Classifying polysilicon

Abstract

A method for mechanically classifying polycrystalline silicon chunks or granules with a vibratory screening machine, involves setting silicon chunks or granules present on one or more screens each comprising a screen lining in vibration such that the silicon chunks or silicon granules perform a movement which causes the silicon chunks or silicon granules to be separated into various size classes, wherein a screening index is greater than or equal to 0.6 and less than or equal to 9.0.

Claims

1. A method for decreasing an overlap of particle sizes between adjacent size fractions in mechanically classifying polycrystalline silicon chunks or granules with a vibratory screening machine, comprising: introducing the chunks or granules into a gravity screening machine; setting the silicon chunks or granules present on one or more screens of the gravity screening machine in vibration, each screen comprising a screen lining such that the silicon chunks or silicon granules perform a throwing movement and which causes the silicon chunks or silicon granules to be separated into various size classes, wherein the screening index is greater than or equal to 1.6 and less than or equal to 3.0, and the screens of the gravity screening machine are characterized by an amplitude of vibration of 0.5 to 8 mm, a speed of rotation of 400 to 2000 rpm and a throwing angle of 30 to 60 relative to a screen plane, with the screen plane inclined by an angle of 0 to 15 relative to the horizontal.

2. The method of claim 1, wherein the screens of the gravity screening machine are characterized by an amplitude of vibration of 1 to 4 mm, a speed of rotation of 600 to 1500 rpm and a throwing angle of 40 to 50 relative to a screen plane, with the screen plane inclined by an angle of 0 to 10 relative to the horizontal.

3. The method of claim 1, wherein the screening machine comprises a plurality of screen decks arranged one on top of another.

4. The method of claim 2, wherein the screening machine comprises a plurality of screen decks arranged one on top of another.

5. The method of claim 1, wherein the screen linings are each secured on a frame of plastic or a frame comprising a plastic lining.

6. The method of claim 1, wherein one or more of the screen linings consist of an elastomer having a Shore A hardness of greater than 65 or have a surface composed of an elastomer having a Shore A hardness of greater than 65.

7. The method of claim 2, wherein one or more of the screen linings consist of an elastomer having a Shore A hardness of greater than 65 or have a surface composed of an elastomer having a Shore A hardness of greater than 65.

8. The method of claim 3, wherein one or more of the screen linings consist of an elastomer having a Shore A hardness of greater than 65 or have a surface composed of an elastomer having a Shore A hardness of greater than 65.

9. The method of claim 1, wherein one or more of the screen linings or the surfaces of one or more of the screen linings and all the further components and linings thereof that come into contact with the chunk silicon or granular silicon consist of plastics having a total contamination of less than 2000 ppmw.

10. The method of claim 3, wherein one or more of the screen linings or the surfaces of one or more of the screen linings and all the further components and linings thereof that come into contact with the chunk silicon or granular silicon consist of plastics having a total contamination of less than 2000 ppmw.

11. The method of claim 1, wherein perforated silicon fillets are used in one or more of the screen linings, holes in the silicon fillets, at least in part, having a conical shape.

12. The method of claim 1, wherein both perforated silicon fillets and plastic are used as screen linings, wherein a screen with a perforated Si fillet is used at least in a first screening step.

13. The method of claim 1, wherein the screens have orifices through which silicon chunks or granules smaller than the size of the orifices pass, the size of the orifices increasing in an outlet direction of the gravity screening machine.

14. The method of claim 1, further comprising dedusting the silicon chunks or granules being classified to remove fine silicon particles having sizes of less than 10 m by directing a gas flow through the gravity screening machine to an off-gas exit, the gas velocity such that the fine silicon particles are entrained in the gas.

15. The method of claim 1, wherein the amplitude of vibration of screens of the gravity screening machine decreases in the direction of an outlet of the gravity screening machine.

16. The method of claim 1, wherein the weight percent overlap of one classified fraction with a next larger or smaller classified fraction is 5 weight percent or less based on the total weight of the one classified fraction.

Description

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(1) The screening index is defined as the ratio of the acceleration generated by the screening motion to the acceleration due to gravity vertical to the screening plane:
K.sub.v=r*.sup.2*sin(+)/(g*cos()),
where
r: amplitude of vibration;
: angular velocity;
: throwing angle;
angle of screen inclination;
g: gravitational constant.

