JP2006273668A - Method for production of semiconductor ingot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for production of a semiconductor ingot having less variation in specific resistance, uniform quality and excellent characteristics. <P>SOLUTION: The method for production of the semiconductor ingot comprises solidifying a semiconductor melt 107 containing a first dopant material 106 defining a first conductive type in a casting mold 103 in a prescribed direction. In that case, a second dopant material 109 which defines a second conductive type being contrary to the above first type is supplied into the semiconductor melt 107. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for manufacturing a semiconductor ingot, and more particularly to a technique for adjusting the specific resistance of a semiconductor ingot.

  At present, silicon ingots are often used as semiconductor substrates for solar cells, and further improvements in the quality of the solar cells are demanded for higher performance of solar cells.

  Among them, in order to adjust the specific resistance of the silicon ingot within a predetermined range, the dopant amount is calculated by a specific calculation formula, and a dopant material which is an III-valent or V-valent impurity element is added to the silicon melt. Is used. This dopant material is usually supplied in the form of granular silicon crystals containing an impurity element at a high concentration, and this is melted simultaneously with the silicon material to obtain a silicon melt having a desired dopant concentration.

  When such a silicon melt is solidified, there is a problem that the resistivity value varies in the crystal growth direction.

  That is, the dopant element in the silicon melt is taken into the silicon solid crystal at a certain ratio by the segregation phenomenon during the solidification growth. At this time, the dopant concentration (CS) in the crystal is expressed by CS = kCL (k: segregation coefficient, CL: dopant concentration in the melt). The dopant concentration that lowers the temperature increases from the bottom of the mold to the top. Accordingly, the specific resistance is highest at the bottom of the mold and decreases in the height direction. Therefore, there has been a problem that the desired specific resistance value is deviated and the portion cannot be used as a product.

  In order to solve this problem, the following method has been proposed as a specific resistance adjustment technique when single crystal silicon is grown by the CZ method.

For example, in the middle of pulling a single crystal from a silicon melt containing a first dopant material, a method of supplying a second dopant material exhibiting a conductivity type opposite to that of the first dopant material (for example, see Patent Document 1), or The specific resistance of the entire silicon ingot by the method of adding phosphorus of the second dopant material at a concentration of 25 to 35% with respect to the boron concentration of the first dopant material (for example, see Patent Document 2) to the initial silicon melt There are those that make the values uniform.
Japanese Patent Laid-Open No. 7-309963 JP 2003-137687 A

  However, in the method of supplying the second dopant material exhibiting the opposite conductivity type while pulling up the single crystal as in Patent Document 1, the tail portion of the pulled silicon single crystal is controlled depending on the timing of supplying the second dopant material. There is a problem that the specific resistance varies greatly. In particular, unlike the CZ method, in which the silicon solidification rate can be easily estimated from the pulling height, when the silicon melt is solidified from the bottom of the mold by the casting method, it is difficult to observe the solid-liquid interface from the outside. It was extremely difficult to control the supply timing.

  Further, in the method of adding the second dopant material in advance to the initial melt as in Patent Document 2, various conditions for controlling various quality characteristics of the silicon ingot (temperature profile, silicon pulling speed, and rotation of the seed crystal) Speed, etc.) is limited as a condition only to achieve uniform resistivity, and it is considered difficult to satisfy other conditions for improving other quality characteristics at the same time. From the viewpoint of homogenizing ingot quality Even when viewed, it was not a preferable method.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to propose a method for producing a semiconductor ingot having a small specific resistance variation, uniform quality and excellent characteristics in a method of manufacturing a semiconductor ingot by a casting method. To do.

  The method for producing a semiconductor ingot according to the present invention comprises solidifying a semiconductor melt containing a first dopant material defining a first conductivity type in a predetermined direction in a mold. A second dopant material defining a second conductivity type opposite to the conductivity type is supplied into the semiconductor melt.

  The second dopant material is formed on the inner surface of the mold above the surface of the semiconductor melt before the start of solidification, and is dissolved in the semiconductor melt whose liquid level has increased during the solidification process. And

  Furthermore, the liquid level of the semiconductor melt is measured in the solidification process, and the second dopant material is supplied into the semiconductor melt according to the measured value.

