Process for determining surface contamination of polycrystalline silicon

Abstract

The invention provides a process for determining surface contamination of polycrystalline silicon, including the steps of: a) providing two polycrystalline silicon rods by deposition in a Siemens reactor; b) determining contaminants in the first of the two rods immediately after the deposition; c) conducting the second rod through one or more systems in which polycrystalline silicon rods are processed further to give rod pieces or polysilicon fragments, optionally cleaned, stored or packed; d) then determining contaminants in the second rod; wherein the difference in the contaminants determined in the first and second rods gives surface contamination of polycrystalline silicon resulting from systems and the system environment.

Claims

1. A process for determining surface contamination of polycrystalline silicon, comprising steps of a) providing two polycrystalline silicon rods by deposition in a Siemens reactor; b) determining a first concentration of non-carbon contaminants and a first concentration of carbon contaminants in a first rod of the two polycrystalline silicon rods after the deposition, wherein: i) a first rod wafer is removed from the first rod, ii) the first rod wafer is analyzed by FTIR to determine the first concentration of carbon contaminants, and iii) the first rod after removal of the first rod wafer is converted by a float zone process to a first monocrystalline rod, and the first concentration of non-carbon contaminants is determined by photoluminescence on a first monocrystalline rod wafer removed from the first monocrystalline rod; c) conducting a second rod of the two polycrystalline silicon rods through at least one system for further processing polycrystalline silicon rods, wherein the further processing comprises comminution to provide rod pieces or polysilicon fragments, wherein the second rod is not comminuted and the first rod is not conducted through the at least one system; d) determining contaminants in the second rod by processing the second rod by a float zone process to provide a second monocrystalline rod; e) removing a FTIR wafer and a photoluminescence wafer from the second monocrystalline rod; f) performing a FTIR analysis of the FTIR wafer to determine a second concentration of carbon contaminants; g) performing a photoluminescence analysis of the photoluminescence wafer to determine a second concentration of non-carbon contaminants; and h) determining the surface contamination of polycrystalline silicon resulting from the at least one system and a system environment from a difference between the first concentration of the non-carbon contaminants and the second concentration of the non-carbon contaminants.

2. The process as claimed in claim 1, wherein the non-carbon contaminants of the first rod and the second rod are members selected from the group consisting of boron, phosphorus, aluminum and arsenic.

3. The process as claimed in claim 2, wherein the first rod is packed in a polyethylene bag after the deposition.

4. The process as claimed in claim 3, wherein in step c) the second rod is conducted through a comminution system and a packaging system.

5. The process as claimed in claim 3, wherein the further processing conducted by the at least one system in step c) comprises comminution, cleaning, storage and packaging of polysilicon, wherein the second rod is packed in a polyethylene bag but is not comminuted, cleaned or stored.

6. The process as claimed in claim 1, wherein the first rod is packed in a polyethylene bag after the deposition.

7. The process as claimed in claim 1, wherein the further processing conducted by the at least one system in step c) comprises comminution, cleaning, storage and packaging of polysilicon, wherein the second rod is packed in a polyethylene bag but is not comminuted, cleaned or stored.

8. The process as claimed in claim 1, wherein the second rod is cleaned, stored and packaged prior to step (d).

9. The process as claimed in claim 1, wherein the second rod is cleaned and packaged prior to step (d) but not stored.

10. The process as claimed in claim 1, wherein the second rod is cleaned and stored prior to step (d) but not packaged.

11. The process as claimed in claim 1, wherein the second rod is stored and packaged prior to step (d) but not cleaned.

12. The process as claimed in claim 1, wherein the second rod is packaged prior to step (d) but not cleaned or stored.

13. The process as claimed in claim 1, wherein the second rod is cleaned prior to step (d) but not stored or packaged.

14. The process as claimed in claim 1, wherein the second rod is stored prior to step (d) but not cleaned or packaged.

15. The process as claimed in claim 1, wherein the surface contamination in step (h) is determined from the difference between the first concentration of the non-carbon contaminants and the second concentration of the non-carbon contaminants and a difference between the first concentration of carbon contaminants and the second concentration of carbon contaminants.