(2) This indicates the maximum vertical acceleration of an object relative to the earth's gravitational acceleration g. If the screening index is <1, there is pure sliding motion (without throwing motion), since the resulting vertical acceleration is smaller than gravitational acceleration. For A throwing motion, the screening index must be >1.

(3) It has been found that, surprisingly, both processes having a screening index of less than 0.6 and processes having a screening index of greater than 9.0 result in much poorer screening results than within the inventive range of 0.6-9.0.

(4) Preferably, the screening index is greater than or equal to 0.6 and less than or equal to 5.0. Classifying at a screening index of 0.6 to 5.0 achieved a further improvement in the screening results. More particularly, the separation sharpness is better than at a screening index of greater than 5.0.

(5) More preferably, the motion of chunk or granular silicon is a throwing motion, with a screening index of 1.6 to 3.0. It has been found that another improvement in screening results, more particularly an even higher separation sharpness between the different size classes, is achieved as a result.

(6) The amplitude of vibration is preferably 0.5 to 8 mm, more preferably 1 to 4 mm. The speed of rotation /2 is preferably 400 to 2000 rpm, more preferably 600 to 1500 rpm. The throwing angle is preferably 30 to 60, more preferably 40 to 50, and the angle of screen inclination relative to the horizontal is preferably 0 to 15, more preferably 0 to 10.

(7) The screening machine preferably comprises a feed region in which the screening material is introduced, and an outlet region in which classified screening material is conducted away.

(8) Preferably, the size of the screen orifices increases in the outlet direction. Fractions/chunk sizes are preferably separated by means of outlets arranged in series.

(9) Preferably, the screening machine comprises screen decks arranged one on top of another. This has the advantage that large chunks cannot damage fine-mesh screen linings. Preferably, fractions/chunk sizes are separated by outlets arranged one on top of another.

(10) Preferably, the screening machine comprises a frame/screen system. This enables rapid screen changing. Monitoring of any contamination is also facilitated. A frame/screen system of this kind comprises screw connection, adhesive bonding, insertion or casting of screen linings in frames, the frames consisting of wear-resistant plastic (preferably PP, PE, PU), optionally with steel reinforcement, or at least being lined with wear-resistant plastic. The frames are preferably sealed by being braced vertically. It is thus possible to avoid contamination and material loss.

(11) It is preferable to use screen linings of particularly wear-resistant plastics, namely elastomers having a Shore A hardness of greater than 65, more preferably having a Shore A hardness of greater than 80. Shore hardness is defined in standards DIN 53505 and DIN 7868. It is possible here for one or more screen linings or surfaces thereof to consist of such an elastomer.

(12) Either one or more screen linings or surfaces thereof or all the components and linings that make contact with the product preferably consist of plastics having a total contamination (metals, dopants) of less than 2000 ppmw, preferably less than 500 ppmw and more preferably less than 100 ppmw.

(13) The maximum contamination of the plastics with the elements Al, Ca, P, Ti, Sn and Zn should be less than 100 ppmw, more preferably less than 20 ppmw.

(14) The maximum contamination of the plastics with elements Cr, Fe, Mg, As, Co, Cu, Mo, Sb and W should be less than 10 ppmw, more preferably less than 0.2 ppmw. The contaminations are determined by means of ICP-MS (mass spectrometry with inductively coupled plasma).

(15) Preferably, the screen linings made of plastics comprise a reinforcement or filling composed of metals, glass fibers, carbon fibers, ceramic or composite materials for stiffening.

(16) Preferably, the screening material is dedusted. The mechanical screening mobilizes the majority of the fine dust adhering to the bulk material on the individual screen decks. This effect is utilized in the invention in order to dedust the bulk material during the screening process.

(17) What is important here is that the fine dust released is transported into an offgas pathway through an appropriate gas flow, in order that it cannot get back into the product. The gas flow can be generated either by suction or by a gas purge. Suitable sifting gases are cleaned air, nitrogen or other inert gases. In the screening machine, there should be a gas velocity of 0.05 to 0.5 m/s, more preferably of 0.2 to 0.3 m/s. A gas velocity of 0.2 m/s can be established, for example, with a gas throughput or a suction performance of 720 m.sup.3 (STP)/h per m.sup.2 of screen area. Fine dust is understood to mean particles smaller than 10 m.