  The method for producing a semiconductor ingot according to the present invention comprises solidifying a semiconductor melt containing a first dopant material defining a first conductivity type in a predetermined direction in a mold. Since the second dopant material defining the second conductivity type opposite to the conductivity type is supplied into the semiconductor melt, the apparent segregation coefficient of the first dopant material as the main additive can be increased, so that the specific resistance A semiconductor ingot with small variations and uniform quality can be manufactured.

  In addition, the second dopant material is formed on the inner surface of the mold above the surface of the semiconductor melt before the start of solidification, and is dissolved in the semiconductor melt that has risen in the solidification process. Accordingly, it is possible to control the timing and amount of the second dopant material to be melted by a simple method in which the semiconductor melt is adjusted to be expanded by solidification and the formation position of the second dopant material is adjusted.

  Further, the liquid level of the semiconductor melt is measured in the solidification process, and the second dopant material is supplied into the semiconductor melt according to the measured value. The dopant material can be supplied, and the specific resistance variation can be more effectively suppressed.

  Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

<Manufacturing equipment>
FIG. 2 is a schematic cross-sectional view showing a configuration of a general silicon casting apparatus, and FIG. 3 is a schematic cross-sectional view of a solidified portion in the silicon casting apparatus. In the figure, 101a is a melting crucible, 101b is a holding crucible, 102 is a pouring spout, 103 is a mold, 104 is a heating means, 105 is a silicon material (semiconductor material), 106 is a first dopant material, 107 is a silicon melt (semiconductor) (Melt), 108 is a release material layer, 110 is a mold heat insulating material, 111 is a cooling means, and 112 is a mold heating means.

  As shown in FIG. 2, the melting crucible 101 a is for pouring a silicon melt, which is heated and melted while holding the first dopant material 106 and the silicon material 105, into the mold 103. A silicon ingot obtained by cooling and solidifying the silicon melt 107 melted in the melting crucible 101a and poured into the mold 103 is used, for example, as a polycrystalline silicon substrate material for solar cells. The first dopant material 106 may be the doping element itself, or may be a silicon material doped with a doping element at a high concentration in advance.

  As the melting crucible 101a, known high-purity quartz or the like can be used so that impurities are not contained in the molten silicon melt. However, melting, evaporation, softening, and deformation are performed at a temperature higher than the melting temperature of the silicon material. The purity is not particularly limited as long as it does not easily cause decomposition and does not deteriorate the solar cell characteristics. Further, since the melting crucible 101a is softened at a high temperature and cannot keep its shape, it is held by a holding crucible 101b made of graphite or the like. Further, the dimensions of the melting crucible 101a and the holding crucible 101b need to be a dimension capable of containing a silicon material corresponding to the melting amount to be melted at one time. The melting amount of the silicon material is, for example, in the range of 1 kg to 250 kg.

  A heating means 104 is disposed around the melting crucible 101a and the holding crucible 101b. The silicon material in the melting crucible 101a is heated and melted by the heating means 104 to obtain a silicon melt. As the heating means 104, for example, a resistance heating type heater or an induction heating type coil can be used.

  As shown in FIG. 3, the mold 103 disposed in the lower part of the melting crucible 101a and the holding crucible 101b has an open part opened upward, and the poured silicon melt is poured by this open part. In addition, the silicon melt is held in the interior and solidified in one direction from below to above.

  The mold 103 is made of graphite, silicon oxide, or the like, and is configured by an integrally molded mold or a divided mold that can be divided and assembled. Further, a release material layer 108 is provided on the inner surface of the mold 103. It is possible to prevent the inner wall of the mold 103 and the silicon ingot from being fused after the silicon melt 107 inside the mold 103 is solidified. As a material of the release material layer 108, for example, silicon nitride, silicon carbide, silicon oxide, or the like can be used. As a method for providing the release material layer 108, a known method of mixing the above-mentioned powder in a solution composed of an appropriate binder and a solvent and stirring it to form a slurry, which is coated by means such as coating or spraying. Use it.

