiaotian A^ with N times N turns can grow and test NXN test ingots. This procedure further ... a method of raw material quality. Step 22 201144224 Figures 22 and 23 - Curves®, which show the actual measured resistivity for different _MG__ batches. Note that the large UMG-Si raw material culture is seen from the batch in the figure to the resistivity (and thus the yield) of the batch in Fig. 23. The graphs of Figures 22 and 23 show the resistivity plots of the two ingots grown from two batches (from the same material) (in ohm-cm pairs from the bottom to the top of the ingot). Note that the 'resistivity curve' reflects the change in conductivity pattern caused by the segregation characteristics of the butterfly and phosphorus in the ingot. In Fig. 22, the 15 mm high ingot has a p/N change at about 75 mm (the ingot moves from the P-type to the η-type point) 'remaining about 45% yield of pSUMG_Si (by reference numeral m) Explicit (4). Table 219 provides the resistivity data for the batch in the resistivity graph of the 帛22 diagram, including the average, median, minimum and maximum resistivity values (in ohm-cm). 15 mm The ingot in the first magic map has a P/N change at about 110 mm, leaving approximately 74% yield of P-type UMG-Si (shown by reference numeral 220) for use. Table 221 is provided at The resistivity data for the batches described in the resistivity plots for the graph, including the mean, median, minimum, and maximum resistivity values (in ohm-centimeter). This large variability is primarily due to the feed material. (such as, but not limited to, quartz and coal.) This disclosure suggests a method and process for controlling this variability before using this material to cast an industrial size ingot, and then use it outside the day of the prayer block. To make solar cells. Figure 24 shows four different kinds of resistance in the four-stack operation ( Material 1, Batch 2, Batch 3 and Batch 4) Example of simultaneous synchronous directional solidification operation. ICPMS data of B (boron) and P (phosphorus) of the test block. Table 224 shows the resistance i and the batch. The measured boron and phosphorus concentrations of material 2. Table 226 shows the measured boron and phosphorus concentrations for the resist 3 and the resist 23 201144224 4. The corresponding resistivity curves are shown in the figure (batch 1), 26 Figure (batch 2), 27 (batch 3) and 28 (batch 4). Here, the resistivity data is not consistent with the ICPMS-based expectations. For example, 'based on boron and phosphorus measurements The values, which are expected to be batched as shown in Figure ) and Batch 3 (as shown in Figure 27), have similar resistivity plots. The measured resistivity plots for each batch allow for practical evaluation of the raw material and not for chemical analysis. At the same time, it is also possible to modify the individual raw material batches according to the resistivity curve to determine the possible amount of common impurities, thereby ensuring high p-type yield and useful resistivity range after co-doping. Fig. 25 is a graph showing the measured resistivity data of the resisting 在 in Fig. 24 with respect to the height of the ingot from the bottom to the top in resistance (in ohm-cm) (resistivity graph 230) ). Batch 1 was changed at about 12 mm, leaving about 73% yield of UMG-Si (shown by reference numeral 234). Table 232 provides the resistivity data for Batch 1, including average, median, and most Maximum resistivity value in ohm-cm. Figure 26 is a graph showing the amount of the material of the batch 2 in Figure 24 from the height of the ingot to the top of the ingot (in ohm-cm). Graph 236). Batch 2 was p/N changed at about 45 mm, leaving about 26% yield of UMG-Si (shown by reference numeral 240). Table 238 provides the resistivity data for Batch 2, including the average, median, and maximum resistivity values in ohm-cm. Figure 27 is a graph of 'resistance (in ohm-cm) pairs, and the height of the ingot to the top shows the measurement of batch 3 in Figure 24 201144224 Resistivity data (resistivity) Graph 240). Batch 3 was changed at about 50 mm to leave about 28% yield of UMG-Si (shown by reference numeral 246). Table 242 provides resistivity data for Batch 3, including average, median, maximum and maximum. Resistivity value in ohm-cents. Figure 28 is a graph of 'the resistance (in ohms-cm,) oblique "_ to the top of the prayer block south shows the batch 4 + " _ μ in the 24th figure! The amount of resistivity data (resistivity curve 248). Batch 4 is at about 70 mcal rpm, leaving about 41% yield of UMG-Si (shown by reference numeral 252) 250 to provide resistivity data for batch 4, including average, medium to maximum resistivity values ( In ohm-cm). Figure 29 is a photograph of the read block from the bottom to the top. The present invention describes a method for controlling the quality of UMG-Si raw materials. Field/hunting is to produce a small test ingot from the raw material of each batch, and then measure the resistivity from the bottom to the top. This method is capable of measuring the growth conditions of the ingot. A typical example produces a 450 kg rust block and specifies the growth conditions used for the addition of yield and resistivity control. However, this disclosure makes other such growth conditions possible. Figure 30 graphically depicts a specific example of a crystal grower_configuration in accordance with the disclosed subject matter. In order to control the raw material f, this: measuring the raw material material and phosphorus concentration, the use of which can be suitable for W type assembly (such as shown in Figure 30) (4) crystal growth (four) step according to directional solidification method) test Ingot, measure the resistivity curve from bottom to top. Crystallizer form 262 has a 2x2 design that can grow up to 4 test ingots during the same run. Crystal grower form 264 has 25 201144224