(18) As well as dedusting in the screening machine, dedusting is optionally conducted by means of countercurrent wind sifting in the removal lines for the individual screen fractions. This involves feeding in the sifting gas in the lower region of the removal lines and conducting the dust-laden offgas away in the upper region, immediately upstream of the screening machine. Useful sifting gases are again the abovementioned media. The advantage of this dedusting method is that the sifting stream can be matched to the particle size of the screen fraction. In the case of a coarse screen fraction, it is possible, for example, to set a high sifting flow rate without discharging fine product as well. This gives a very good dedusting outcome and the desired low fine dust fraction in the product.

(19) Preferably, the rotational speed is increased temporarily up to 4000 rpm, in order to free the screen linings from lodged grains. For this purpose, it is alternatively also possible to increase the amplitude of vibration temporarily to up to 15 mm. It is likewise preferable to use impact balls made from plastic or ultrapure silicon, in order to free the screen linings from lodged grains.

(20) Preferably, the amplitude of vibration decreases toward the outlet. More preferably, the ratio of the amplitude of vibration at the exit is up to 50% lower than at the inlet. It has been found that this can further reduce both wear and product contamination.

(21) Useful types of drive for the screening machine include linear, circular or elliptical oscillators. The drive preferably provides a vertical acceleration component in order to reduce screen wear and avoid lodged particles.

(22) It is preferable to use particular shapes for the screen orifices.

(23) Advantageous shapes have been found to be rectangular orifices. Lower wear is found as a result of smaller contact areas. Lodged/jammed grains can be avoided more easily. Round orifices, in contrast, lead to a higher separation sharpness with respect to particle size. Square orifices are likewise preferable. These can combine advantages of rectangular and round orifices.

(24) Preferably, the screen trough and the screen outlets are lined completely on the inside with silicon or with a thermoplastic or elastomer.

(25) Steel base structures of the screening machines are preferably provided with welded PP lining segments. Preference is also given to the use of inner PU linings.

(26) Particularly suitable lateral linings have been found to be steel-reinforced PU castings.

(27) The screen frames can preferably be fixed using quick-release devices.

(28) It is also preferable to use perforated silicon fillets as the screen lining. It is possible for one or more screen linings to be configured in this way. These preferably comprise square bars of ultrapure silicon provided with holes. These holes preferably have a conical shape at least in part, meaning that a cross-sectional area at the top is smaller than at the bottom. This contributes to avoidance of lodged grains. The cone preferably has an angle of 1 to 20, more preferably 1 to 5. Preferably, edge rounding of the holes with a radius of 0.1 to 2 mm is provided at the top of the screen, in order to prevent loss of material and wear, which would lead to deterioration in the separation sharpness. Preferably, only the lower part of each hole is conical and the other part is cylindrical, in order that the hole is not widened too quickly as a result of wear.

(29) Preference is given to providing plastic-sheathed metal support fillets for stabilization in the event of fracture of the Si fillets, for avoidance of contamination and for safeguarding against losses of chunks in the event of fillet fracture.

(30) Preferably, individual Si fillets are equipped with concluding cemented carbide fillets, which are clamped horizontally or vertically. Thus, inexpensive exchange of individual fillets according to wear is possible. The cemented carbide used is preferably WC, SiC, SiN or TiN.

(31) Preferably, the perforated Si screen is laid onto, bonded to or screwed onto a substrate. This enables higher strength; larger areas and the use of thinner or thicker screens is possible. Fracture is easier to avoid.

(32) It is most preferable to use both perforated Si screens and screens made from plastic or screens having a plastic lining.

(33) Preferably, the first screen cut used is a perforated Si screen having a hole diameter of 5 mm to 50 mm. In this case, the large chunks are able to clear away jammed grains and hence prevent blockage. For further separation of the fines fractions, one or more screens made from plastic or having plastic linings are used.