A mold heat insulating material 110 is installed around the mold 103 to suppress heat removal from the mold side surface. A material such as carbon felt is generally used for the mold heat insulating material 110 in consideration of heat resistance, heat insulating properties, and the like. In addition, a cooling means 111 such as a cooling plate for circulating a coolant such as water is installed inside the metal plate in order to remove the poured silicon melt from below and cool and solidify it below the mold 103. May be. Further, a mold heating means 112 made of a carbon heater or the like capable of heating the mold 103 from above may be disposed, whereby the surface of the silicon melt 107 poured into the mold 103 is heated appropriately and from below to above. The directed temperature gradient can be controlled more accurately. At this time, the density of solid silicon is 2.33 g / cm ³ , whereas the density of liquid silicon is 2.54 g / cm ³ , and volume expansion of nearly 10% occurs when the liquid solidifies. By doing so, it is possible to solidify while allowing the volume that increases with solidification to escape upward. Therefore, the surface of the silicon melt rises as solidification proceeds.

  These casting apparatuses are preferably placed in a vacuum vessel (not shown) and performed in a reducing atmosphere such as an inert gas from the viewpoint of preventing contamination and oxidation of impurities.

<First embodiment>
Hereinafter, a first embodiment according to a method for manufacturing a semiconductor ingot of the present invention will be described with reference to FIG. The first dopant material 106 will be described as a P-type dopant (first conductivity type), and the second dopant material 109 having a second conductivity type opposite to the first dopant material 106 will be described as an N-type dopant.

  As shown in FIG. 1, on the inner surface of the mold 103 on which the release material layer 108 is formed, the second dopant material 109 is fixedly formed above the liquid surface of the semiconductor melt 107 before the start of solidification.

  As the second dopant material 109, for example, a slurry-like material prepared by mixing and stirring fine powders formed by pulverizing and washing N-type single crystal silicon top, tail, etc., with an appropriate organic binder. Use. The slurry-like second dopant material 109 is within an arbitrary range located from the vicinity of the surface of the silicon melt before solidification (solidification rate 0%) to the surface of the silicon melt when final solidification (coagulation rate 100%). Apply. In this way, by setting the application position of the N-type second dopant material 109 on the upper side of the silicon melt and further on the region above the surface of the silicon melt before solidification, the initial stage of solidification of the silicon melt Since the elution amount of the donor element from the second dopant material 109 is suppressed and the apparent segregation coefficient of the P-type element as the main additive does not change, the specific resistance value at the bottom of the silicon ingot deviates from a desired value. There is nothing. Since the silicon melt increases as the solidification of the silicon melt proceeds, the elution amount of the donor element from the N-type second dopant material 109 increases, and the apparent segregation coefficient of the P-type element as the main additive Therefore, it is possible to suppress a decrease in specific resistance, and it is possible to manufacture a polycrystalline silicon ingot having a uniform quality with small variations in specific resistance.

  The position where the second dopant material 109 is fixed is located from the position above the silicon melt surface before solidification (solidification rate 0%) to the surface of the silicon melt when final solidification (solidification rate 100%). It is desirable to be within an arbitrary range. Thus, if the second dopant material 109 is fixed at a position above the surface of the silicon melt before solidification (solidification rate 0%), the silicon melt is solidified and expanded, and the surface of the silicon melt rises. The donor element elutes from the dopant material 113 into the silicon melt before reaching the fixing position of the dopant material, and the specific resistance value at the bottom of the silicon ingot is reliably prevented from deviating from a desired value. Can do.

  Furthermore, it is more desirable if it is within an arbitrary range located from the surface of the silicon melt when the solidification rate of the silicon melt is 50% to the surface of the silicon melt when the solidification rate is 100%. Can produce polycrystalline silicon ingots with small variations in quality and uniform quality. Furthermore, there are no restrictions on the manufacturing conditions, and various temperature conditions can be changed freely to control the solidification rate. Therefore, it is possible to simultaneously satisfy various conditions for improving the solar cell characteristics.

  The N-type impurity concentration in the second dopant material 109 is desirably set to 20% or more and less than 40% of the P-type impurity concentration in the first dopant material 106. When the N-type impurity concentration is less than 20%, the effect of canceling the P-type impurity is reduced, and when the N-type impurity concentration is 40% or more, the specific resistance is greatly increased in the upper part of the ingot, or the conductivity type This is not preferable because reversal of this occurs.

  The specific resistance value of the polycrystalline silicon ingot manufactured by the above manufacturing method is such that the difference in specific resistance between the bottom and top of the silicon ingot is 0.2Ω · cm to 0.5Ω · cm. Even if the desired specific resistance range is narrow, the problem that the deviated portion cannot be used as a product without deviating from the desired range can be suppressed.