3X3 hanging net form, which can grow up to 9 test blocks during the same run. The knot B bucket E gauge form 266 has a 4x4J# crash form. It can grow up to 16 test ingots in the same month. The crystallizer form 268 has a 6X6 坩埚 form that can be grown up to 36 test casts during the same run or other 坩埚 forms, such as larger configurations (such as 7x7) or long, such as 2x3, 3x2, 3x4 or 4x3) or any change thereof. In a typical example, the weight of the test ingot can be grown in the range of ^ ▲ and the δ Hai method from the different raw materials during the same operation. This scale test proves that the county material (four) and this The method is 60001 = levy. Typically, the raw material batch can range from 2000 kilograms to the other, which can be supplemented by the side SiC contaminants in the raw material. This can be done by using the "inclusions" in the ingot (from this original rod) using 1R surface speculation. The hazard of the batch containing the raw material in the household stone map may be covered by a lid made of high purity stone mortar or other such material, which is mutually polluted. The configuration shown in the 篦 丄 丄 brother's, ., 〇 〇 〇 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步 同步In another embodiment, other configurations of the shape, such as a round shape. The first call-photograph shows that it shows a modified solid-running single _ hot group stationed in four (mother 10,000 (1) raw condensate 7 ΛΡ, 坩埚 heat configuration. Early 坩埚 crystal growth benefit 270 has been modified into four knots day growth 272 Therefore, for each direction, with reference - 26 201144224 In this specific example, the individual ingot sizes allow the manufacture of six (6) england solar cells. The method of the present disclosure allows for a fast and reliable B/P ratio. Control. The specific example shown in Figure 31 can be scaled up, for example, to thirty-six (36) ingots per run, corresponding to the raw material size of the UMG-Si stock material 5 〇 MT. The relatives of those tested. The small size allows for the good control of crystallinity using industrial grade crystallization furnaces with specially designed components to ensure heat and gas symmetry. The B/P ratio for each batch of raw materials is confirmed using a one-block test, followed by further analysis. (such as detection of SiC inclusions).

Figure 33 is a process flow showing the disclosed UMG-Si control Figure 32 is a photographic example showing the tantalum carbide (SiC) impurities actually observed in actual UMG-Si ingots. Note the degree of change in SiC content in ingot 276, ingot 278, and ingot 280. Sic inclusions can be measured by infrared imaging (IR) methods. Depending on the process conditions at the location of the raw material supplier, the SiC content can often be formed from one batch to the other. Since the raw material containing the SiC inclusions produces ingots containing inclusions, a better control method at the location of the raw material supplier allows for (9) no inclusion at the location of the user (such as a solar cell manufacturer). Scales of things. In a specific example, the main steps of a specific example of a process that allows for the reliable management of raw material materials. In step 290, selected for

Analytical raw material γ.,.., one (2) to six (6) ΜΤ 27 201144224 The size of the raw material batch should be greater than 2χ1 (Τ3. This ratio is greater than the chemical analysis based on UMG-S i raw materials. Current implementation 3 - 4 series large / J \ 测试 In step 292, test batch crystallization occurs. The size and shape of the crucible typically allows the ingot to produce wafers on a 156 mm x 56 mm level. Because the disclosed method analyzes a plurality of batches in a single run, it should exhibit symmetrical heat and gas flow conditions for all of the helium and test batches. In step 294, the resistivity curves of the individual ingots are measured. The boron and phosphorus concentrations can be determined from this resistivity measurement. The optional step 296 measurement can be used to increase the ingot yield and produce an appropriate resistivity profile (based on the amount of boron and phosphorus in the raw material batch being analyzed). a co-dopant. The optional step 298 measures the SiC content in the test ingot via IR analysis. And the optional step 300 produces a test wafer for extensive evaluation of the stock batch. Revealed subject matter Providing a quality control method for determining the impurity concentration in the UMG-Si raw material batch according to the resistivity curve of the test ingot (which is obtained from the UMG-Si of the batch). According to the directional solidification method Simultaneously growing a plurality of test ingots (each corresponding to the UMG-Si stock batch). Although the disclosed