FIELD OF THE INVENTION

The present invention relates to a method for verifying the source of a silicon containing material. In particular, the invention relates to determining the ratio of silicon isotopes in a sample of the material and then comparing the ratio against data for silicon of known source to verify the source of the material.

BACKGROUND

Verification of the source of silicon and silicon containing materials is becoming increasingly important worldwide.

Silicon is a strategic resource for many nations due to its importance as a semiconductor for computers and photovoltaics. There is growing political demand to verify the source of silicon. There are also growing concerns around a lack of sustainability, damage to the environment caused by silicon manufacturers, and damage to human health by some manufacturers of silicon and silicon containing materials.

Consumer awareness of the origin of silicon and silicon containing materials is growing in many markets. There is increasing demand from consumers that products including silicon containing materials meet their ethical expectations. Governmental regulation is also tightening around manufacturers that use silicon containing materials to verify that the materials have been produced legally or obtained from approved sources.

Silicon and silicon containing materials are often highly refined, for example in the case of elemental silicon for the photovoltaic and semiconductor industries, or in the case of fumed silica. Governments and consumers are demanding that the source or location where the silicon containing material is refined can be verified. It is also useful to be able to verify the geographical or geological region where the silicon containing material was extracted.

For some materials, forensic chemistry and statistics can be combined to establish an inherent "fingerprint" based on naturally occurring chemical elements and their isotopes contained in the material. For example, the geographic region where a particular mineral was extracted can sometimes be determined by analysis of trace elements or isotopic ratios. In other cases, the production process can impart a fingerprint that is unique to that batch or production unit.

There are four isotopes of silicon in the natural environment: ²⁸ Si, ²⁹ Si, ³⁰ Si and ³² Si. The isotopes ²⁸ Si, ²⁹ Si, ³⁰ Si are stable isotopes and make up virtually the entirety of natural silicon in any terrestrial sample. ³² Si is radiogenic with a half-life of about 153 years. Only traces of ³² Si are detectible in natural silicon samples. Other isotopes of silicon have a half-life too short and abundances too low to detect in samples of silicon. There is little variation in isotopic ratios of silicon across geographic areas or in geological samples of silicon and silicon containing materials. This is believed to be due to silicon having one valence state (Si ⁴⁺ ) and not forming volatile compounds readily.

Elemental silicon is produced to a number of grades, depending on the intended use. Metallurgical grade silicon (MGS) has a purity of between about 95% and 99%. MGS may be refined to produce polycrystalline silicon (polysilicon) or monocrystalline silicon (monosilicon), in which impurities are less than about 1 part per million (ppm). Polysilicon is made up of many silicon crystallites, and therefore comprises grain boundaries at adjacent crystallites. Monosilicon is a single crystal of silicon, having a continuous crystal structure, and is substantially devoid of grain boundaries. For the semiconductor industry, polysilicon or monosilicon generally have impurities of less than about 1 part per billion (ppb).

Because of the high purity requirements, polysilicon and monosilicon are produced by chemical vapour deposition methods. Silicon is converted to a gaseous phase silicon compound, such as silane (SiF ) or trichlorosilane (SiHCh), and purified via distillation. Purified gas is then pyrolised to elemental silicon and collected. The purified silicon may be polysilicon or monosilicon, depending on the deposition methodology. In some cases, the purified silicon may be recrystallised to increase its purity, reduce the number of grain boundaries, or to produce monosilicon.

Use of traceability frameworks or knowledge of source in the silicon sector has been a relatively recent requirement, largely stemming from concerns around the working conditions in some production facilities and attempts to account for carbon use in the supply chain. The opacity and lack of reliability of traceability information when relying on vendor self-declarations creates a need for manufacturers and suppliers to know the source of silicon containing materials in their supply chain and to be able to prove the source through scientifically reliable traceability techniques.

The applicant has now found that the use of silicon isotopes ratios can be used to accurately verify the origin of the silicon. It is therefore an object of the invention to provide a method of verifying the source or origin of a silicon containing material, or to at least provide an alternative to existing methods.

SUMMARY OF THE INVENTION

In one aspect of the invention there is provided a method of verifying the source of a silicon containing material, comprising:

(a) determining a ratio of silicon isotopes of the silicon present in a sample of the material; and

(b) comparing the ratio against silicon isotope ratio data for silicon of known source to verify the source of the material. In certain embodiments of the invention, the ratio of silicon isotopes is any ratio of ²⁸ Si, ²⁹ Si and ³⁰ Si. Preferably, the ratio is the ratio ²⁹ Si/ ²⁸ Si or the ratio ³⁰ Si/ ²⁸ Si, or a combination thereof.

In certain embodiments of the invention, the silicon is elemental silicon. In other embodiments, the silicon is in the form of a silicon dioxide, a silicate, or a silicic acid or a salt thereof.

In certain embodiments of the invention, the silicon containing material is at least 99.99999% silicon, for example at least 99.999999999% silicon.

