JP2009184922A - Flowable chip, and method for using the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve such a problem that conventional granular polycrystalline silicon has too low purity irrespective of a process used for recharging a granular material depending on applications. <P>SOLUTION: Flowable chips are a polycrystalline silicon piece group which comprises polycrystalline silicon pieces prepared by a chemical vapor deposition process and having bulk impurities at a level not exceeding 0.03 ppma and surface impurities at a level not exceeding 15 ppba, and which moreover has a controlled particle size distribution and generally has nonspherical morphology. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention is based on US Provisional Application Serial No. 60 / 358,851 filed February 22, 2002 and US Patent Application Publication No. Claims priority based on the specification of 10 / 298,129.

  The present invention relates to a flowable chip, a device for manufacturing and using the flowable chip, and a method thereof. The flowable chip is effective in a method of reloading the crucible in the Czochralski method.

  Many semiconductor chips used in electronic devices are made from single crystal silicon manufactured by the Czochralski (CZ) method. In the CZ method, when producing a single crystal silicon ingot, the polycrystalline silicon raw material in the crucible is melted, the crucible and the molten raw material are stabilized at an equilibrium temperature, the seed crystal is immersed in the molten raw material, and the molten raw material is seeded. A single crystal ingot is grown by pulling out the seed crystal while crystallizing upward, and when the ingot grows, the ingot is pulled up. Melting is performed in an inert gas atmosphere at a temperature of 1420 ° C. and a low pressure. Crystals are grown while the crucible is continuously rotated about a substantially vertical axis. The speed at which the ingot is pulled up from the molten raw material is selected according to the desired diameter of the ingot to be produced.

  Polycrystalline silicon can be obtained using a fluidized bed reactor process to form particulates. Alternatively, polycrystalline silicon can be produced using chemical vapor deposition (CVD) in a vacuum vessel. Polycrystalline silicon produced by the CVD method may be crushed or cut into pieces of an appropriate size such as rods, chunks, chips or aggregates thereof, and then placed in a crucible. Polycrystalline silicon is melted to form molten silicon.

  One of the disadvantages of the CZ method is that when the loaded polycrystalline silicon melts, only half of the crucible contains molten silicon. This is because a gap remains in the crucible loaded with irregularly shaped fragments, which makes the use of the crystal pulling device inefficient. Therefore, it is desirable to develop a method for efficiently filling the charge after melting and before the start of seed crystallization.

  Another disadvantage of the CZ method is that the crucible usually has to be replaced after every pull-up, since the crucible may deteriorate with use and impurities may be mixed into the molten silicon. New crucibles are expensive and disposal of used crucibles is also expensive. For this reason, the development of an improved crucible that can be continuously used for pulling up an ingot many times and in which contamination (contamination) in molten silicon is reduced has been promoted. Therefore, it is necessary to efficiently reload the crucible during or after the first ingot and subsequent ingots are pulled up. So far, various methods have been proposed for filling the melt and reloading the crucible.

  One method is to leave granular polycrystalline silicon produced in a fluidized bed process (such as granular material from Ethyl or MEMC) in the crucible after pulling up the ingot, that is, to fill the initial charge melt. It is to load the melt heel. However, this method has a drawback that hydrogen is mixed in granular polycrystalline silicon produced by the fluidized bed method. When granular polycrystalline silicon is added over the heel, hydrogen is released, which may cause the granular material to rupture. This can cause the molten silicon to splash and break the crucible.

  Another method is to add granular polycrystalline silicon to the crucible while pulling up the ingot. However, this method has a drawback that it is difficult to melt the granular polycrystalline silicon even for a sufficient period of time so as to achieve an appropriate addition speed because the particle size is small. Since more heat is required to melt these small particles, new costs are incurred and crucible degradation is accelerated. Accelerating crucible degradation shortens crucible life and increases costs. When the granular polycrystalline silicon addition speed is too high and the granular material is not sufficiently melted, the surface of the ingot to be pulled is damaged, dislocation occurs, and single crystallinity is impaired. In addition, granular polycrystalline silicon may contain a large amount of dust. Dust can cause contamination problems in the housing of the pulling device and can stick to the surface of the pulled ingot, causing dislocations and reducing crystal yield. This also increases process time because remelting and redrawing is required.

  Eventually, the granular polycrystalline silicon is too low in purity, depending on the application, regardless of the process used to reload the granular material.

  Attempts to use polycrystalline silicon rods produced by chemical vapor deposition and broken into pieces have not been directed to crucible refilling due to purity or size issues. The use of relatively large sized pieces of polycrystalline silicon for crucible refilling has the disadvantage of damaging the crucible and reloading device. If the polycrystalline silicon piece is broken into smaller sizes, the polycrystalline silicon piece becomes unsuitable for use in the crucible reloading process due to contamination by impurities.

The present invention relates to a flowable chip, a device for manufacturing and using the flowable chip, and a method thereof.
A flowable chip is manufactured by chemical vapor deposition and has a bulk silicon impurity level not exceeding 0.03 ppma and a polycrystalline silicon fragment group having a surface impurity level not exceeding 15 ppba. And consisting of a group of polycrystalline silicon pieces having a controlled particle size distribution and generally having a non-spherical morphology.
Flowable chips are:
a) subdividing the polycrystalline silicon rod;
b) controlling the particle size distribution by sorting the product from step a) using a step deck sorter;
c) removing impurities from the product in step a) or step b) or both steps. Step b) may be performed using a step deck sorter. Step c) may comprise exposing the product of step a) or step b) or both to a magnetic field. Step c) may comprise surface cleaning the product from step b).

The present invention further relates to a method of reloading a crucible used in the Czochralski method using a flowable tip. This method is:
a) pulling the silicon ingot from the crucible by the Czochralski method;
b) adding a flowable tip to the molten silicon in the crucible, wherein the flowable tip is manufactured by chemical vapor deposition and has a low level of bulk impurities and a low level of surface impurities A stage comprising a group of polycrystalline silicon particles composed of silicon particles and having a controlled particle size distribution and generally a non-spherical morphology;
c) optionally adding a dopant to the crucible.
The flowable chip used in the present invention comprises a group of polycrystalline silicon pieces composed of polycrystalline silicon pieces having low level bulk impurities and low level surface impurities, and has a controlled particle size distribution. And it is preferable that it consists only of a group of polycrystalline silicon having a non-spherical morphology.

The present invention further includes:
a) lifting the silicon ingot from the crucible by the Czochralski method and leaving the heel in the crucible;
b) solidifying at least the surface of the heel;
c) adding a flowable tip to the surface of the heel, wherein the flowable tip is manufactured by chemical vapor deposition and has low levels of bulk impurities and low levels of surface impurities And comprising a group of polycrystalline silicon particles having a controlled particle size distribution of 1 mm to 12 mm and generally having a non-spherical morphology;
d) optionally adding a dopant to the crucible.
The flowable chip used in the present invention is a group of polycrystalline silicon particles composed of polycrystalline silicon particles having low level bulk impurities and low level surface impurities, and has a controlled particle size distribution of 1 mm to 12 mm. It preferably comprises only a group of polycrystalline silicon particles having and generally having a non-spherical morphology.

It is a schematic block diagram of a CZ apparatus. It is a schematic block diagram of the vibration feeder apparatus which reloads the crucible used by CZ method. It is a schematic block diagram of the canister feeder apparatus which reloads the crucible used by CZ method. It is a schematic sectional side view of the jaw crusher used by the method of manufacturing a fluid tip. It is a schematic sectional side view of the step deck sorter used with the method of manufacturing a flowable chip. FIG. 6 is an enlarged side cross-sectional view of the second and third decks of the step deck sorter of FIG. 5. It is a schematic plan view of the 2nd step deck of the step deck sorter of FIG. It is a schematic sectional side view of the 2nd deck of the step deck sorter of FIG.

  All quantities, ratios and proportions are given by weight and are otherwise indicated. The following is a definition list used in this specification.

<Definition>
“A” and “an” both mean one or more.

  “Blinding” means that the gap between two decks is clogged in the step deck sorter, and the piece of polycrystalline silicon cannot pass through the gap, so that the step deck sorter does not function as a sorting means. Means.

  “Charge maximization” is the process of increasing the amount of melt over that obtained by random filling, by using polycrystalline silicon with various sizes and shapes for containers such as molds or crucibles. Means the process of filling

  “Filling” means the process of filling a container, such as a mold or crucible, with polycrystalline silicon, where the polycrystalline silicon is melted and then further polycrystalline silicon is added to increase the amount of melt. It is added.