subject matter has been described in detail, it should be understood that a variety of variations, substitutions, and changes can be made without leaving The present invention is as defined by the scope of the appended claims. C-Simple Description 3 28 201144224 The first (previously described) system - used to reduce the process of stone, dish and inscription in Shi Xizhong Process; Figure 2 shows the actual measured impurity of different UMG raw materials; Figure 3 shows the concentration of boron and phosphorus in the block; Figure is a graph showing the resistivity curve of the UMG-Si ingot measured in Figure 3 (calculated versus measured resistivity); Figure 5 shows the directional solidification of the UMG-Si ingot After wearing a face image; Figure 6 is a UM The cross-sectional image of the G-Si ingot after directional solidification and the cutting line generated according to the disclosed subject matter; Figure 7 is a graphical illustration of the 3D solidification interface of the ingot ingot; Figure 8 is a graph, which is a graph A graph showing the concentration of boron, phosphorus and aluminum in the UMG-Si ingot; Fig. 9 is a cross-sectional image of the aluminum concentration of the UMG-Si ingot described in Fig. 8; A cross-sectional image of the phosphorus concentration of the UMG-Si ingot described in Figure 8; Figure U is a cross-sectional image of the boron concentration of the UMG-Si ingot described in Figure 8; Figure 12 - Process flow , which shows a side view of the tantalum material solidified in the bidirectional solidification furnace; Fig. 13 is a process flow diagram showing a top view of the tantalum material solidified in the bidirectional solidification furnace 29 201144224; Graphical illustration of the solidification interface of the 31) cast ingot produced in a biaxial solidification furnace; Figure 15 is a graph showing resistivity curves and cut lines of various impurity concentrations; Figure is a graph illustrating the resistivity plot and impurity concentration of the tantalum ingot The relationship between the graphs; Fig. 19 is a graph showing the resistivity curve of the block in the 16th-18th image (in ohm-cm to solidified component); Figure 20 is shown in Figure 19. Corresponding impurity concentration curve of the resistivity graph in the middle; Figure 21 shows a schematic diagram of the process flow for reducing boron, filling and aluminum in the prior art; Figures 22 and 23 are a graph showing different The actual measured resistivity of the UMG Si raw material batch; Figure 24 shows the ICPMS data for the B and P of the sample test ingots for simultaneous directional solidification; Figure 25 is a graph showing the same in Figure 24 The resistivity data of the batch 1 is measured; Fig. 26 is a graph showing the measured resistivity data of the batch 2 in Fig. 24; Fig. 27 is a graph showing The resistivity data of the batch 3 measured in Fig. 24; Fig. 28 is a graph showing the resistivity data of the batch No. 4 201144224 in Fig. 24; Fig. 29 Photograph of the cast ingot; Figure 30 is drawn graphically according to the disclosed A specific example of the crystal grower configuration of the event; Figure 31 is a photo representation showing a modification, each configuration of the thermal configuration of the thermal configuration to four (4) 坩埚; Figure 32 is a photo The examples, which show the impurities actually detected in the actual UMG-Si ingot; and the 33rd, a process flow, show the main steps of a specific example of the disclosed UMG-Si control method. [Main component symbol description] 2...Step 72...Left ingot block edge 4...Step 74...Right ingot block edge 6...The first DSS passes through 76...the flat cutting line 8: The second DSS passes through 77...layer 40...aluminum 78...layer 42...boron 80...cold zone 44...phosphorus 82...hot zone 50".boron 84...cut line 52. "Phosphorus 86...Ingot brick 54...B/P ratio 88...Ingot brick 60...Measured resistivity 90...cast block, solidified layer 62...calculated resistance Rate 92... ingot brick, solidified layer 70... impurity line 94... ingot brick, solidified layer 31 201144224 100··· boron 102". phosphorus 104.. .abs(B-P+Al)

106.. . IS 110.. . impurity line 112.. impurity line 114.. impurity line 120.. side heater 122.. top heater 124.. side heater 126.. . Step 130...Step 132.. Side Heater 134.. Side Heater 136...Step 138. Step 140.. Step 150.. Solidified Layer 152.. Solidified Layer 154.. Solidified layer 156.. Side 160.. Resistivity curve 162.. Resistivity curve 164.. Resistivity curve 166.. Cutting line 168... Cutting line 170... Cutting line 180.. Tangent 182.. . Resistivity curve 184.. Phosphorus 186.. Boron 188.. . Aluminum 200.. . Cutting line 202.. . Resistivity curve 204.. . Phosphorus 206. " Aluminum 208... Boron 210...Step 212.. Step 214...Step 216.. Product 218.. Cut Line 219... Table 220.. Cut Line 221... Table 222.. Cut Line 224.. Resistivity Graph, Table 32 201144224 226.. Table, phosphorus 228... boron 230.. aluminum, resistivity curve 232.. Table 234.. cutting line 236.. resistivity curve 238.. , resistivity curve 242.. Table 246.. cutting line 248.. . resistivity curve 250.. . Table 252.. cutting 262.. Crystallizer format 264.. Crystallizer format 266.. Crystallizer format 268.. Crystallizer format 270... Single crystallizer 272.. Four crystal growth 274. . . four test ingots 276.. · ingots 278.. ingots 280.. ingots 290-300...Step 33