In certain embodiments of the invention, the silicon containing material is polysilicon. In other embodiments, the silicon containing material is monosilicon.

In certain embodiments of the invention, the ratio of silicon isotopes is determined by any known method, including isotope ratio mass spectrometry (IRMS), thermal ionisation mass spectrometry (TIMS), or inductively coupled plasma mass spectroscopy (ICP-MS). The preferred method is multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS).

In certain embodiments of the invention, the source is a silicon refining facility, a polysilicon refining facility, or a monosilicon refining facility.

In certain embodiments of the invention, the silicon containing material is a silicon semiconductor, a silicon photovoltaic cell, a silicon chip, or a silicon integrated circuit.

In certain embodiments of the invention, the method further comprises determining a ratio of oxygen isotopes of oxygen present in the sample. Preferably, the ratio of oxygen isotopes is the ratio of ¹⁸ O/ ¹⁶ O.

In certain embodiments of the invention, the method further comprises determining the concentrations of one or more trace elements in the sample. Preferably, the trace elements include, but are not limited to, at least one of Li, B, Na, Mg, Al, K, Ca, Ti, Ni, As, Rb, Sr, Sn, Sb, Cs and Ba.

In certain embodiments of the invention, the sample comprises trace elements at a concentration of less than 100 ppm, preferably less than 10 ppb, more preferably less than 1 ppb.

In certain embodiments of the invention, the concentrations of one or more trace elements are obtained via analysis of a subsample of the sample of the silicon, wherein the subsample has a location in the sample that includes a grain boundary.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 shows isotopic fingerprints of samples from different sources. The squares represent samples obtained from USA Plant 1. The open squares represent MGS samples and the filled squares represent polysilicon samples. The circles represent samples obtained from USA Plant 2. The open circles represent polysilicon samples and the filled circles represent monocrystal silicon wafer samples.

Figure 2 shows the Si isotopic values of samples from the same sources but at different stages of production. The solid black symbols represent samples of 100% virgin silicon doped with phosphorus. The open circles represent samples that are also 100% polysilicon doped with gallium.

DEFINITIONS

The term "isotope ratio" means the ratio of atomic abundances of two isotopes of the same chemical element.

The term "trace element" means a chemical element having a very low concentration and includes, but is not limited to, an element having a concentration of less than 100 ppm or less than 100 pg/g. The term "trace metal" has a correspnding meaning.

The term "polycrystalline silicon" or "polysilicon" means elemental silicon (Si) in the form of a solid made up of silicon crystallites and comprising grain boundaries at adjacent crystallites.

The term "monocrystalline silicon" or "monosilicon" means elemental silicon (Si) in the form of a single crystal of silicon and being substantially devoid of grain boundaries.

The term "metallurgical grade silicon" or "MGS" is silicon having a purity of between about 95% and 99%, and is usually prepared by purifying silicon using heat and a reducing agent in an arc furnace.

The term "wafer" means a thin slice of semiconductor material such as crystalline silicon.

Unless specifically defined otherwise, all technical and scientific terms used herein shall be taken to have the same meaning as commonly understood by one of ordinary skill in the art.

It is intended that reference to a range of numbers disclosed herein (e.g. 1 to 10) also incorporates reference to all related numbers within that range (e.g. 1, 1.1, 2, 3, 3.9, 4, 5, 6, 6.5, 7, 8, 9 and 10) and also any range of rational numbers within that range (for example 2 to 8, 1.5 to 5.5 and 3.1 to 4.7) and, therefore, all sub-ranges of all ranges expressly disclosed herein are expressly disclosed. These are only examples of what is specifically intended and all possible combinations of numerical values between the lowest value and the highest value enumerated are to be considered to be expressly stated in this application in a similar manner.

The term "and/or", e.g., "X and/or Y", shall be understood to mean either "X and Y" or "X or Y" and shall be taken to provide explicit support for both meanings or for either meaning. The term "a" or "an" may refer to one or more than one of the entity specified. As such, the terms "a" or "an", "one or more" and "at least one" can be used interchangeably.

Throughout this specification the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps.

Throughout this specification, unless specifically stated otherwise or the context requires otherwise, reference to a single step, composition of matter, group of steps or group of compositions of matter shall be taken to encompass one and a plurality (i.e. one or more) of those steps, compositions of matter, groups of steps or group of compositions of matter.

DETAILED DESCRIPTION

The invention is predicated, at least in part, on the finding that the sourcing and manufacturing process of silicon containing materials differs between providers with respect to ²⁸ Si : ²⁹ Si : ³⁰ Si isotope ratios. Therefore, it is possible to verify the source of a silicon containing material by comparison of its silicon isotope ratio against corresponding data for silicon containing materials of known source.

It can also be advantageous to combine silicon isotope ratio analyses with analyses for trace element concentrations. There is often an abundance of trace elements in silicon containing materials which can indicate the source of the material.