<Second Embodiment>
Hereinafter, a second embodiment according to a method for manufacturing a semiconductor ingot of the present invention will be described.

  First, a predetermined amount of silicon material 105 and an amount of the first dopant material 106 containing P-type impurities such as boron and gallium, which are adjusted so that the specific resistance at the bottom of the ingot becomes a desired value, are charged into the melting crucible 101a. Thereafter, the silicon material 105 in the melting crucible is melted by the heating means 104 to obtain a silicon melt 107 having a desired temperature. At this time, the dopant material may be the impurity element itself, or may be a silicon material in which the impurity element is previously doped at a high concentration.

  Next, the melting crucible 101 a is tilted and the silicon melt 107 heated and melted is poured into the mold 103. The mold 103 filled with a predetermined amount of the silicon melt 107 moves to a solidification step, and the cooling plate 111 is brought into contact with the lower surface of the mold holding table, and at the same time, the silicon melt 107 is heated from above by the mold heating means 112. The output of the mold heating means 112 is adjusted so that the silicon melt 107 gradually solidifies from the bottom to the top of the mold 103.

Here, while the density of solid silicon is 2.33 g / cm ³ , the density of liquid silicon is 2.54 g / cm ³ , and attention is paid to the fact that the liquid expands by about 10% when the liquid solidifies. Estimating the solidification rate of the silicon melt 107 using the height of the liquid surface of the melt 107 as an index, and when the liquid surface height of the silicon melt 107 reaches a height corresponding to the desired solidification rate, A second dopant material 109 having a conductivity type opposite to that of the first dopant material 106 is supplied into the mold 103.

  By using such a method, since the second dopant material 109 is not previously added into the silicon melt 107 at the initial stage of solidification of the silicon melt 107, the apparent appearance of the P-type impurity element which is the main additive is obtained. Since the segregation coefficient does not change, the specific resistance value at the bottom of the silicon ingot does not deviate more than the desired value. Then, the solidification rate of the silicon melt 107 is estimated from the level of the silicon melt 107, and the second dopant material 109 is added to increase the apparent segregation coefficient of the P-type impurity element as the main additive. Therefore, it is possible to suppress a decrease in specific resistance, and it is possible to manufacture a polycrystalline silicon ingot having a small variation in specific resistance and uniform quality.

  Further, since the silicon melt 107 is solidified from the bottom of the mold 103 and the second dopant material 109 is supplied to the surface of the silicon melt 107 heated by the mold heating means 8, the P-type near the solid-liquid interface of silicon. The apparent segregation coefficient of the impurity element does not change abruptly and is uniformly diffused by thermal convection generated in the silicon melt 107. Furthermore, the degree of freedom of manufacturing conditions is not restricted, and various temperature conditions can be freely changed to control the solidification rate, so that various conditions for improving solar cell characteristics can be satisfied at the same time. It becomes possible.

  For example, the solidification rate of the silicon melt 107 in the mold 103 can be estimated by observing the height of the liquid surface of the silicon melt 107 as needed from a viewing window provided in the casting apparatus. In addition, by providing a mark such as a scale on the surface of the release material layer 108 previously applied to the side surface of the mold 103, the solidification rate can be estimated more accurately. Further, the supply of the second dopant material 109 can be controlled by monitoring the liquid level using a video camera or the like.

  Further, when the first dopant material 106 is a P-type impurity, the second dopant material 109 is made of an N-type impurity such as phosphorus or antimony having the opposite conductivity type, and the second dopant material 109 is usually an impurity element. Is supplied in the form of granular or granular silicon crystals contained in a high concentration, and after being completely melted into the silicon melt 107 by adjusting the output of the mold heating means 112, solidification proceeds again, and the specific resistance is increased. Silicon ingots with small variations can be cast. Alternatively, the silicon melt 107 containing the second dopant material 109 can be separately prepared and supplied into the mold 103.