(34) Preferably, for chunk silicon having particle sizes of greater than 15 mm (max. particle length), an additional pre-screen having a plastic lining and having a mesh ratio relative to the screen deck beneath of 1.5:1 to 10:1 is used. This can reduce plastic wear on the lower screen deck. The outputs from the two screen decks are combined. The pre-screen deck preferably has a lower screen stress. This serves to minimize wear.

(35) The method of the invention (throwing motion, screen index 1.6-3.0) leads to polycrystalline silicon chunks having a sharp particle size distribution without any great overlap, or to polycrystalline silicon granules classified with a high separation sharpness, which was not achievable as such in the prior art to date.

(36) The invention therefore also relates to classified polycrystalline silicon chunks, characterized by a particle size classification into chunk size classes 2, 1, 0 and F, where the following applies to the chunks: chunk size 2 has max. 5% by weight smaller than 11 mm and max. 5% by weight larger than 27 mm; chunk size 1 has max. 5% by weight smaller than 3.7 mm and max. 5% by weight larger than 14 mm; chunk size 0 has max. 5% by weight smaller than 0.6 mm and max. 5% by weight larger than 4.6 mm; chunk size F has max. 5% by weight smaller than 0.1 mm and max. 5% by weight larger than 0.8 mm.

(37) The chunk size is defined as the longest distance between any two points on the surface of a silicon chunk (=max. length).

(38) The following chunk sizes are found: chunk size F (CS F) in mm: 0.1 to 0.8; chunk size 0 (CS 0) in mm: 0.6 to 4.6; chunk size 1 (CS 1) in mm: 3.7 to 14; chunk size 2 (CS 2) in mm: 11 to 27.

(39) In each case, at least 90% by weight of the chunk fraction is within the size range mentioned. This results in an overlap range of the 5% by weight quantile of the coarse chunk size to the 95% by weight quantile of the fine chunk size of:

(40) chunk size 2 to chunk size 1: max. 3 mm;

(41) chunk size 1 to chunk size 0: max. 0.9 mm;

(42) chunk size 0 to chunk size F: max. 0.2 mm.

(43) The polycrystalline silicon chunks having the improved particle size classification preferably have very low surface contamination:

(44) Tungsten (W):

(45) chunk size 1100,000 pptw, more preferably 20,000 pptw;

(46) chunk size 01,000,000 pptw, more preferably 200,000 pptw;

(47) chunk size F10,000,000 pptw, more preferably 2,000,000 pptw;

(48) Cobalt (Co):

(49) chunk size 25000 pptw, more preferably 500 pptw;

(50) chunk size 150,000 pptw, more preferably 5000 pptw;

(51) chunk size 0500,000 pptw, more preferably 50,000 pptw;

(52) chunk size F5,000,000 pptw, more preferably 500,000 pptw;

(53) Iron (Fe):

(54) chunk size 250,000 pptw, more preferably 1000 pptw;

(55) chunk size 1500,000 pptw, more preferably 10,000 pptw;

(56) chunk size 05,000,000 pptw, more preferably 100,000 pptw;

(57) chunk size F50,000,000 pptw, more preferably 1,000,000 pptw;

(58) Carbon (C):

(59) chunk size 21 ppmw, more preferably 0.2 ppmw;

(60) chunk size 110 ppmw, more preferably 2 ppmw;

(61) chunk size 0100 ppmw, more preferably 20 ppmw;

(62) chunk size F1000 ppmw, more preferably 200 ppmw;

(63) Cr, Ni, Na, Zn, Al, Cu, Mg, Ti, K, Ag, Ca, Mo, for each individual element:

(64) chunk size 21000 pptw, more preferably 100 pptw;

(65) chunk size 12000 pptw, more preferably 200 pptw;

(66) chunk size 010,000 pptw, more preferably 1000 pptw;

(67) chunk size F100,000 pptw, more preferably 10,000 pptw;

(68) Fine dust (silicon particles having a size of less than 10 m):

(69) chunk size 25 ppmw, more preferably 2 ppmw;

(70) chunk size 115 ppmw, more preferably 5 ppmw;

(71) chunk size 025 ppmw, more preferably 10 ppmw;

(72) chunk size F50 ppmw, more preferably 20 ppmw.