  “Chemical vapor deposition” refers to all chemical vapor deposition methods that do not include a fluidized bed reactor method to produce polycrystalline silicon. As an example of chemical vapor deposition, there is a Siemens growth method.

  “Subdivide” means crushing, cutting, or grinding into small particles. Fragmentation includes any method that reduces polycrystalline rods to fragments and is not limited to cutting the rods and then crushing them in various ways.

  “Controlled particle size distribution” means that at least 75% of the particles in the particle group have a specific range of particle sizes. For example, a controlled particle size of 4 mm to 12 mm means that at least 75% of the particles are in the range of 4 mm to 12 mm and the remaining up to 25% of the particles are outside the range of 4 mm to 12 mm.

  “Donor” means an atom that donates electrons to silicon. Donors include antimony, arsenic and phosphorus.

  “Fluidity” means that a large number of solid particles pass through a transfer device without bonding between particles, including when vibrational energy is applied to the device to prevent the formation of a network structure where particles are bonded. Means the ability to move.

A “flowable chip” is a group of polycrystalline silicon particles composed of polycrystalline silicon particles having low levels of bulk impurities and low levels of surface impurities, having a controlled particle size distribution and generally non-spherical morphology. Means a group of polycrystalline silicon particles.

  “Granular” and “granular” both mean polycrystalline silicon particles produced by a fluidized bed process and having a particle size of 6 mm or less. Granules are usually spherical or nearly spherical.

  “Heel” means a certain amount of silicon left in the container. The residue includes a lump of silicon remaining in the crucible after the ingot is lifted from the crucible, and a lump of molten silicon generated by melting the charge in the container before filling.

  “Granularity” means the longest distance between two points on a particle. For example, in the case of spherical particles, the particle size is the diameter.

  The abbreviation “ppba” means one billionth of an atom with respect to the number of silicon atoms.

  The abbreviation “ppma” means 1 / 1,000,000 atom number with respect to silicon atom number.

<Fluidity chip>
A flowable chip is a group of polycrystalline silicon particles having a controlled particle size distribution. The controlled particle size distribution may be from 0.2 mm to 45 mm, alternatively from 1 mm to 25 mm, alternatively from 1 mm to 20 mm, alternatively from 3 mm to 20 mm, alternatively from 4 mm to 12 mm, or It may be 4 mm to 10 mm, or 1 mm to 12 mm, or 1 mm to 8 mm. However, a precisely controlled particle size distribution is selected based on a variety of factors including the method in which the flowable tip is used and the equipment used to supply the flowable tip. For example, the control particle size distribution may be from 2 mm to 45 mm for flowable chips used in the CZ method or electronic grade applications or both. In addition, in a flowable chip used for solar cell grade applications such as casting, the control particle size distribution may be 0.2 mm to 45 mm.

  Depending on the application, the controlled particle size distribution may be 4 mm to 12 mm, 4 mm to 8 mm, or 4 mm to 6 mm. While not intending to be limited by theory, the splash when adding a flowable tip to a crucible containing molten silicon by setting the control particle size distribution to be lower in the range of 4 mm to 12 mm. It is believed that scattering is minimized. Depending on the application, the control particle size distribution may be 9 mm to 12 mm, or 10 mm to 12 mm. While not intending to be limited by theory, when adding a flowable tip to a crucible containing an at least partially solidified heel by setting the control particle size distribution to be lower in the range of 4 mm to 12 mm Splashing is believed to be minimal.

  The ranges disclosed herein disclose not only the ranges themselves, but also any ranges that are encompassed by the ranges, including the boundaries of the ranges. For example, disclosure of the range of 4 mm to 12 mm includes not only the range of 4 mm to 12 mm but also other numerical values included in this range, such as 4 mm, 5 mm, 7 mm, 11 mm, and 12 mm. Further, for example, the disclosure of the range of 4 mm to 12 mm includes 4 mm to 8 mm, 9 mm to 10 mm, 9 mm to 12 mm, and 10 mm to 12 mm as well as other partial ranges included in the range. In addition, ranges equivalent to the ranges disclosed herein are also included.

  The morphology of flowable chips is usually non-spherical. The exact morphology depends on the method used to make the flowable chip. For example, when a flowable chip is produced by manually breaking a polycrystalline silicon rod by hitting it with a low-contamination impact tool such as disclosed in the method disclosed herein, such as disclosed in European Patent Application No. 0539097, etc., Becomes irregular.

  The flowable tip has low levels of bulk impurities of boron, donor, phosphorus, carbon and all metals. The level of bulk impurities may be less than or equal to 0.2 ppma, may be less than or equal to 0.03 ppma, or may be less than or equal to 0.025 ppma. The boron level may be less than or equal to 0.06 ppba. However, depending on the application, for example, when boron is used as a dopant, the level of boron may be less than or equal to 20 ppba or in the range of 5 ppba to 20 ppba.

  The donor level may be less than or equal to 0.30 ppba. The flowable chip may have a level of phosphorus less than or equal to 0.02 ppba, or less than or equal to 0.015 ppba. The level of carbon may be less than or equal to 0.17 ppba. The total level of bulk metal impurities may be less than or equal to 4.5 ppba, or may be less than or equal to 1 ppba. Bulk metal impurities include Cr, Cu, Fe and Ni. The flowable tip may have a bulk level of Cr less than or equal to 0.01 ppba. The flowable tip may have a bulk level of Cu that is less than or equal to 0.01 ppba. The flowable tip may have a bulk level of Fe that is less than or equal to 0.01 ppba. The flowable tip may have a bulk level of Ni less than or equal to 0.01 ppba.

  Bulk metal impurities can be obtained by the float zone method disclosed in U.S. Pat. No. 4,912,528, U.S. Pat. No. 5,361,128, and U.S. Pat. It can be measured by a known method such as the method shown in Example 3 of the specification.

  The flowable chip has a low level of total surface impurities. The total level of surface impurities may be less than or equal to 30 ppba, may be less than or equal to 15 ppba, or may be less than or equal to 4.5 ppba. The surface impurities include Co, Cr, Cu, Fe, Na, Ni, W, and Zn.

  When the flowable chip is manufactured by the method described below using the jaw crusher as shown in FIG. 4 and the step deck sorter as shown in FIGS. 5 to 8, the flowable chip has a smaller amount of surface. May have impurities. For example, the flowable chip is less than or equal to 0.06 ppba, or less than or equal to 0.02 ppba, or less than or equal to 0.01 ppba, or more than 0.004 ppba. It can have a small or equal amount of Cr. Is the flowable chip less than or equal to 0.15 ppba, or less than or equal to 0.03 ppba, or less than or equal to 0.02 ppba, or less than 0.01 ppba? Alternatively, it may have an equal amount of Cu. The flowable chip has an amount of Fe less than or equal to 18 ppba, or less than or equal to 10 ppba, or less than or equal to 9 ppba, or less than or equal to 7 ppba. Can do. Is the flowable chip less than or equal to 0.9 ppba, or less than or equal to 0.8 ppba, or less than or equal to 0.5 ppba, or less than 0.4 ppba? Or it may have an equal amount of Na. The flowable chip has an amount less than or equal to 0.1 ppba, or an amount less than or equal to 0.07 ppba, or an amount less than or equal to 0.04 ppba, or 0 ppba Ni. obtain. Is the flowable chip less than or equal to 0.6 ppba, or less than or equal to 0.5 ppba, or less than or equal to 0.4 ppba, or less than 0.3 ppba? Alternatively, it may have an equivalent amount of Zn.

  Surface purity can be measured by known methods, such as the method disclosed in US Pat. No. 5,851,303.

  Flowable chips can also have low levels of dust. While not intending to be limited by theory, it is believed that when flowable chips are added to the crucible, low levels of dust facilitate melting and reduce the crystal dislocation rate.