Description

DESCRIPTION OF THE INVENTION

(1) This object is achieved by a process for determining surface contamination of polycrystalline silicon, comprising the steps of

(2) a) providing two polycrystalline silicon rods by deposition in a Siemens reactor;

(3) b) determining contaminants in the first of the two rods immediately after the deposition;

(4) c) conducting the second rod through one or more systems in which polycrystalline silicon rods are processed further to give rod pieces or polysilicon fragments, optionally cleaned, stored or packed;

(5) d) then determining contaminants in the second rod;

(6) wherein the difference in the contaminants determined in the first and second rods gives surface contamination of polycrystalline silicon resulting from systems and the system environment.

(7) First of all, two polycrystalline silicon rods are provided, by depositing polycrystalline silicon in a Siemens reactor, giving rise to U-shaped polycrystalline silicon bodies each comprising two polycrystalline silicon rods.

(8) The reaction gas used in the deposition typically comprises a silicon-containing component, preferably trichlorosilane, and hydrogen.

(9) The deposition is preferably effected in a small test reactor.

(10) In practice, an aforementioned U-shaped body, after the deposition, can be deinstalled from the reactor, and then the bridge and respective rod ends are removed, so as to obtain two polycrystalline silicon rods from one and the same batch.

(11) The two polycrystalline silicon rods provided in step a) were preferably joined to one another (brother rods) via a bridge (U shape) during the deposition.

(12) In the case of use of a small reactor, the two poly-crystalline silicon rods may typically have a length of about 20 cm and a diameter of about 1.6 cm.

(13) One of the two rods is preferably packed in a PE bag immediately after the deposition and the separation of bridge and rod end. The two rods are preferably each packed in a PE bag.

(14) This first rod is subsequently analyzed for contamination.

(15) Preference is given to determining dopants and carbon.

(16) In an analytical laboratory to which the packed rod is transported, preference is given to removing the rod from the PE bag, separating a wafer from the poly-crystalline rod and sending it to the FTIR analysis.

(17) This determines the carbon concentration.

(18) The remaining rod is preferably converted by means of FZ to a monocrystalline rod.

(19) The concentration of dopants is determined therein by means of photoluminescence.

(20) The values thus determined for dopant and carbon concentrations serve as reference values for the second rod.

(21) The second rod, after removal from the PE bag, is preferably conducted through the systems for the production of polycrystalline silicon chunks (comminution, packaging) and optionally through the systems for cleaning of polycrystalline silicon chunks).

(22) In the course of this, the rod takes up the contaminants in terms of dopants and carbon while passing through the systems.

(23) After passing through the cleaning systems or the production line for uncleaned chunk poly, the contaminated rod is preferably packed again in a high-purity PE bag.

(24) Preferably two labels are stuck onto the PE bag:

(25) Label No. 1: Label with the batch number of the original batch (comparison with first rod)

(26) Label No. 2: Label with a new batch number

(27) The contaminated rod is used to produce a monocrystalline rod by means of FZ.

(28) Subsequentlyas described above for the first roddopants are determined by means of photoluminescence, and carbon by means of FTIR.

(29) In contrast to the first rod, the determination of the carbon concentration by means of FTIR is effected not on a polycrystalline wafer but on a monocrystalline wafer.

(30) In the course of FZ pulling of the contaminated rod, the carbon-containing particles migrate from the surface into the bulk and thus become amenable to carbon measurement by means of FTIR.

(31) The values measured for the first rod are subtracted from the values for the second rod conducted through the systems.

(32) The differences between the first and second rods then give rise to the value which can be attributed to the surface of the polycrystalline silicon after processing, cleaning, packaging.

(33) The process according to the invention thus makes it possible to determine indirectly how polysilicon is contaminated at the surface in the course of the processing steps such as comminution, cleaning, packaging, or in transport operations.

(34) The process thus gives surface contaminations for all possible products, such as polysilicon rods, cut rods and polysilicon chunks of different size classes (etched or unetched).

(35) The process also enables monitoring and optimization of individual production steps with regard to surface contamination:

(36) For example, the second rod can be conducted only through the cleaning system or only through the comminution system. In that case, the process separately gives the influence of the comminution system and the environment thereof, or of the cleaning system and the environment thereof, on surface contamination. The same applies to the packing of poly-silicon or the transport of the polysilicon from one system to another system.

(37) The determination of the surface contamination is reproducible.

(38) For testing, twelve rods in twelve process dishes were run through the cleaning system at the same time.

(39) Subsequently, the dopant concentrations were determined by means of photoluminescence.