(73) The invention also relates to classified polycrystalline silicon granules, classified at least into the two size classes of screen target size and screen undersize, with a separation sharpness between screen target size and screen undersize of more than 0.86.

(74) Preference is given to classified polycrystalline silicon granules, classified into screen target size, screen undersize and screen oversize, with a separation sharpness between screen target size and screen undersize and between screen target size and screen oversize of more than 0.86 in each case.

(75) Classified polycrystalline silicon granules preferably have the following contaminations by metals at the surface: Fe: <800 pptw, more preferably <400 pptw; Cr: <100 pptw, more preferably <60 pptw; Ni: <100 pptw, more preferably <50 pptw; Na: <100 pptw, more preferably <50 pptw; Cu: <20 pptw, more preferably <10 pptw; Zn: <2000 pptw, more preferably <1000 pptw.

(76) Classified polycrystalline silicon granules preferably have contamination by carbon at the surface of less than 10 ppmw, more preferably less than 5 ppmw.

(77) Classified polycrystalline silicon granules preferably have contamination by fine dust at the surface of less than 10 ppmw, more preferably less than 5 ppmw. Fine dust is defined as silicon particles having a size of less than 10 m.

EXAMPLES AND COMPARATIVE EXAMPLES

(78) The advantages of the invention are shown hereinafter by examples and comparative examples.

(79) Example 1 and comparative example 2 relate to the classifying of polycrystalline silicon chunks into chunk sizes 2, 1, 0 and F.

(80) Example 3 and comparative example 4 relate to the classifying of polycrystalline silicon granules (screen target size 0.75-4 mm).

Example 1

(81) Table 1a shows the main parameters of the screening machine.

(82) TABLE-US-00001 TABLE 1a Screen width b [mm] 600 Screen length l [mm] 1600 Frequency n [Hz] 25 Rotational speed [rpm] 1500 Angular velocity [1/s] 157.1 Stroke [mm] 3 Amplitude r [mm] 1.5 Angle of inclination [] 0 Throwing angle [] 50 Screening index Kv [] 2.9 Throughput [kg/h] 700 N.sub.2 sifting gas [m.sup.3 (STP)/h] 50

(83) Table 1b shows which screen set was used in the example. Three screen decks with different mesh sizes of the screens were used.

(84) TABLE-US-00002 TABLE 1b Mesh size [mm] Material Deck 1 9 polyurethane Deck 2 1.9 polyamide Deck 3 0.3 polyamide

(85) Table 1c shows the composition of the screen linings.

(86) TABLE-US-00003 TABLE 1c Element Polyurethane: Polyamide: Al [ppmw] 17 0.7 Ca [ppmw] 14 9.1 Cr [ppmw] <0.2 0.3 Fe [ppmw] 0.7 0.9 K [ppmw] 0.7 <0.2 Mg [ppmw] 0.4 0.2 Na [ppmw] 0.3 0.6 P [ppmw] 63 <20 Sn [ppmw] 5.4 <0.2 Ti [ppmw] 570 0.2 Zn [ppmw] 8.5 <0.2 As, B, Ba, Cd, Co, Cu, <0.2 <0.2 Li, Mn, Mo, Ni, Sr, V [ppmw] Be, Bi, Pb, Sb, W [ppmw] <0.2 <0.2

(87) The screening results achieved with respect to particle size distribution are shown in tables 1d and 1e.

(88) TABLE-US-00004 TABLE 1d Chunk Chunk Chunk Chunk size 2 size 1 size 0 size F 5% by weight 11.3 3.9 0.65 0.12 length quantile: [mm] 95% by weight 26.7 13.9 4.4 0.72 length quantile: [mm]

(89) TABLE-US-00005 TABLE 1e CS 2/1 CS1/0 CS0/F Overlap of 5% by weight/ 2.6 0.5 0.07 95% by weight [mm]

(90) Table 1f shows the contaminations of the classified chunks by surface metals, carbon, dopants and fine dust.