The flowable chip can have a low residual gas content. The flowable chips may contain no hydrogen at all, or may contain less levels of hydrogen than granules produced by the fluidized bed process. The hydrogen content of the flowable chip can be from 0 to 3600 ppba, alternatively from 0 to 1300 ppba, alternatively from 0 to 800 ppba, alternatively from 800 to 1300 ppba. The flowable tip can contain low levels of chlorine. The chlorine content of the flowable chip can be from 0 to 300 ppba, alternatively from 20 to 120 ppba, alternatively from 25 to 110 ppba, alternatively from 30 to 100 ppba, alternatively from 50 to 65 ppba.
The flowable chip of the present invention is a group of polycrystalline silicon pieces composed of polycrystalline silicon pieces having a bulk impurity level not exceeding 0.03 ppma and a surface impurity level not exceeding 15 ppba, Preferably, it consists solely of a group of polycrystalline silicon pieces having a controlled particle size distribution and generally having a non-spherical morphology.

<Method for producing a flowable chip>
Flowable chips are:
a) crushing or cutting the polycrystalline silicon rod;
b) selecting the product of step a) to control the particle size distribution;
Optionally, it may be produced by a method comprising: c) surface cleaning the product of step a) or step b) or both.

The flowable chip product is further added to the method described above:
d) packaging the product of step a), step b), or step c).

And the flowable chip is:
a) subdividing the polycrystalline silicon rod;
b) selecting the product of step a) to control the particle size distribution;
c) removing impurities from the product of step a) or step b) or both steps.

The flowable chip product is further to the method described above:
d) packaging the product of step a), step b) or step c).

<Manufacture of polycrystalline silicon>
The polycrystalline silicon rod can be manufactured by a known method. For example, a polycrystalline silicon rod can be manufactured by a chemical vapor deposition method including chemical vapor deposition on a heated substrate of high-purity chlorine gas or silane gas. "Handbook of Semiconductor Silicon Technology" (William C. O'Mara), edited by Robert B. Herring and Lee P. Hunt, Neuss Publishing, Park Ridge , New Jersey, USA, 1990), Chapter 2, pages 39-58.

<Division of polycrystalline silicon>
Polycrystalline silicon rods can be subdivided, for example, by sawing or by tapping with a low-contamination impact tool such as disclosed in EP 0 590 097. Alternatively, the polycrystalline silicon rod may be subdivided with a jaw crusher. Alternatively, the polycrystalline silicon rod may be subdivided by hitting with a low-contamination impact tool, and the subdivided rod may be further subdivided by a jaw crusher. Alternatively, the polycrystalline silicon rod may be subdivided by cutting with a saw, then hit with a low-contamination impact tool, and further subdivided with a jaw crusher. An example of a suitable jaw crusher is shown in FIG. The jaw crusher 400 includes a frame assembly 401 on which a fixed jaw plate 402 is mounted. The movable jaw plate 403 faces the fixed jaw plate 402. A jaw cavity 404 is formed between the jaw plates 402 and 403. Polycrystalline silicon can be transferred from hopper 425 to jaw cavity 404.

  The movable jaw plate 403 is provided on the pitman carrier assembly 405. The pitman carrier assembly 405 is coupled to a pitman bearing 406 that surrounds an eccentric shaft 407 at one end and a tension rod pin 408 at the other end. The eccentric shaft 407 is provided on the flywheel 409. Motor 410 drives belt 411 around flywheel 409. The flywheel 409 rotates the eccentric shaft 407 to move the movable jaw plate 403 in an elliptical motion with respect to the fixed jaw plate 402. The rotation speed may be 300 to 400 times / minute (r.p.m). The motor 410 is provided on the base. By the movement of the movable jaw plate 403, the polycrystalline silicon in the jaw cavity 404 is crushed. The grain size of the resulting polycrystalline silicon piece is sufficiently reduced for the polycrystalline silicon piece to exit the jaw cavity through the discharge slot 418.

  In order to apply tension to the rod pin 408, the horizontal spring assembly includes a tension rod 413. The tension rod extends through adjustment wheel 414, outer spring collar 415, tension spring 416, and inner spring collar 417. The adjustment wheel 414 provided on the outer spring collar 415 can rotate to adjust the tension spring. A horizontal spring assembly can be used to hold the pitman 405 in contact with the toggle plate 424.

  The vertical assembly includes a wedge-shaped adjustment rod 419 that extends through an adjustment wheel 420 and a cross bar 421. The adjustment wedge 422 is provided on the bearing wedge 423. A toggle plate bearing wedge 423 is provided on the toggle plate 424. A toggle plate 424 is mounted on the pitman carrier assembly 405 above the tension rod pin 408. This vertical assembly can be used to adjust the width of the discharge slot 418. The position of the toggle plate 424 (in the groove of the bearing wedge 423) determines the movement of the pitman 405 and the bottom of the movable jaw plate 403.

  Polycrystalline silicon can be supplied from the hopper 425 to the jaw crusher 400. When polycrystalline silicon is applied to jaw crusher 40, it is divided into smaller polycrystalline silicon pieces by movable jaw plate 403. Polycrystalline silicon pieces can be resized from dust to lumps, debris, flakes, and oversized chunks. The particle size distribution of the polycrystalline silicon pieces depends on various factors including the width of the discharge slot 418 and the residence time in the fracture cavity 404.

The jaw plates 402, 403 are made of a material that minimizes silicon contamination, such as a material having a hardness greater than or comparable to polycrystalline silicon. Jaws is tungsten carbide, tungsten carbide containing cobalt binder, tungsten carbide containing nickel _{binder, Cr 2 C 3, Cr 2} C 3 including NiCr binder, or may consist of these combinations. By using a tungsten carbide-containing material, the level of iron contaminants mixed into the silicon by the fragmentation process can be reduced. The hopper 425 supplying the polycrystalline silicon rod and / or the oversized polycrystalline silicon piece and both, and the discharge chute (not shown) from the jaw crusher 400 are made of the same material as the jaw plates 402 and 403 or silicon contamination. It may be composed of another constituent material that minimizes the above, or may be lined with the material. Such materials include ultra high molecular weight polyethylene (UHMWPE), polypropylene, perfluoroalkoxy resin (PFA), polyurethane (PU), polyvinylidene difluoride (PVDF), Teflon (registered trademark), tungsten carbide, silicon And ceramic.

  Multiple jaw crushers placed in series to obtain the desired shape distribution and / or the above-mentioned particle size distribution, or to recycle oversized (too large) polycrystalline silicon pieces, or both Those skilled in the art will recognize that these may be used. "Introduction to Particle Technology" (John Wiley & Sun, New York), where the material of the components of the fragmentation device that contacts the silicon contains materials that minimize silicon contamination State, April 1999), Chapter 10, pp. 241-263, conventional devices such as jaw crushers, gyratory crushers, crushing roll devices, cone crushers, and table milling machines Those skilled in the art will recognize that can be used in the present invention. As suitable jaw crushers, Morse Jaw Crushers sold by Metso Minerals Industories, Inc. of Danville, Pa. Can be used.

  Those skilled in the art will recognize that other conventional fragmentation devices may be used in the method of the present invention in conjunction with or in place of the jaw crusher. Suitable subdividing devices are U.S. Pat. No. 4,815,667, U.S. Pat. No. 5,346,141, U.S. Pat. No. 5,464,159, EP 0573855. , Japanese Patent No. 0256759 and Japanese Patent No. 58145611.

<Selection of polycrystalline silicon pieces>
Assuming that the portion of the sorter that contacts silicon is made of a material that minimizes silicon contamination as described above, a piece of polycrystalline silicon (crushed rod) is disclosed in US Pat. No. 5,165,548. Rotating silicon screens disclosed in US Pat. No. 3,905,556, US Pat. No. 5,064,076 or 5,791,493. Sorting can be done using a device such as a device or manually.

The polycrystalline silicon piece may be sorted using an apparatus including a step deck sorter. An apparatus for sorting out polycrystalline silicon pieces
I) a vibration motor assembly;
II) a step deck sorter provided in the vibration motor assembly;
Is provided.

Step deck sorter
i) the first deck,
a) a polycrystalline silicon piece inlet to the groove forming region;
b) The groove forming region extends from the polycrystalline silicon piece inlet, or extends downstream from the polycrystalline silicon piece inlet, and each groove has a ridge and a valley.
c) A slope formed so that the ridge portion of the groove protrudes beyond the first gap between the first deck and the final deck than the trough portion of the groove. A first deck with one deck exit end;
ii) a final deck located downstream of the first gap and below the first deck,
a) polycrystalline silicon piece inlet,
b) a groove forming region extending from the polycrystalline silicon piece inlet or formed downstream of the polycrystalline silicon piece inlet, wherein each groove comprises a ridge and a valley;
c) a final deck comprising an outlet for polycrystalline silicon pieces;
iii) a collection container provided under the first gap for collecting the pieces of polycrystalline silicon that have fallen through the first gap;
iv) a collection container for oversized polycrystalline silicon pieces provided under the exit of the final deck to collect oversized polycrystalline silicon pieces that did not fall through the first gap;
It has.