  Here, the solidification rate of the silicon melt 107 when the second dopant material 109 is supplied is preferably between 30% and 80%. When the solidification rate of the silicon melt 107 is smaller than 30%, the second dopant concentration necessary for canceling the increase in the specific resistance at the ingot head is undesirably high. Further, there is a problem that the rise in the liquid level of the silicon melt 107 is small and difficult to detect. Further, when the solidification rate of the silicon melt 107 is advanced to 80%, the specific resistance is about 30% higher than that of the ingot bottom due to segregation of the first dopant material, and the second solidification rate is higher than that. Even if this dopant material is supplied, the effect of reducing variation in specific resistance is small.

  The N-type dopant concentration of the second dopant material 109 is preferably set to 20% to 40% with respect to the P-type dopant concentration of the first dopant material 106. When the N-type dopant concentration is less than 20%, the effect of canceling the P-type impurities is reduced. When the N-type dopant concentration is 40% or more, the specific resistance is greatly increased in the upper part of the ingot or the conductivity type is not reversed. Since it will occur, it is not preferable.

  Further, the second dopant material 109 may be supplied into the silicon melt 107 in a plurality of times.

  It should be noted that the embodiment of the present invention is not limited to the above-described example, and it is needless to say that various modifications can be made without departing from the gist of the present invention.

  For example, when the silicon melt 107 is poured from the melting crucible into the mold 103, it may be poured by a method other than the above method. For example, a discharge port may be provided at the bottom of the melting crucible, and the silicon melt 107 may be poured from the bottom into the mold 103 installed at the bottom. In this case, before the silicon material is completely melted, a mechanical plug or a discharge port is closed near the discharge port so that the silicon material before melting or the partially melted silicon melt 107 does not leak from the discharge port. A pouring control means capable of controlling pouring, such as installing a silicon material, is provided.

  Further, in the above description, the pouring type silicon casting apparatus provided with the melting crucible on the upper side has been described. However, the present invention is not limited to this, and the silicon material is melted inside the mold 103 to form the silicon melt 107. An in-mold melting type silicon casting apparatus may be used. A silicon material is introduced into the mold 103, and the silicon material is melted while suppressing heating of the portion where the second dopant material 109 is applied by heating means provided on the bottom and side surfaces of the mold 103 to form a silicon melt 107. Thus, the second dopant does not dissolve during the melting of the silicon material, and the surface of the silicon melt 107 rises in the solidification step, so that the second dopant dissolves into the silicon melt 107. Can be met.

  Furthermore, even if the first dopant material 106 is an N-type dopant and the second dopant material 109 having a conductivity type opposite to that of the first dopant material 106 is a P-type dopant, the effects of the present invention can be satisfied. Further, donor elements such as P (phosphorus), As (arsenic), and Sb (antimony) are used as the N-type dopant, and B (boron), Ga (gallium), In (indium), and Al are used as the P-type dopant. An acceptor element such as (aluminum) is used.

  A semiconductor ingot according to the present invention was manufactured as follows.

Example 1
A slurry made of fine N-type silicon powder and an appropriate binder is applied to the side wall of the quartz mold 103 having the release layer 108 formed on the inner surface at a position where the solidification rate of the silicon melt 107 is 50% to 100%. The prepared mold 103 (Example) and the mold (comparative example) to which the second dopant material 109 was not applied were prepared and placed in a vacuum container (not shown) of the silicon casting apparatus.

  Next, the high-purity quartz melting crucible 101a was held by a graphite holding crucible 101b, and 100 kg of silicon material and a specific resistance at the bottom of the ingot were calculated in advance so as to be 1.5 Ω · cm in the melting crucible 101a. A silicon melt 107 obtained by containing a predetermined amount of the first dopant material 106 and heating and melting the silicon material 105 and the first dopant 106 in the melting crucible 101a by the heating means 104 was prepared under the above-described conditions. Poured into mold 103. After pouring, the silicon melt 107 in the mold 103 was solidified in one direction from the bottom to the top to obtain a polycrystalline silicon ingot. Subsequently, the center longitudinal cross section of the obtained silicon ingot was cut, and the specific resistance value of the cross section was measured with a four-terminal measuring instrument.