(91) TABLE-US-00006 TABLE 1f Metals, carbon, dopants, Chunk Chunk Chunk Chunk fine dust size 2 size 1 size 0 size F Fe [pptw] 80 170 1200 12,800 Cr [pptw] 10 60 270 7300 Ni [pptw] <10 10 110 5400 Na [pptw] 20 40 430 6300 Zn [pptw] <10 40 210 5000 Al [pptw] 30 80 40 6200 Cu [pptw] <10 <10 30 <5000 Mg [pptw] <10 20 70 5600 Ti [pptw] <10 20 170 <5000 W [pptw] 1500 6340 57,600 969,000 K [pptw] 20 10 160 <5000 Ag [pptw] <10 <10 <10 <5000 Ca [pptw] 60 110 350 <5000 Co [pptw] 270 730 9300 135,000 V [pptw] <10 10 130 <5000 Pb [pptw] <10 <10 90 <5000 Zr [pptw] <10 <10 860 <5000 Mo, As, Be, Bi, Cd, In, <10 <10 <10 <5000 Li, Mn, Sn [pptw] C [ppbw] 72 278 896 5857 B [pptw] 6 15 41 106 P [pptw] 35 131 208 574 As [pptw] 3 7 15 51 Fine dust (<10 m) 1.9 3.8 8.4 17.2 [ppmw]

Comparative Example 2

(92) Table 2a shows the essential parameters of the screening machine used therefor.

(93) TABLE-US-00007 TABLE 2a Screen width b [mm] 600 Screen length l [mm] 1600 Frequency n [Hz] 20 Rotational speed [rpm] 1200 Angular velocity [1/s] 125.7 Stroke [mm] 2.4 Amplitude r [mm] 1.2 Angle of inclination [] 0 Throwing angle [] 45 Screening index Kv [] 1.4 Throughput [kg/h] 700 N.sub.2 sifting gas [m.sup.3 (STP)/h] NN

(94) Table 2b shows which screen set was used in comparative example 2. Three screen decks with different mesh sizes of the screens were used.

(95) TABLE-US-00008 TABLE 2b Mesh size [mm] Material Deck 1 9 polyurethane Deck 2 1.9 polyamide Deck 3 0.3 polyamide

(96) Table 2c shows the composition of the screen linings used.

(97) TABLE-US-00009 TABLE 2c Element Polyurethane: Polyamide: Al [ppmw] 43 2.3 Ca [ppmw] 35 44 Cr [ppmw] <0.2 2.0 Fe [ppmw] 4.5 4.7 K [ppmw] 5.1 0.6 Mg [ppmw] 2.6 0.8 Na [ppmw] 3.8 6.1 P [ppmw] 114 28 Sn [ppmw] 18 1.1 Ti [ppmw] 1220 0.7 Zn [ppmw] 19 1.5 Ni [ppmw] 1.2 0.8 Cu [ppmw] 0.8 0.6 B [ppmw] 4.4 1.9 As, B, Ba, Cd, Co, Li, <0.2 <0.2 Mn, Mo, Sr, V [ppmw] Be, Bi, Pb, Sb, W [ppmw] <0.2 <0.2

(98) The screening results achieved with respect to particle size distribution are shown in Tables 2d and 2e.

(99) TABLE-US-00010 TABLE 2d Chunk Chunk Chunk Chunk size 2 size 1 size 0 size F 5% by weight length quantile 10 3 0.5 0.11 [mm] 95% by weight length quantile 40 15 5 0.81 [mm]

(100) TABLE-US-00011 TABLE 2e CS 2/1 CS1/0 CS0/F Overlap of 5% by weight/ 5 2 0.31 95% by weight [mm]

(101) The overlap is much higher than in example 1. This is attributable to the altered parameters in the screening machine, especially to the lower screening index.

(102) Table 2f shows the contaminations of the classified chunks by surface metals, carbon, dopants and fine dust.