The step deck sorter may comprise one or more additional decks between the first and final decks, each additional deck being
a) polycrystalline silicon piece inlet,
b) a groove forming region extending from the polycrystalline silicon piece inlet or downstream of the polycrystalline silicon piece inlet, wherein each groove has a ridge and a valley;
c) an additional deck exit, the ridge portion of the groove being more inclined than the trough portion of the groove so as to protrude over the gap at the exit end, Yes.

  The polycrystalline silicon sorting device is further possible to those skilled in the art without undue experimentation, either under the collection container of iii), a dust removal device located at the downstream of the first gap or both, or both Such a deformation device may be provided. One skilled in the art will recognize that one or more step deck sorters can be used in series to sort polycrystalline silicon pieces.

  Examples of apparatus for sorting polycrystalline silicon pieces and including a step deck sorter are shown in FIGS. FIG. 5 is a side view of the apparatus. The step deck sorter 500 has an inlet 502 for polycrystalline silicon pieces. The vibration motor assembly 501 moves the polycrystalline silicon piece across the first deck 531. The polycrystalline silicon piece first passes through the fluidized bed region 503 where the dust is removed by the air flow indicated by the arrow 504, and proceeds to the dust collector 532 through the perforated plate 505. The polycrystalline silicon piece moves through the fluidized bed region 503 to the groove forming region 506. Depending on the size and shape, the polycrystalline silicon piece will either fit into the trough 520 of the groove 512 (shown in FIGS. 6-8) or remain at the top of the ridge 519 of the groove 512. When the polycrystalline silicon piece reaches the end of the first deck 531, the polycrystalline silicon piece smaller than the gap 507 passes through the gap 507 and falls onto the conveyor 508. The dropped piece of polycrystalline silicon is transported to a collection container for a piece of polycrystalline silicon 509. A large piece of polycrystalline silicon passes over the gap and falls onto the second deck 510.

  FIG. 7 is a plan view of the second deck 510, and FIG. 8 is a cross-sectional view of the second deck along the line AA. There are a plurality of grooves 512 at the top of the second deck 510. Each groove 512 has a ridge 519 and a valley 520. The groove 512 is semicircular. The side wall 530 extends beyond the ridge 519 to prevent the polycrystalline silicon pieces from falling from the side of the second deck 510. The polycrystalline silicon piece moves from the inlet end 511 of the second deck 510 to the outlet end 518 of the second deck 510.

  FIG. 6 is a cross-sectional view of the second deck 510, the third deck, and the gap 516 therebetween. The input port end 511 of the second deck 510 is orthogonal to the horizontal. The groove 512 is cut into the top of the second deck 510. Polycrystalline silicon pieces such as chunks 513 and fragments 514 will rest in the troughs 520 of the grooves. A piece of polycrystalline silicon such as flake 515 will rest on top of ridge 519 of groove 512. The outlet end 518 of the second deck 510 is inclined so that the ridge 519 of the groove 512 protrudes beyond the gap 516 than the valley 520 of the groove 512. As the second deck 510 vibrates, the mass 513 falls through the gap 516, while the debris 514 and flake 515 roll off the exit end 518, over the gap 516, and carried to the third deck 517. While not intending to be limited by theory, the angle of inclination of the outlet end 518 of the second deck 510 will make the blinding effect as small as possible. The deck is made thin at the exit to further reduce the blind effect.

  The step deck sorter 500 divides the polycrystalline pieces 513, 514 and 515 based on the size of the gaps 507, 516, 518, 534, 522 and 524 between the decks 531, 510, 517, 533, 521, 523 and 525. Separated into controlled particle size distribution. The gaps 507, 516, 518, 534, 522 and 524 are increasing in size in the transport direction. Smaller polycrystalline silicon pieces fall through the smaller gaps 507, 516 and 518 and are collected in a collection container 509 for smaller polycrystalline silicon pieces. Larger pieces of polycrystalline silicon fall through the larger gaps 534, 522 and 524 and are collected in a collection vessel 526 for larger pieces of polycrystalline silicon. Oversized polycrystalline silicon pieces are collected at the end of the step deck sorter 500 in a collection container 527 for oversized polycrystalline silicon pieces. Oversized polycrystalline silicon pieces can be recycled to the fragmentation device. Polycrystalline silicon pieces with different controlled particle size distribution, step deck sorting through gaps between decks by changing gap size, number of collection containers and number and position of conveyors moving polycrystalline silicon pieces through collection containers Those skilled in the art will recognize that they are withdrawn from various decks of vessels.

  To directly fill a collection container 509, 526, 527 such as a bag, or to stop moving and change the collection container when a predetermined fill weight is reached, a weigh scale 528 is connected to the controller of the vibratory feeder 501. It may be incorporated.

  Those skilled in the art will understand that the number of decks; the width, depth and shape of the grooves in each deck; the size of the gap between the decks; and adjusting the particle size distribution collected by varying the number of collection containers . The grooves may have different widths, depths and shapes. For example, the groove may be triangular, trapezoidal, or semicircular.

  Similar to the device used to subdivide the polycrystalline silicon, the portion of the device used for sorting that contacts the polycrystalline silicon piece is made of a material that does not contaminate the silicon, such as the component materials described above for the jaw crusher 400. .

  In addition, the step deck sorter described above can be adapted to a larger size (eg, 45 mm or more) by changing the size of the deck; the width, depth and shape of the grooves in each deck; the size of the gap between the decks; Those skilled in the art will recognize that materials other than flowable chips, such as polycrystalline silicon pieces, can be used to sort materials.

<Contaminant removal as an option>
Polycrystalline silicon pieces having a controlled particle size distribution obtained as described above may optionally be exposed to a magnetic field to remove contaminants. For example, a piece of polycrystalline silicon may be passed through a chamber containing a magnet to remove contaminants, or a magnet may be passed over the piece of polycrystalline silicon. The magnet may be a rare earth magnet or an electromagnet or a combination thereof. The magnet may be in direct contact with the polycrystalline silicon piece or may be in the vicinity of the polycrystalline silicon piece. The magnet removes most of the fine particles having the appropriate magnetic susceptibility. These particles include ferromagnetic impurities such as iron and cobalt, paramagnetic impurities such as tungsten carbide, and other ferromagnets used to generate materials for the equipment used to subdivide and sort silicon. Impurities and paramagnetic impurities are included.

  Alternatively, contaminants can be found in U.S. Pat. No. 3,905,556, U.S. Pat. No. 4,125,191, U.S. Pat. No. 4,157,953, U.S. Pat. No. 4,250,025. Specification, US Pat. No. 4,345,995, US Pat. No. 4,525,336, US Pat. No. 5,297,744, or US Pat. No. 5,830,282. It may be removed by a method as disclosed in the specification. Contaminants may be removed by chemical methods as disclosed in EP 0215121 in addition to or instead of methods involving magnetic fields.

  Depending on the purity of the polycrystalline silicon rod used as the starting material and the method of fragmenting and sorting the silicon, the product at this stage can be either a solar grade single crystal silicon wafer or an electronic grade single crystal silicon wafer. It may have sufficient purity for use in manufacturing. However, if the purity is not sufficient for electronic product grade applications or both, silicon may be surface cleaned to further remove impurities.

<Surface cleaning as an option>
The surface of the polycrystalline silicon piece can be cleaned by a known method. Surface cleaning may be performed in addition to or instead of the contaminant removal method described above. For example, the crushing rod is continuously contacted with the method disclosed in US Pat. No. 5,851,303, that is, the crushing rod is in contact with an aqueous solution containing hydrogen fluoride gas, at least 0.5% hydrogen peroxide. It can be cleaned by a method comprising a step and subsequently drying the crushing rod. Alternatively, the crushing rod can be cleaned by the method disclosed in JP-A-5-4811. Alternatively, the crushing rod can be surface cleaned by anisotropic etching as disclosed in Canadian Patent No. 954425 or US Pat. No. 4,971,654. Other suitable surface cleaning methods include those disclosed in US Pat. Nos. 4,588,571 and 6,004,402.