(Example 2)
The molten crucible 1 made of high-purity quartz was held by a holding crucible 2 made of graphite, and 100 kg of silicon material and the specific resistance at the bottom of the cast ingot were calculated in advance so that the specific resistance was 1.5 Ω · cm. A first dopant composed of a predetermined amount of boron was added. The material in the crucible was heated and melted by the heating means 104 to form a silicon melt, and poured into the quartz mold 5. In the solidification process, the mold 5 poured with the silicon melt 9 was removed from the bottom by the cooling plate 7 and heated from the top by the mold heating means 8 to start unidirectional solidification from the bottom to the top. Then, the surface of the silicon melt is monitored using a video camera to estimate the solidification rate of the silicon melt. When the solidification rate reaches 50%, the concentration of the first dopant is 20%. A second dopant comprising phosphorus was fed into the mold 5. After adjusting the output of the mold heating means 8 and confirming that the second dopant was completely melted, the silicon melt was solidified to obtain a polycrystalline silicon ingot. Subsequently, the center longitudinal cross section of the obtained silicon ingot was cut, and the specific resistance value of the cross section was measured with a four-terminal measuring instrument.

(Comparative example)
The molten crucible 1 made of high-purity quartz was held by a holding crucible 2 made of graphite, and 100 kg of silicon material and the specific resistance at the bottom of the cast ingot were calculated in advance so that the specific resistance was 1.5 Ω · cm. A predetermined amount of the first dopant raw material was charged. The raw material in the crucible was heated and melted by the upper heating means 3a and the side heating means 3b to form a silicon melt, which was poured into the quartz mold 5. In the solidification process, the mold 5 poured with the silicon melt 9 is removed from the bottom by the cooling plate 7 and heated from the top by the mold heating means 8 to perform unidirectional solidification from the bottom to the top. Got. Subsequently, the center longitudinal cross section of the obtained silicon ingot was cut, and the specific resistance value of the cross section was measured with a four-terminal measuring instrument.

  As shown in FIG. 4, the difference in specific resistance between the top and bottom of the polycrystalline silicon ingots of Examples 1 and 2 is 0.2 Ω · cm, whereas the top and bottom of the silicon ingot of the comparative example The difference in specific resistance value was as large as 0.7 Ω · cm.

  Moreover, the general bulk type solar cell element was produced using the polycrystalline silicon substrate (the said Example and comparative example) obtained by the said method, and the conversion efficiency of the solar cell element was evaluated as the characteristic. As a result, in Examples 1 and 2, the conversion efficiency was 15.9%, but in the comparative example, it was 15.4%. As described above, the solar cell element formed using the polycrystalline silicon substrate of the example obtained better characteristics than those of the conventional example, but this was achieved using the method for manufacturing a semiconductor ingot of the present invention. The formed silicon ingot has the quality of uniform impurity concentration and small specific resistance variation, and distortion is reduced by doping phosphorus with a large ionic radius in the vicinity of boron with a small ionic radius as the main additive. Therefore, it is presumed that the effect of stress-induced dislocations is reduced.

It is a figure showing 1st embodiment which concerns on the manufacturing method of the semiconductor ingot of this invention. It is sectional drawing which shows the crucible and mold used for the manufacturing method of the semiconductor ingot of this invention. It is sectional drawing which shows the casting_mold | template used for the manufacturing method of the semiconductor ingot of this invention. It is a graph which shows the specific resistance of the silicon ingot manufactured by the manufacturing method of the semiconductor ingot of this invention.

Explanation of symbols

101a: melting crucible 101b: holding crucible 102: pouring spout 103: mold 104: heating means 105: silicon material (semiconductor material)
106: First dopant material 107: Silicon melt (semiconductor melt)
108: Release material layer 109: Second dopant material 110: Mold heat insulating material 111: Cooling means 112: Mold heating means

Claims (3)

In a method for producing a semiconductor ingot obtained by solidifying a semiconductor melt containing a first dopant material defining a first conductivity type in a predetermined direction in a mold,
A method for producing a semiconductor ingot, wherein a second dopant material defining a second conductivity type opposite to the first conductivity type is supplied into the semiconductor melt in the solidification process. The second dopant material is formed on the inner surface of the mold above the surface of the semiconductor melt before the start of solidification, and is dissolved in the semiconductor melt that has risen in the solidification process. The manufacturing method of the semiconductor ingot of Claim 1. 2. The semiconductor ingot according to claim 1, wherein a liquid level of the semiconductor melt is measured in the solidification process, and the second dopant material is supplied into the semiconductor melt according to the measured value. Method.

Images