(103) TABLE-US-00012 TABLE 2f Chunk Chunk Chunk Chunk Surface contaminations size 2 size 1 size 0 size F Fe [pptw] 200 340 1640 19,800 Cr [pptw] 30 50 310 11,000 Ni [pptw] <10 40 180 6800 Na [pptw] 40 50 480 7900 Zn [pptw] 20 30 360 6100 Al [pptw] 70 120 160 8400 Cu [pptw] <10 20 60 <5000 Mg [pptw] <10 30 80 9700 Ti [pptw] <10 40 160 <5000 W [pptw] 1640 5830 60,700 1,067,000 K [pptw] 10 30 140 <5000 Ag [pptw] <10 <10 <10 <5000 Ca [pptw] 50 130 380 <5000 Co [pptw] 300 790 11,300 12,800 V [pptw] <10 <10 100 <5000 Pb [pptw] <10 20 80 <5000 Zr [pptw] <10 <10 670 <5000 Mo, As, Be, Bi, Cd, In, <10 <10 <10 <5000 Li, Mn, Sn [pptw] C [ppbw] 103 387 1431 7299 B [pptw] 6 16 48 133 P [pptw] 32 164 216 614 As [pptw] 2 8 22 60 Fine dust [ppmw] 4.8 11.5 19.3 44.2

(104) The contaminations are higher throughout than in example 1. This shows the influence of the composition of the screen linings on the surface contamination of the chunks after classification.

Example 3

(105) Table 3a shows the essential parameters of the screening machine.

(106) TABLE-US-00013 TABLE 3a Screen width b [mm] 500 Screen length l [mm] 1100 Frequency n [Hz] 24.3 Rotational speed [rpm] 1460 Angular velocity [1/s] 152.9 Stroke [mm] 2.4 Amplitude r [mm] 1.2 Angle of inclination [] 3 Throwing angle [] 40 Screening index Kv [] 1.95 Si-throughput [kg/h] 1000 N.sub.2 sifting gas [m.sup.3 (STP)/h] 55

(107) Table 3b shows which screen set was used in example 3. Three screen decks with different mesh sizes of the screens were used.

(108) TABLE-US-00014 TABLE 3b Mesh size [mm] Material Deck 1 9 polyurethane Deck 2 4.0 polyamide Deck 3 0.75 polyamide

(109) Table 3c shows the composition of the screen linings.

(110) TABLE-US-00015 TABLE 3c Element: Polyurethane: Polyamide: Al [ppmw] 17.1 <0.2 Ca [ppmw] 11.3 18.6 Cr [ppmw] <0.2 <0.2 Fe [ppmw] 0.6 0.3 K [ppmw] 0.9 NN Mg [ppmw] 0.3 0.2 Na [ppmw] 0.4 0.9 P [ppmw] 53.2 <20 Sn [ppmw] 5.8 NN Ti [ppmw] 560 <0.2 Zn [ppmw] 7.5 <0.2 B, Ba, Cd, Co, Cu, Li, <0.2 <0.2 Mn, Mo, Ni, Sr, V [ppmw] As, Be, Bi, Pb, Sb, W [ppmw] <0.2 NN

(111) The results achieved with respect to particle size distribution are shown in tables 3d and 3e.

(112) TABLE-US-00016 TABLE 3d Screen Screen target undersize size Screen oversize Waste (<0.75 mm) (0.75-4 mm) (4-9 mm) (>9 mm) 5% by weight 0.35 0.81 3.61 NN quantile [mm] 95% by weight 0.79 2.86 7.68 NN quantile [mm]

(113) TABLE-US-00017 TABLE 3e Screen target Screen size/screen oversize/screen undersize target size Separation sharpness [] 0.862 0.876

(114) Table 3f shows the contaminations of the classified granules by surface metals, carbon, dopants and fine dust.

(115) TABLE-US-00018 TABLE 3f Screen target Screen Screen undersize size oversize Surface metals: (<0.75 mm) (0.75-4 mm) (4-9 mm) Fe [pptw] 1700 860 380 Cr [pptw] 150 100 80 Ni [pptw] 120 80 40 Na [pptw] 390 230 150 Zn [pptw] 2620 2120 1530 Al [pptw] 260 150 140 Cu [pptw] 40 25 15 Mg [pptw] 120 70 60 Ti [pptw] 210 90 90 W [pptw] 60 50 <10 K [pptw] 70 45 40 Ca [pptw] 580 360 320 Mo, As, Sn, Ag, Co, V, <10 <10 <10 Pb, Zr [pptw] C [ppbw] 564 252 204 B [ppta] 27 25 23 P [ppta] 123 120 114 As [ppta] 8 6 6 Fine dust [ppmw] NN 3.6 NN

Comparative Example 4

(116) Table 4a shows the essential parameters of the screening machine.