  The produced flowable chips can be packaged by any convenient means, i.e. manually or automatically by placing the flowable chips in a polyethylene bag.

<How to use fluidity chip>
The flowable chips described above can be used in solar cell grade applications or electronic product grade applications depending on the particle size distribution and purity. Special applications of flowable tips include initial loading related applications such as initial loading maximization and initial loading filling, CZ crucible reloading and reloading maximization and reload filling. Reload related applications are included.

  In addition to or instead of the silicon described herein, flowable chips can be formed using U.S. Pat. No. 4,176,166, U.S. Pat. No. 4,312,700, U.S. Pat. No. 5,382,838, US Pat. No. 4,572,812, US Pat. No. 5,254,300, US Pat. No. 5,431,869, US Pat. No. 5,492. No. 5,079, US Pat. No. 5,510,095, Chinese Patent No. 1176319, German Patent No. 4441911, European Patent No. 0869102, European Patent No. 0095757 , Japanese Patent No. 10190025, Japanese Patent No. 11116386, Japanese Patent No. 58026019, Japanese Patent No. 58099115, Japanese Patent 62108515 and JP may be used in Japanese Patent solar cell casting method as disclosed in the 9301709 JP. Casting may include a step of injecting molten silicon into a heated mold, a step of melting polycrystalline silicon in the crucible, and a step of gradually cooling and solidifying the silicon.

For example, a suitable batch casting method is:
1) introducing a semiconductor material into a mold with walls defining a desired cross-sectional shape;
2) melting the semiconductor material;
3) After step 2), solidifying the semiconductor material to produce a cast ingot having a desired cross-sectional shape. The implementation of step 2) can be any of before, during and after step 1). The cast ingot may be removed from the mold after step 3) and then the method repeated. The flowable chip described above may be used to load the mold in step 1).

A continuous casting method can also be used. The continuous casting method is:
1) melting a semiconductor material continuously supplied to a bottomless container disposed in an induction coil;
Optionally 2) blowing hot plasma gas over the surface of the melt for purification;
And 3) continuously releasing the solidified silicon from the bottomless container downstream. At least the shaft portion of the bottomless container is divided into a plurality of conductive pieces spaced apart in the circumferential direction. The semiconductor material may comprise the above-described flowable chip.

  The apparatus used in the preferred continuous casting method also includes a plurality of conductive members arranged side by side to define a container-like region having an open top and an open bottom. The apparatus further comprises means for including a high frequency AC current in each conductive member. A telescopic support member is provided through the open bottom of the container-like region. The support member functions to support the semiconductor material in the container-like region.

The preferred continuous casting method is:
1) introducing a semiconductor material into the container-like region of the device described above;
2) melting the semiconductor material;
3) Inducing a first current in each conductive member by applying a voltage to the means for inducing current;
4) Inducing a second current in the semiconductor material using the first current, wherein the second current is flowing in a substantially opposite direction to the direction of the first current. When,
5) using a first current and a second current so as to avoid the semiconductor material from repelling and contacting the conductive member during the casting process.

This method further:
6) pulling the support member away from the container-like region so that the molten semiconductor material supported by the support member is repelled from the conductive member and the molten semiconductor material is solidified in the cast ingot;
And 7) additionally supplying a semiconductor material on the container-like region. Steps 1) to 7) may be repeated. The material used in step 1) or step 7) or both may include the flowable tip described above.

  The flowable tip may be used in a regular length method such as a regular edge thin film growth method (EFG) for producing silicon ribbons. The EFG method is based on the article “EFG-Invention and Sapphire Growth” published in the Journal of CrystalGrowth, Vol. 8-17 (1980) by HELaBelle.Jr et al. "EFG The Invention and Application to Sapphire Growth", and K. Koliwad et al. On April 15, 1984, "Plate Solar Array Project for High-Speed Crystal Growth and Crystal Characterization for Solar Cells" Research Forum Proceedings (edited by Catherine Damas, Jet Propulsion Laboratory (JPL), California Institute of Technology, Pasadena, California, US Department of Energy), pages 22-24 . A suitable EFG method may include the step of lifting the silicon ribbon from the molten silicon meniscus defined by the die edge. Ribbon dimensions are controlled by the meniscus. The ribbon production rate and the thermal balance of the equipment are carefully controlled.

  The flowable chip is made up of Japanese Patent No. 10182124 and Crystal Growth (Brian R. Pamplin) issued by Pergamon Press, Ltd. (Oxford) in 1975. Ed.), Chapter 9, pp. 343-344, and may be used in an induction plasma method as disclosed in the paper “Growth, Measurement and Control of Crystal Growth Environment”. As an example of the induction plasma method, for example, silicon particles are melted using a high frequency plasma torch of 4 megahertz (MHz) or higher. The torch is generated by ionizing it through a high frequency electric field in which an inert gas such as argon is applied between the cathode and the anode. Once the argon stream is ionized into a high temperature plasma, powdered silicon can be commonly supplied from the hopper to the plasma jet. Silicon melts in the jetting zone and the molten silicon can be directed to a water-cooled crucible or onto a growing crystalline silicon body. A flowable silicon tip in a suitable size range for the plasma torch cavity can be used as a silicon source for such a method.

  The flowable tip may be used in an electron beam melting method as disclosed in US Pat. No. 5,454,424 and Japanese Patent No. 62260710. In an embodiment of the electron beam melting method, melting the polycrystalline silicon by scanning the polycrystalline silicon with an electron beam and casting the resulting molten silicon by any of the casting methods described herein. Is provided. The polycrystalline silicon may include a flowable chip.

  The flowable chip may be used in a heat exchange method (HEM). The HEM furnace may comprise a chamber containing a crucible surrounded by a heating element using a helium heat exchanger coupled to the bottom of the crucible. Polycrystalline silicon is placed on top of the seed crystal to fill the crucible. The chamber is evacuated and the heating element is heated to melt the silicon. The seed crystal is avoided from melting by flowing helium gas through the heat exchanger. When the gas flow is gradually increased, the temperature of the heat exchanger is lowered, the silicon is gradually solidified, and the crystal grows outside the seed crystal. The temperature of the molten silicon is controlled by a heating element; however, the temperature of the solid crystals is independently controlled by a heat exchanger. By controlling both heating and cooling in a binary manner, the position and movement of the solid-liquid interface of the crystal growth process can be controlled. The term HEM refers to the paper “HEM Silicon” published in Crystal Systems (National Renewable Energy Laboratory, Golden, Colorado), and Federick Schmid and Chandra P. Khattak's paper “The cost of cutting silicon wafers for photovoltaic cells (Optical Spectra, May 1981)”.

Flowable tips may be used in the string ribbon method as disclosed in US Pat. No. 4,689,109. In the string ribbon example:
1) pulling the two strings and seed crystal vertically out of the shallow silicon melt;
2) Wetting the strings and seed crystals with molten silicon to fill the spaces between the strings;
3) cooling the product of step 2) to produce a silicon ribbon. The string ribbon method is continuous and can be initially charged with a melt and reloaded with a flowable tip.

U.S. Pat. No. 4,095,329, U.S. Pat. No. 4,323,419, U.S. Pat. No. 4,447,289, U.S. Pat. No. 4,519,764. Specification, US Pat. No. 4,561,486, US Pat. No. 4,561,717, US Pat. No. 5,178,840, German Patent 3210492, European Patent You may use by the method of casting a silicon | silicone on a board | substrate like the method disclosed by the specification of 0079567 and Japanese Patent 6168898. An example of the method is:
1) melting polycrystalline silicon and supplying a molten silicon reservoir to a crucible;
2) applying molten silicon from a crucible onto a substrate, thereby forming a silicon wafer. The substrate can move through the molten silicon, for example, the substrate can be a rotating wafer chuck or other moving substrate. Alternatively, the substrate may be stationary and molten silicon may be supplied thereon. The flowable tip may be used to load or reload the crucible or both.

The flowable tip may be used in sintering methods as disclosed in US Pat. No. 5,006,317 and US Pat. No. 5,499,598. One example of this sintering method is:
1) filling the container with polycrystalline silicon pieces;
2) locally heating the container in a local heating region to melt a portion of the polycrystalline silicon piece to form a sintered portion and a molten portion;
3) The local heating region is moved in the long axis direction of the container to alternately solidify the molten part, melt the sintered part, and form a new sintered part; Forming a step.

