METHOD FOR PRODUCING SILICON FRAGMENTS

Abstract

The present disclosure relates to a method for producing polycrystalline silicon fragments. The process includes (a) providing a polycrystalline silicon rod, (b) working the surface of the silicon rod by means of a hammer or needle hammer to remove at least a portion of a layer of the surface of the polycrystalline silicon rod, and (c) reducing the silicon rod to fragments. Wherein an amount of impact energy expended by the hammer and/or needle hammer is from 1 J to 15 J.

Claims

1-11. (canceled)

12. A process for producing silicon chunks, comprising the steps of: a) providing a silicon rod; b) treating a surface of the silicon rod with a hammer and/or a needle hammer for at least partial removal of a layer of the surface of the silicone rod; c) comminuting the silicon rod into chunks using a thermal crushing process or a high-voltage impulse crushing or an apparatus selected from the group comprising jaw crusher, roller crusher, chisel and sledgehammer; and wherein the surface layer is removed to a depth of 1 to 10 mm by an impact energy applied with the hammer and/or needle hammer of 1 to 15 J.

13. The process of claim 12, wherein the impact energy applied with the hammer and/or needle hammer is 2 to 10 J, preferably 3 to 8 J.

14. The process of claim 12, wherein the needle hammer comprises 6 to 24, preferably 6 to 18, particularly preferably 6 to 12, needles.

15. The process of claim 12, wherein the needles of the needle hammer each have a circular impact surface having a radius of 0.5 to 1.5 mm, particularly preferably of 0.5 to 1 mm.

16. The process of claim 12, wherein the needle hammer is an electrically, pneumatically or hydraulically operated needle hammer.

17. The process of claim 15, wherein the needle hammer is operated with an impact speed of 2000 to 5000 min.sup.−1, preferably 2500 to 4000 min.sup.−1, particularly preferably 2800 to 3500 min.sup.−1.

18. The process of claim 12, wherein the needles of the needle hammer and/or at least the polysilicon-contacting part of the hammer is made of a material selected from the group comprising carbide, metal-ceramic, ceramic and combinations thereof.

19. The process of claim 12, wherein the needles of the needle hammer and/or at least the polysilicon-contacting part of the hammer is made of a material selected from the group comprising tungsten carbide, tungsten carbide with a cobalt binder, tungsten carbide with a nickel binder, titanium carbide, chromium carbide (for example Cr.sub.3C.sub.2) with a nickel-chromium alloy binder, tantalum carbide, niobium carbide, silicon nitride, silicon carbide in a matrix (e.g. Fe, Ni, Al, Ti or Mg), aluminum nitride, titanium carbide with cobalt and titanium carbonitride, nickel, nickel-cobalt alloy, iron and combinations thereof.

20. The process of claim 12, wherein the surface layer is removed to a depth of 1 to 5 mm, preferably 1 to 3 mm.

21. The process of claim 12, wherein the roughness parameters Ra and/or RPc according to DIN EN ISO 4287/4288 and DIN EN 10049 of the silicon rod are determined between steps a) and b) and/or between steps b) and c).

22. The process of claim 12, wherein the step b) is repeated after a determination of the roughness of the treated silicon rod.

Description

[0055] FIG. 1: Manual needle hammer according to the invention

[0056] FIG. 2: Detailed representation of a needle

[0057] FIG. 1 shows a manual needle hammer 1 having a handle 2 and a head 3. Both the handle 2 and the head 3 have a stainless steel core sheathed in a plastic (for example polyethylene, polypropylene or polyurethane). The head 3 has a first and a second flat, rectangular front end (face) 4, 5. Eight needles 6 are arranged on the first front end 4 and 12 needles 7 are arranged on the second front end 5. The needles 6, 7 may be screwed together with the stainless steel core for example and are not covered with the plastic. A detailed representation of a needle 6 is shown in FIG. 2. The needles 6 have a larger diameter than the needles 7. A symmetric arrangement of the needles 6, 7 is not compulsory. A typical hammer 1 has a weight of 1 kg, for example, and the needles 6, 7 have a diameter d.sub.1 of 4 mm or 3 mm.