(117) TABLE-US-00019 TABLE 4a Screen width b [mm] 500 Screen length l [mm] 1100 Frequency n [Hz] 20 Rotational speed [rpm] 1200 Angular velocity [1/s] 125.7 Stroke [mm] 2.6 Amplitude r [mm] 1.3 Angle of inclination [] 3 Throwing angle [] 40 Screening index Kv [] 1.4 Si-throughput [kg/h] 1000 N.sub.2 sifting gas [m.sup.3 (STP)/h] 45

(118) Table 4b shows which screen set was used in comparative example 4. Three screen decks with different mesh sizes of the screens were used.

(119) TABLE-US-00020 TABLE 4a Mesh size [mm] Material Deck 1 9 polyurethane Deck 2 4.0 polyamide Deck 3 0.75 polyamide

(120) Table 4c shows the composition of the screen linings used.

(121) TABLE-US-00021 TABLE 4c Element Polyurethane: Polyamide: Al [ppmw] 57.2 1.3 Ca [ppmw] 45.2 32.5 Cr [ppmw] 1.5 1.3 Fe [ppmw] 14.0 3.1 K [ppmw] 6.5 0.4 Mg [ppmw] 3.6 1.4 Na [ppmw] 9.5 11.1 P [ppmw] 180 25.1 Sn [ppmw] 12.5 0.6 Ti [ppmw] 1400 0.3 Zn [ppmw] 25.3 5.8 Ni [ppmw] 0.7 0.6 Cu [ppmw] 0.5 0.3 B [ppmw] 5.3 0.4 Ba, Cd, Co, Li, Mn, Mo, Sr, <0.2 <0.2 V, s, Be, Bi, Pb, Sb, W [ppmw]

(122) The screening results achieved with respect to particle size distribution are shown in Tables 4d and 4e.

(123) TABLE-US-00022 TABLE 4d Screen Screen target Screen undersize size oversize Waste (<0.75 mm) (0.75-4 mm) (4-9 mm) (>9 mm) 5% by weight 0.38 0.74 3.56 NN quantile: [mm] 95% by weight 0.78 2.63 7.30 NN quantile: [mm]

(124) TABLE-US-00023 TABLE 4e Screen target Screen size/screen oversize/screen undersize target size Separation sharpness [] 0.803 0.874

(125) The separation sharpness in the case of screen target size/screen undersize is worse than in example 3. This is attributable to the lower screening index compared to example 3.

(126) Table 4f shows the contaminations of the classified granules by surface metals, carbon, dopants and fine dust.

(127) TABLE-US-00024 TABLE 4F Screen Screen target undersize size Screen oversize Surface metals: (<0.75 mm) (0.75-4 mm) (4-9 mm) Fe [pptw] 3500 1490 720 Cr [pptw] 270 210 140 Ni [pptw] 300 150 80 Na [pptw] 750 530 520 Zn [pptw] 3270 2610 2230 Al [pptw] 360 220 170 Cu [pptw] 70 60 30 Mg [pptw] 610 320 130 Ti [pptw] 340 120 130 W [pptw] 50 50 <10 K [pptw] 210 170 110 Ca [pptw] 2520 810 720 Sn 40 30 <10 Mo, As, Ag, Co, <10 <10 <10 V, Pb, Zr [pptw] C [ppbw] 728 311 292 P [ppta] 202 148 133 As [ppta] 15 11 8 Fine dust [ppmw] NN 8.3 NN

(128) The contaminations are higher throughout than in example 3.

(129) The measurement methods which follow were used to determine the parameters specified.

(130) Contamination by carbon is determined by means of an automatic analyzer. This is described in detail in U.S. application Ser. No. 13/772,756, which is yet to be published, and in German application number 102012202640.1.

(131) The dopant concentrations (boron, phosphorus, As) are determined to ASTM F1389-00 on monocrystalline samples.

(132) The metal contaminations are determined to ASTM 1724-01 by ICP-MS.

(133) The fine dust measurement is effected as described in DE 10 2010 039 754 A1.

(134) The particle sizes (minimum chord) are determined by means of dynamic image analysis according to ISO 13322-2 (measurement range: 30 m-30 mm, type of analysis: dry measurement of powders and granules).