  Liquid chips can be found in Chapter 13, pages 497-555 of Crystal Growth (edited by Brian R. Pamplin, Pergamon Press, Ltd., Oxford, 1975). It may be used in various crystal pulling methods as disclosed in the published paper “Crystal Pulling”. Among them, a CZ method using a crucible and a crucible free method such as a pedestal method, a cooling furnace method and a cooling crucible method may be included. Another cooling crucible method is the paper by TFCiszek, “Some Applications of Cold Crucible Technology for Silicon Photovoltaic Material Preparation” (Journal of Electrochemical Society) of the Electrochemical Society), Vol. 132, No. 4, April 1985) ”.

  The invention further relates to a method for reloading a crucible used in the CZ process. The method includes pulling up at least one ingot from the crucible and loading the crucible with a flowable tip to reload the crucible.

  FIG. 1 shows an example of a CZ device 100 in which the present invention can be used. The apparatus 100 comprises a growth chamber 101 below the pulling chamber 102. The growth chamber 101 can be insulated from the chamber 102 by a vacuum valve 103. The growth chamber 101 contains a crucible 104 mounted on a shaft 105. The shaft 105 is rotatably connected to a motor 106 for rotating the shaft 105 and the crucible 104. The crucible 104 is surrounded by a heater 107 that forms a hot zone 114 around the crucible 104. Crucible 104 contains molten silicon 108. The ingot 109 raises the seed 110 from the crucible 104 by immersing the seed 110 in the molten silicon 108 and then pulling the seed 110 and the ingot 109 upward. The seed 110 and the ingot 109 rotate in the opposite direction to the crucible 104. The seed 110 is attached to the seed support cable 111. The seed support cable 111 is pulled up by the lifting mechanism 112. A crystal weight reading device 113 may be attached to the pulling mechanism 112.

  The ingot is removed from the CZ apparatus without introducing air into the growth chamber. Flowable chips can also be added during or after pulling up the ingot. Flowable chips can be added in either continuous mode or batch mode.

  The flowable tip is added to the crucible when the growth chamber is under vacuum and / or under an inert gas atmosphere.

  Flowable chips are added to the crucible while it is hot. While not intending to be limited by theory, it is possible that the crucible may break if the temperature of the crucible is too low.

  Flowable tips can be added to the crucible so as not to contaminate the silicon heel.

  The flowable tip can be added to the crucible in either batch mode or continuous mode. The flowable tip can be added to a crucible containing molten silicon or at least partially solidified silicon heel. The flowable tip can be added using various feeder devices.

  FIG. 2 shows a vibration feeder device 200 for reloading the crucible 104 used in the CZ method as shown in FIG. The vibration feeder device 200 includes a hopper 201 that accommodates the flowable chip 202. The hopper 201 has an input port 203 into which the flowable chip 202 can be input, and an outlet 204 through which the flowable chip 202 exits to the supply tray 205. The supply tray 205 is attached to a vibration feeder 206 that vibrates the supply tray 205. After the flowable chip 202 exits from the outlet 204, it moves along the supply tray 205 to a supply tube 207 extending from the end of the supply tray 205 to the growth chamber 101 of the CZ apparatus. A lance 208 attached to the end of the supply pipe 207 extends from the end of the supply pipe 207 to the crucible 104. At least the input ports of the input port 203, the hopper 201, the vibration feeder 206, and the supply pipe 207 are mounted in a housing 209 that can maintain a vacuum. In order to isolate the housing 209 from the growth chamber 101 while pulling up the ingot (not shown) from the crucible 104, a load isolation lock 210 is provided at the end of the supply tray 205 and the inlet of the supply tube 207.

Flowable chips are:
i) evacuating the hopper containing the flowable chip and / or filling the hopper with an inert gas;
ii) supplying flowable chips from the hopper to the feed device;
iii) vibrating all or part of the vibratory feeder device, thereby moving the flowable tip through the feed device to the crucible;
Can be added to the crucible using a vibration feeder device.

  Inert gas injection into the hopper can be performed by evacuating the hopper and purging the hopper at least once with an inert gas such as semiconductor grade argon, helium or nitrogen.

  The feed device comprises a supply tube, optionally a supply tray from the hopper to the supply tube, and optionally a lance from the supply tube to the crucible. To move the flowable tip, the hopper, feed tray, feed tube, lance or combination thereof can be vibrated.

  Step iii) can also be carried out by vibrating all or part of the vibrating feeder device at the resonant frequency of the flowable tip.

  Flowable chips can be added to the crucible in batch mode. The flowable tip can be added to a crucible containing an at least partially solidified heel. The flowable chips are described in canister feeder devices (eg, US Pat. No. 5,488,924 and the article by Daud, T and Kachare, A, “Improved Czochralski for Photovoltaic Modules”. Advanced Czochralski Silicon Growth Technology for Photovoltaic Modules (DOE / JPL-1012-70, Distribution Category UC-63b, 5101-2-7 Flat Solar Array Project, JPL Publication 82-35, September 1982) Day))).

  FIG. 3 shows an embodiment of a canister feeder apparatus 300 for reloading the crucible 104 used in the CZ method as shown in FIG. The canister feeder device 300 includes a canister or hopper 301 filled with a flowable tip 302. The canister 301 is located in the pulling chamber for the CZ method. The pulling chamber 102 is closed and evacuated. The canister 301 is attached to the cable 303. The cable 303 lowers the canister 301 to the vicinity of the crucible 104, that is, to a level above the at least partially solidified surface 304 of the heel 305 in the crucible 104. The canister 301 has a cone 306 attached to the outlet of the canister 301. Mechanism 307 descends cone 306 down to a level close to but above the at least partially solidified surface 304 of heel 305 in crucible 104. This allows the flowable tip 302 to exit the canister 301 and falls onto the at least partially solidified surface 304 of the heel 305 in the crucible 104.

Flowable chips are:
Optionally, a) at least partially solidifying the heel in the crucible;
b) filling the canister with flowable chips;
c) evacuating the canister;
d) moving the canister to a height above the heel;
e) opening the canister and allowing the flowable chip to discharge the canister into the crucible;
f) Repeat steps b), c), d) and e) until the crucible is filled to the desired depth, and can be added to the crucible using a canister feeder device in the manner provided.

  US Pat. No. 3,998,686, US Pat. No. 4,002,274, US Pat. No. 4,095,329, US Pat. No. 4,176,166, US Patent US Pat. No. 4,312,700, US Pat. No. 4,323,419, US Pat. No. 4,382,838, US Pat. No. 4,394,352, US Pat. No. 4,447,289, U.S. Pat. No. 4,519,764, U.S. Pat. No. 4,557,795, U.S. Pat. No. 4,561,486, U.S. Pat. No. 4,572. No. 4,812, US Pat. No. 4,661,324, US Pat. No. 4,689,109, US Pat. No. 4,968,380, US Pat. No. 5,006,317. Specification, US Pat. No. 5,080 873, US Pat. No. 5,098,229, US Pat. No. 5,161,717, US Pat. No. 5,178,840, US Pat. No. 5,229,082. Specification, US Pat. No. 5,242,667, US Pat. No. 5,254,300, US Pat. No. 5,431,869, US Pat. No. 5,454,424 US Pat. No. 5,462,010, US Pat. No. 5,488,924, US Pat. No. 5,499,598, US Pat. No. 5,510,095, US Patent No. 5,492,079, US Pat. No. 5,690,733, US Pat. No. 5,762,491, US Pat. No. 5,868,835, US Pat. No. 5,902,395, and U.S. Patent No. 6,217,649, a method using a polycrystalline silicon disclosed in which are incorporated by reference herein. Those skilled in the art will appreciate that crucibles can be used in addition to or in place of the silicon starting materials and reload materials described herein, as well as other methods of processing polycrystalline silicon.

<Feeder device>
Instead of granules in a feeder device configured to supply granules, a flowable tip can be used. In a feeder device of any one of a quantitative feeder device, a canister feeder device, a heavy belt feeder device, a vibration feeder device, a chip thruster feeder device, a pneumatic transport feeder device, a stagnant flow feed lance feeder device, a rotating disk feeder device, or an auger feeder device A flowable chip can be used.