Claims (1)

201144224 VII 1. 2. 3. 4. 5. 6. Patent application scope: Kind to use. A method for controlling the quality of UMG-Si raw materials, the method comprising the steps of: introducing directional solidification of molten uMG_Si from a batch of UMG-Si raw materials to form a test ingot; measuring the ingot of the test ingot The top to bottom resistivity; plot the resistivity of the tantalum test ingot; calculate the phosphorus and boron concentrations of the UMG-Si stock batch based on the resistivity plot of the tantalum test ingot. The method of calculating the range of the selected UM G_si raw material batch and the boron concentration according to the resistivity curve of the financial test scale further includes, according to the test of the ingot from the crucible The step of determining the phosphorus and boron concentrations of the selected UMG-Si raw material batch is determined by the yield of the test ingot as determined by the resistivity plot. The method of claim </ RTI> further includes the step of measuring the co-dopant, wherein the sho is used to increase the yield of the lumps and the concentration of boron and phosphorus according to the selected UMG-Si raw material batch. Generate an appropriate resistivity plot. The method of claim 1, further comprising the step of measuring the content of S i C in the ingot by the Han imaging analysis. The method of claim 3, further comprising the step of manufacturing a test wafer from the crucible test ingot. The method of claim 1, wherein the weight ratio of the niobium test ingot to the UMG-Si raw material batch is greater than 2Χι〇_3. 34 201144224 7. The method of claim 1, wherein the crucible is tested in an ingot weight of about 15 kg. 8. The method of claim 1, wherein the step of directional solidification uses a bidirectional solidification furnace for concentrating impurities on the top and one side of the crucible test ingot. 9. A method for controlling the quality of UMG-Si raw materials, the method comprising the steps of: UMG-Si synchronous directional solidification in a single crystal grower from a plurality of UMG-Si raw material batches Forming a plurality of flaw test ingots each of which corresponds to a particular UMG-Si stock batch; measuring the resistivity of each of the tantalum test ingots from top to bottom; drawing The enthalpy tests each resistivity profile of the ingot; the resistivity curves of each of the corresponding ingots are tested to calculate the phosphorus and boron concentrations for each of the UMG-Si stock batches. The method of claim 9, wherein the simultaneous directional solidification of the molten UMG-Si is performed from a plurality of UMG-Si raw material batches in a single crystal grower to form a plurality of ruthenium test ingots. And the step of wherein the plurality of ruthenium test ingots each correspond to a special UMG-Si raw material batch, further comprising melting from a plurality of UMG_Si raw material batches in a single multi-turn, and a helium grower The simultaneous directional solidification of the UMG-Si to form a plurality of test ingots, wherein each of the four (4) test casting blocks corresponds to a special UMG-Si raw material batch. The method of the sigma π patent fegl item 9 'in which the phosphorus and boron concentrations of each UMG-S i raw material batch are further calculated according to the resistivity curve of each of the corresponding ingots 35 201144224 ingots Including, according to the yield curves of each of the ingots of the ingot test, each of the selected UMG-S i materials is used to calculate the yield of each of the selected UMG-S i materials. And the step of boron concentration. 12. The method of claim 9 further comprising the step of measuring a co-dopant, wherein the co-dopant is used to increase the ingot yield and the boron and phosphorus according to the selected UMG-Si raw material batch. The concentration produces an appropriate resistivity plot. 13. The method of claim 9, further comprising the step of measuring the content of S i C in the ingot by the imaging analysis. M. The method of claim 9, further comprising the step of manufacturing a test wafer from each of the test pieces. 15. The method of claim 9, wherein the weight ratio of each of the bismuth test ingots to each of the UMG-Si raw material batches is greater than 2 χ 1 (Τ3. 16. The method of claim 9, wherein the weight of each of the ingots is about 15 kg. 17. The method of claim 9, wherein the step of directional solidification uses a bidirectional solidification furnace, The crucibles are tested for concentrating impurities on the top and side of each of the ingots. 18. The method of claim 9, wherein the plurality of UMG-Si raw material batches are smelted in a single crystal grower Simultaneous directional solidification of UMG-Si to form a plurality of ruthenium test ingots, wherein the composite 36 201144224 plurality of ruthenium test ingots each correspond to a particular UMG-Si raw material batch, further comprising from a plurality of UMG- Si raw material batch, synchronous directional solidification of molten UMG-Si in a single multi-turn crystallizer having an NxN坩埚 form to form a plurality of tantalum test ingots, wherein the plurality of tantalum test ingots are each Special UMG-S i original The batch is corresponding. 37