(40) Theoretically, the twelve brother rods, even though they originate from different batches, should have the same analysis values for boron, phosphorus, aluminum and arsenic, since they have been run through the cleaning system simultaneously under the same conditions.

(41) Table 1 shows the values determined for boron, phosphorus, aluminum and arsenic in ppta.

(42) The values measured for the first rod were subtracted from the values for the second rod conducted through the systems.

(43) TABLE-US-00001 TABLE 1 B P Al As #1 20.95 23.37 0.98 4.95 #2 12.74 <1 <0.5 <0.5 #3 14.40 1.25 <0.5 2.04 #4 16.04 5.49 <0.5 0.52 #5 20.96 10.09 <0.5 <0.5 #6 17.79 7.22 0.52 <0.5 #7 12.28 <1 <0.5 <0.5 #8 14.03 <1 <0.5 <0.5 #9 22.15 13.85 <0.5 1.60 #10 21.98 7.51 <0.5 2.03 #11 12.49 1.52 <0.5 1.71 #12 22.91 14.86 0.50 <0.5

(44) The following reproducibilities and detection limits were estimated:

(45) Boron

(46) Reproducibility: +/5 ppta

(47) Detection limit: 5 ppta

(48) Phosphorus

(49) Reproducibility: +/5 ppta

(50) Detection limit: 5 ppta

(51) Aluminum

(52) Reproducibility: +/0.25 ppta

(53) Detection limit: 1 ppta

(54) Arsenic

(55) Reproducibility: +/0.5 ppta

(56) Detection limit: 5 ppta

(57) The process according to the invention can also be used to determine the content of carbon particles on the silicon surface with a reproducibility of +/10 ppba at a detection limit of 10 ppba.

(58) Example

(59) The example shows how the second rod is conducted through the cleaning system and then analyzed for dopant concentration. The first rod (brother rod of the second rod from a U-shaped body after deposition) was analyzed as described above for dopants by means of photoluminescence.

(60) The PE bag in which the second rod (length 20 cm, diameter 1.6 cm) has been packed is opened with scissors, preferably ceramic scissors. The rod is removed, using an ultraclean glove for manual removal. Subsequently, the rod is placed into a process dish.

(61) A suitable ultraclean glove (PE-Tyvek glove) is disclosed in US 2011-0083249, which is fully incorporated here by reference. Tyvek from DuPont is a paper web-like fibrous functional textile composed of thermally welded fibers of high-density polyethylene (HDPE).

(62) The process dish filled with the rod is run through the cleaning system.

(63) In the course of this, the silicon rod, in a pre-purification, is washed with an oxidizing cleaning solution containing the compounds hydrofluoric acid (HF), hydrochloric acid (HCl) and hydrogen peroxide (H.sub.2O.sub.2). In a main cleaning operation, the rod is washed with a cleaning solution comprising nitric acid (HNO.sub.3) and hydrofluoric acid (HF). Subsequently, the rod is washed with an oxidizing cleaning solution and thus hydrophilized. With regard to the cleaning process, EP 0 905 796 B1 is fully incorporated by reference.

(64) After the rod has been cleaned, it is dried and, after cooling, grasped with an ultraclean glove, preferably a PE-Tyvek glove, and packed in a high-purity PE bag which is sealed.

(65) Two labels are stuck onto the PE bag:

(66) Label No. 1: Label with the batch number of the original batch (enables comparison of the measurements with first rod)

(67) Label No. 2: Label with a new batch number

(68) The contaminated rod is processed by means of FZ to give a monocrystalline rod. As described above, dopants are determined by means of photoluminescence. It would likewise also be possible to analyze carbon by means of FTIR.

(69) The values for the boron, phosphorus, aluminum and arsenic dopants measured for the first rod are subtracted from the corresponding values for the second rod.

(70) The differences between the first and second rods then give the values which can be attributed to the surface of the polysilicon.

(71) Table 2 shows the differences determined for surface contaminations of boron, phosphorus, aluminum and arsenic.

(72) TABLE-US-00002 TABLE 2 B P Al As 44.06 15.46 0.03 1.29 119.32 405.97 194.63 22.78 19.10 4.28 0.89 4.66 128.55 250.91 145.57 17.18 7.70 79.68 0.87 0.52 3.58 21.01 1.53 2.66 3.86 16.17 6.71 4.39 6.57 0.22 0.25 1.24 9.11 2.68 1.37 1.08 10.10 1.37 14.60 0.59 20.47 41.02 7.26 3.18