  An example of a metering feeder device is the paper “Economic feeder for recharging and“ topping off ”” by Journal of Ficket, B and Mihalik, G (Journal of "Journal of Crystal Growth, Siemens Solar Industry, Vol. 211, pages 372-377, 2000" ", U.S. Pat. No. 3,998,686, U.S. Pat. No. 5,080. No. 5,873, U.S. Pat. No. 5,762,491, and JP-A-62-260791. Examples of canister feeder devices include US Pat. No. 4,394,352, US Pat. No. 4,557,795, US Pat. No. 5,229,082, and US Pat. No. 5,488. 924, the disclosure of which is incorporated herein by reference. An example of a heavy belt feeder device is disclosed in US Pat. No. 6,217,649. Examples of the vibration feeder device are disclosed in US Pat. No. 5,462,010 and Japanese Patent Application Laid-Open No. 02-617197. An example of a chip thruster feeder device is disclosed in US Pat. No. 4,661,324. Examples of pneumatic transport feeder devices are disclosed in US Pat. No. 4,968,380 and US Pat. No. 5,098,229. Examples of stagnant flow delivery lance feeder devices are disclosed in US Pat. No. 5,690,733, US Pat. No. 5,868,835, and US Pat. No. 5,902,395. Yes. Examples of turntable feeder devices are disclosed in US Pat. No. 4,002,274 and US Pat. No. 5,242,667. An example of an auger feeder device is the paper by Daud, T and Kachare, A, “Advanced Czochralski Silicon Growth Technology for Photovoltaic Modules for Photovoltaic Modules (DOE / JPL). -1012-70, Distribution Category UC-63b, 5101-2-7 Flat Solar Array Project, JPL Publication 82-35, September 15, 1982)))). It will be appreciated that the flowable tips disclosed herein can be used in other known feeders of a size suitable for processing polycrystalline silicon.

  The following examples are only intended to illustrate the present invention to those skilled in the art and should not be construed as limiting the scope of the invention.

<Example 1>
WC, Co, GC-712 (manufactured by General Carbide, containing 12% Co and 88% WC, powder having a particle size of 0.6 μm) and sintered WC / Co powder are spread on plastic. Record the initial weight of the powder. An Elite (R) rare earth plate magnet is passed through each powder form at a distance of 2 mm or less. Record the final weight of the powder. The results are shown in Table 1.

  Example 1 shows that impurities can be removed from a piece of polycrystalline silicon when a magnet is used. In particular, WC / Co impurities introduced by the jar crusher of FIG. 4 are removed using the method of Example 1.

<Example 2>
Four examples of flowable chips are produced by the following method. A polycrystalline silicon U-shaped rod-shaped material is obtained from the cold wall of the Siemens type vacuum vessel. After removing the carbon socket, the U-shaped silicon rod is crushed into 10-15 cm pieces using a low contamination impact tool on a polyethylene table. The obtained silicon chunk is supplied to the jar crusher 400 shown in FIG. The width of the discharge slot is limited to 15 mm at the closest distance. As the polycrystalline silicon pieces pass through the discharge slot 418, they pass through an aerated dust collection zone that acts to remove dust from the polycrystalline silicon pieces. The resulting polycrystalline silicon pieces are collected in a polyethylene lined bottle.

  The collected polycrystalline silicon pieces are then ground again with the same equipment. After the pulverization stage, the flow rate of the silicon pieces is adjusted, and the collected polycrystalline silicon pieces are transferred to a UHMWPE hopper that serves to feed on the step deck sorter 500 shown in FIGS. Both the hopper 502 and the step deck sorter 500 are attached to a vibration table 501 that operates to move the polycrystalline silicon piece. The deck is adjusted so that the product size is usually maintained in the range of 1 mm to 12 mm. Polycrystalline silicon pieces longer than 12 mm in length are removed from the product and returned to the grinding process for additional grinding passes so that the crushed material is mixed with the polycrystalline silicon pieces exposed to a single grinding pass. Will be. This procedure is repeated many times until a 300 kg piece of polycrystalline silicon is obtained. During a series of treatments, four samples of polycrystalline silicon pieces are collected in an acid treated PTFA container. These samples are submitted for surface metal analysis by the method described in US Pat. No. 5,851,303. The purity of the obtained silicon is shown in Table 2 below. The control particle size distribution of each sample of polycrystalline silicon pieces is 1 mm to 12 mm.

<Example 3>
Polycrystalline silicon U-shaped rods are obtained from the cold water wall of a Siemens type bell jar reactor. After removing the carbon socket end, the U-shaped rod-shaped silicon is crushed into 4-inch pieces on a polyethylene table using a low contamination impact tool. The obtained silicon chunk is supplied to the jar crusher 400 shown in FIG. The width of discharge slot 418 is limited to 15 mm at the closest distance. As the polycrystalline silicon pieces pass through the discharge slot 418, they pass through an aerated dust collection zone that acts to remove dust from the polycrystalline silicon pieces. The resulting polycrystalline silicon pieces are collected in polyethylene-lined bottles. The collected polycrystalline silicon pieces are ground again with the same equipment. After the grinding step, the collected polycrystalline silicon pieces are transferred to a UHMWPE hopper which acts to adjust the flow rate of the silicon pieces to the step deck sorter 500 of FIG. The deck is adjusted so that the product particle size distribution is normally maintained in the range of 1 mm to 12 mm. A piece of polycrystalline silicon weighing 40 kg is processed. Bulk impurities (boron, total donors, carbon, phosphorus, iron, nickel, copper, and chromium) and surface impurities (iron, nickel, copper, chromium, sodium, and zinc) are measured. For surface purity analysis, four samples of polycrystalline silicon pieces are collected in a pickled PTFA container.

  These samples analyze bulk impurities and surface impurities. Bulk metal values are obtained from a floating zone core etched using acid decomposition of a frozen probe. The core is obtained from a polycrystalline silicon rod. The procedure for obtaining bulk metal values is described in US Pat. No. 4,912,528. Metal concentration is measured from atomic absorption analysis.

  Carbon is measured from the floating zone core. Take a slice from the core. The slice is lapped and polished. The carbon concentration in silicon is measured using Fourier transform infrared spectroscopy.

  Phosphorus, boron, aluminum, and arsenic are measured from the floating zone core using a technique known as dispersive photoluminescence (PL). This test is used to chemically polish single crystal silicon slices from the floating zone core. By cooling the slice to the liquid helium temperature, an argon laser is used to cause photon emission in the sample. The measured intensity of electron-hole pair recombination radiation is used to determine the concentration of these impurities.

  The donor is a calculated value obtained from measuring the resistance of the silicon core.

  Surface impurities are measured by the method of US Pat. No. 5,851,303. The purity of the obtained silicon is shown in Tables 3 and 4 below. The control particle size distribution of each sample of polycrystalline silicon pieces is 1 mm to 12 mm.

DESCRIPTION OF SYMBOLS 100 CZ apparatus 101 Growth chamber 102 Lifting chamber 103 Vacuum valve 104 Crucible 105 Shaft 106 Motor 107 Heater 108 Molten silicon 109 Ingot 110 Seed 111 Seed support cable 112 Lifting mechanism 113 Crystal weight reading device 114 Hot zone 200 Vibrating feeder device 201 Hopper 202 Flow Insert 203 204 outlet 205 supply tray 206 vibration feeder 207 supply pipe 208 lance 209 housing 210 loading isolation lock 300 canister feeder device 301 canister or hopper 302 fluidity tip 303 cable 304 surface 305 heel 306 cone 307 mechanism 400 jaw crusher 401 Frame assembly 402 Fixed jaw plate 403 Moving jaw plate 404 Jaw cavity 405 Pitman carrier assembly 406 Pitman bearing 407 Eccentric shaft 408 Tension rod pin 409 Flywheel 410 Motor 411 Belt 412 Base 413 Tension rod 414 Adjustment wheel 415 Outer spring collar 416 Tension spring 417 Outer spring collar 418 Release slot wedge 419 Release slot collar 419 Adjustment rod 420 Adjustment wheel 421 Cross bar 422 Adjustment wedge 423 Bearing wedge 424 Toggle plate 425 Hopper 500 Step deck sorter 501 Vibration motor assembly 502 Input port 503 Fluidized bed region 504 Air flow 505 Perforated plate 506 Groove formation region 507 Gap 508 Conveyor 509 Collection container 510 2nd deck 511 2nd deck Inlet end 512 groove 513 lump 514 fragment 515 thin piece 516 gap 517 third deck 518 second deck outlet 519 ridge 520 valley 521 deck 522 gap 523 deck 524 gap 525 deck 526 collection container 528 collection container 528 collection container 528 collection container 528 531 First Deck 532 Dust Collector 533 Deck 534 Gap