[0058] FIG. 2 shows a detailed representation of a needle 6 from FIG. 1. The needle 6 has a cylindrical part 8 and a point 9. The cylindrical part 8 has a diameter d.sub.1 of for example 4 mm. The point 9 has the shape of a conical frustum and the circular top surface 10 constitutes the impact surface. The top surface 10 typically has a diameter d.sub.2 of 1 mm. The height h of the needle 6 is about 12 mm.

Example

[0059] 5 inventive treatment tools (hammers and needle hammers, lines 1 to 5) and 2 manual hammers (lines 6 and 7) were tested for removal of popcorn on polysilicon rods having a diameter of about 250 mm. The rods originated from the same batch (outer rod circle) from a Siemens reactor having an output of 24 rod pairs. All rods exhibited popcorn in the upper region where the bridge was previously located (the depth of the trenches was up to 4 mm).

[0060] The rods were weighed before and after treatment to calculate yield. Determination of the treated rod area was carried out using a stopwatch and before and after images, and image processing software was utilized for surface determination.

[0061] Contamination was determined using ICP-MS as described. The impact energies of the pneumatic needle hammers were predefined on the instrument. The impact energies for the manual hammers are average values determined by means of piezoelectric force sensors. The hammer was clamped in an apparatus and dropped onto the measuring apparatus from different heights. The test setup was conceived such as to allow reproducible repetition of the test results. Selectivity was determined on the basis of topographic images captured after treatment (by laser-optical scanning), wherein “++” corresponds to very uniform material removal with complete popcorn removal and “−” corresponds to irregular material removal with a high loss of target material. “−−” corresponds to fracture of the polysilicon rod (see table 1).

[0062] The needles of the pneumatic needle hammer comprised a tungsten carbide cobalt (WC—Co) hard metal having a Co proportion of 10%. The particle size of the WC was 0.6 μm. The needles had a round impact surface having a diameter of 1 mm and an overall diameter of 3 mm. The needles of the manual needle hammer (cf. FIG. 1) were likewise made of WO—Co hard metal as described hereinabove. The impact surface of the needles was identical to the needles of the pneumatic needle hammers. The hammer from line 1 was a hammer according to EP 0 539 097 A1 whose impact surface was made of a WO—Co hard metal as described hereinabove. The heads of the manual hammers from lines 6 and 7 were made of a WC—Co hard metal having a Co proportion of 10% and a WC particle size of 2.5 to 4 μm. The impact surface was round and minimally curved outwards.

TABLE-US-00001 TABLE 1 Yield [%] (proportion Treated Con- of rod Treatment Impact tamination target area/ # tool energy W [pptw] Selectivity material) second 1 hammer 10 to   50 pptw +  75%   5 cm.sup.2 (1 kg) 15 J 2 pneumatic 6 J  <10 pptw ++ >95% 1-5 cm.sup.2 needle hammer (6 needles) 3 pneumatic 6 J  <10 pptw ++ 90%  2-8 cm.sup.2 needle hammer (12 needles) 4 pneumatic 6 J   20 pptw ++ 85% 5-10 cm.sup.2 needle hammer (24 needles) 5 manual 10 to  <10 pptw ++ >95%  1-2 cm.sup.2 needle 15 J hammer (9 needles, 1 kg) 6 hammer 100 J  200 pptw − 50%  20 cm.sup.2 (2.5 kg) 7 hammer 300 J 1000 pptw −− 20% − (6 kg)

[0063] It is readily apparent from the examples that using pneumatic needle hammers having 6 or 12 needles results in very good selectivity coupled with very low contamination with tungsten. The yield is even greater for 6 needles than for 12, but this is associated with a longer working time. Economy here especially also depends on the type (the required quality) of the produced polysilicon rod.

[0064] A very good yield and selectivity coupled with only low contamination is likewise realized with the manual needle hammer (cf. FIG. 1). Depending on the fall height, the impact energy varies here within narrow limits. At low impact energy the known hammer of line 1 may also be an alternative to the previously described variants, since contamination is within an acceptable range and relatively large areas can be treated in a short time. This hammer may be an advantageous alternative especially for spot-removal of popcorn.

[0065] Selective removal of the popcorn is not possible with the manual hammers of lines 6 and 7. The 2.5 kg hammer caused very large areas of spalling and thus very imprecise removal of the popcorn layer. The use of the 6 kg hammer resulted in fracture of the rods. Removal of the layer was not possible.