Claims (35)

  A group of polycrystalline silicon pieces made of a polycrystalline silicon piece manufactured by chemical vapor deposition and having a bulk impurity level not exceeding 0.03 ppma and a surface impurity level not exceeding 15 ppba, A flowable chip comprising a group of polycrystalline silicon pieces having a controlled particle size distribution and generally having a non-spherical morphology.   Bulk impurities are boron in an amount of 0.06 ppba or less, donors in an amount of 0.30 ppba or less, phosphorus in an amount of 0.02 ppba or less, carbon in an amount of 0.17 ppma or less, and total bulk metal impurities are 4 The flowable chip according to claim 1, which has an amount of not more than 0.5 ppba.   The flowable chip of claim 1, wherein all bulk metal impurities are present in an amount of 1 ppba or less.   The total bulk metal impurity comprises Cr having an amount of 0.01 ppba or less, Cu having an amount of 0.01 ppba or less, Fe having an amount of 0.01 ppba or less, and Ni having an amount of 0.01 ppba or less. The flowable chip described in 1.   Surface impurities are Cr with an amount of 0.06 ppba or less, Cu with an amount of 0.15 ppba or less, Fe with an amount of 10 ppba or less, Na with an amount of 0.9 ppba or less, and Ni with an amount of 0.1 ppba or less. The flowable chip according to claim 1, wherein Zn is an amount of 0.6 ppba or less.   The flowable tip of claim 1 is used in an application selected from initial loading maximization, initial filling, reloading, reloading maximization, reloading filling, and combinations thereof. A method with to do.   The flowable chip according to claim 1 is casted on a substrate by a solar cell casting method, a constant forming length method, an induction plasma method, an electron beam melting method, a heat exchange method, a string ribbon method, and a substrate. And a method selected from the group consisting of a sintering method and a crystal pulling method. a) pulling the silicon ingot from the crucible by the Czochralski method;
b) adding a flowable tip to the molten silicon in the crucible, wherein the flowable tip is manufactured by chemical vapor deposition and has a low level of bulk impurities and a low level of surface impurities A stage comprising a group of polycrystalline silicon pieces composed of silicon pieces and having a controlled particle size distribution and generally having a non-spherical morphology;
c) optionally adding a dopant to the crucible.   The method of claim 8, wherein step a) and step b) are carried out simultaneously.   The process according to claim 9, wherein step b) is carried out continuously.   9. A method according to claim 8, wherein step a) is carried out before step b).   The method according to claim 11, wherein step b) is carried out batch by batch.   Step b) is a metering feeder device, a canister feeder device, a heavy belt feeder device, a vibration feeder device, a chip thruster feeder device, a pneumatic transport feeder device, a stagnant flow feed lance feeder device, a rotating disk feeder device, or an auger feeder device. The method according to claim 8, wherein the method is performed using any one of the feeder devices. Step b) is performed using a vibration feeder device.
i) evacuating the hopper containing the flowable chip and / or filling the hopper with an inert gas;
ii) supplying flowable chips from the hopper to the feed device;
9. A method according to claim 8, wherein the method comprises: iii) vibrating all or part of the vibratory feeder device, thereby moving the flowable tip through the feed device to the crucible.   The method of claim 8, wherein the flowable chip has a particle size distribution of 0.2 mm to 45 mm.   The method of claim 8, wherein the flowable chip has an irregular morphology.   Whether the flowable chip is boron below 20 ppba, donor below 0.30 ppba, phosphorus below 0.02 ppba, carbon below 0.17 ppma, or below 4.5 ppba. 9. The method of claim 8, having low levels of bulk impurities, either bulk metal impurities or a combination thereof.   9. The method of claim 8, wherein the flowable chip has a low level of surface impurities that are Co, Cr, Cu, Fe, Na, Ni, W or Zn of 30 ppba or less. a) lifting the silicon ingot from the crucible by the Czochralski method and leaving the heel in the crucible;
Optionally b) solidifying at least the surface of the heel;
c) adding a flowable tip to the surface of the heel, wherein the flowable tip is manufactured by chemical vapor deposition and has low levels of bulk impurities and low levels of surface impurities And comprising a group of polycrystalline silicon particles having a controlled particle size distribution of 1 mm to 12 mm and generally having a non-spherical morphology;
d) optionally adding a dopant to the crucible.   The method according to claim 19, wherein step c) is carried out batch by batch.   Step c) comprises a metering feeder device, a canister feeder device, a heavy belt feeder device, a vibration feeder device, a chip thruster feeder device, a pneumatic transport feeder device, a stagnant flow feed lance feeder device, a rotating disk feeder device, or an auger feeder device. The method according to claim 19, wherein the method is carried out using any one of the feeder devices. Step c)
i) filling the canister with flowable chips;
ii) evacuating the canister;
iii) moving the canister to a height above the heel;
iv) opening the canister and releasing the flowable tip from the canister into a crucible;
v) repeating steps i), ii), iii) and iv) until the crucible is filled to the desired depth;
The method of claim 19, wherein the method is performed by a technique comprising   20. The method of claim 19, wherein the flowable chip has a controlled particle size distribution of 4 mm to 10 mm.   20. The method of claim 19, wherein the flowable chip has an irregular morphology.   Whether the flowable chip is boron at a level of 20 pbba or less, a donor at a level of 0.30 ppba or less, phosphorus at a level of 0.02 ppba or less, carbon at a level of 0.17 ppma or less, or a level of 4.5 ppba or less. 20. The method of claim 19, having low levels of bulk impurities, either bulk metal impurities or a combination thereof.   20. The method of claim 19, wherein the flowable tip has a surface impurity level that is Co, Cr, Cu, Fe, Na, Ni, W or Zn of 30 ppba or less. 1) introducing a semiconductor material comprising a flowable chip according to claim 1 into a mold comprising walls defining a desired cross-sectional shape;
2) melting the semiconductor material;
And 3) solidifying the semiconductor material after step 2) to produce a cast ingot having a desired cross-sectional shape. 1) continuously supplying the flowable chip according to claim 1 to a bottomless container disposed in the induction coil;
2) melting a continuously supplied flowable chip;
Optionally 3) blowing hot plasma gas over the surface of the melt;
4) A step of continuously releasing the solidified silicon from the bottomless container downstream.   A method comprising pulling a silicon ribbon from a molten silicon meniscus defined at an edge of a die under the condition that molten silicon is produced by melting the flowable tip of claim 1. 1) melting the flowable tip of claim 1 using a high frequency plasma torch;
And 2) directing the product of step 1) to a water-cooled crucible or onto a growing crystalline silicon body. 1) melting the flowable tip according to claim 1 with an electron beam;
2) casting the product of step 1). A method comprising the steps of manufacturing an ingot in a HEM furnace,
The HEM furnace comprises a chamber containing a crucible surrounded by a heating element having a helium heat exchanger connected to the bottom of the crucible:
1) placing the flowable chip according to claim 1 on top of the seed crystal and filling the crucible;
2) evacuating the chamber;
3) heating the heating element to melt the flowable tip;
4) flowing helium gas through the heat exchanger, thereby avoiding melting of the seed crystal;
5) gradually increasing the gas flow, thereby gradually solidifying the silicon and growing the crystal outside the seed crystal. Under the condition that the flowable chip according to claim 1 is initially loaded with a shallow silicon melt and / or reloaded with the flowable chip according to claim 1 or both. so,
1) pulling up two strings and seed crystal vertically from said shallow silicon melt;
2) Wetting the strings and seed crystals with molten silicon from the shallow silicon melt to fill the spaces between the strings;
3) cooling the product from step 2) to produce a silicon ribbon. Under the condition that the flowable tip of claim 1 is loaded into a crucible, or the crucible is reloaded with the flowable tip of claim 1, or both.
1) melting the flowable tip and supplying a molten silicon pool to the crucible;
2) applying molten silicon from a crucible onto a substrate, thereby producing a silicon wafer. 1) filling the container with the flowable tip according to claim 1;
2) melting the part of the polycrystalline silicon piece to form a sintered part and a molten part by locally heating the container in a local heating region;
3) The local heating region is moved in the long axis direction of the container to alternately solidify the molten part, melt the sintered part, and form a new sintered part; Forming an ingot.

Images