Data relating to the silicon isotope ratio and/or trace element abundance in a sample of a silicon containing material can add relevant identifying information to a dataset (or 'silicon fingerprint') that may be used to determine the source of the silicon containing material. The dataset obtained from a sample of silicon containing material of unknown source may be compared against data for silicon containing materials of known source to verify the source of the material.

There is therefore provided a method for verifying the source of a silicon containing material, comprising:

(a) determining a ratio of silicon isotopes of the silicon present in a sample of the material;

(b) optionally determining the concentrations of one or more trace elements present in a sample of the material; and

(c) comparing the ratio, and optionally the trace element concentrations, against silicon isotope ratio data, and optionally trace element data, for silicon of known source to verify the source of the material. Silicon containing material

Examples of silicon containing materials include elemental silicon, silicon dioxides, silicates, and silicic acids and their salts.

Elemental silicon includes monosilicon, polysilicon, as well as elemental silicon of lower purities such as metallurgical grade silicon (MGS) or raw extracted silicon ore. In some embodiments of the invention, the elemental silicon comprises trace elements at a concentration of less than 100 ppm. In some embodiments, the elemental silicon comprises trace elements at a concentration of less than 10 ppb. In some embodiments, the elemental silicon comprises trace elements at a concentration of less than 1 ppb.

Silicon dioxides can include natural and synthetic silicon dioxides, such as quartz, fused quartz, silica and fumed (or pyrogenic) silica.

Silicates can include natural and synthetic ionic forms of silicon, such as orthosilicates, metasilicates and pyrosilicates.

Silicon containing materials may also include silicic acids and their corresponding salts, such as hexafluorosilicic acid and hexafluorosilicates.

The silicon containing material includes manufactured, processed or semiprocessed silicon containing materials, such as silicon semiconductors, silicon photovoltaic cells, silicon chips, and silicon integrated circuits.

Sources

The source of silicon containing material to be verified by the method of the invention is preferably the facility where the silicon containing material was refined or purified.

The source of silicon containing material may also include the geographic or geological area in which the silicon containing material was extracted.

Isotope ratio

The silicon isotope ratio that is determined in the method of the invention is any ratio of ²⁸ Si, ²⁹ Si and ³⁰ Si. Preferably, the ²⁹ Si/ ²⁸ Si and ³⁰ Si/ ²⁸ Si ratios are determined and the isotope data are reported as delta values which express the deviation from the accepted standard NBS28.

Isotopic ratios are conventionally determined via mass spectrometry (MS). However, in order to use isotopic ratios to verify the source of a silicon containing material, the determination of silicon isotope ratio must be sufficiently precise and sensitive to identify small differences.

Geological quartz (the starting product of polysilicon) is virtually homogenous globally. Measurable differences in Si isotopes have been observed in biogenic silica. Si isotopes undergo fractionation during the purification processes between quartz, MGS and polysilicon. The inventors have now found that the key purification step, where the Si is purified in a gaseous intermediary (trichlorosilane) via distillation, enables isotopic fractionation of the Si isotopes ( ²⁸ Si, ²⁹ Si, ³⁰ Si). As this step is specific to each manufacturer or manufacturing facility, the fractionation may show some specificity between different facilities (origins), observed as a difference in these isotope ratios ( ²⁹ Si/ ²⁸ Si and ³⁰ Si/ ²⁸ Si). Note that due to the low natural abundance of ²⁹ Si and ³⁰ Si, the isotope ratio of ²⁹ Si/ ³⁰ Si cannot be reliably determined.

MS techniques previously considered sufficiently sensitive for determining isotopic ratios required a sample of the material to be in gas phase, which required fluorination of the silicon sample and thus had the attendant hazards and disadvantages of working with fluorine.

The method of the invention involves techniques for determining silicon isotope ratios that include isotope ratio mass spectrometry (IRMS), thermal ionisation mass spectrometry (TIMS), and inductively coupled plasma mass spectroscopy (ICP-MS). ICP- MS is a preferred technique. Examples of variations of ICP-MS that would be suitable for determining silicon isotope ratios include multi-collector ICP-MS (MC-ICP-MS).

Where the silicon containing material is a silicon-oxygen containing material such as silicon dioxide or a silicate, the isotopic ratio of oxygen in the silicon containing material may be determined in addition to determining silicon isotope ratios. The oxygen isotope ratio that is preferably determined in these embodiments of the invention is the ratio of ¹⁸ O/ ¹⁶ O.

Data relating to the oxygen isotope ratio in a silicon and oxygen containing sample can provide datapoints for comparison with data for oxygen isotope ratios of known silicon and oxygen containing materials to assist in verifying the source of the sample. An example of a suitable method for determining the isotopic ratio of the oxygen atoms in the silicon containing material is gas source mass spectrometry.

The determined variation in isotope ratio can be modelled by either Rayleigh kinetics for a closed system, or steady-state model for an open system.

Trace elements

There are many known techniques for analysing trace elements. For example, mass spectrometry, atomic absorption and X-ray fluorescence.

In order to determine the concentration of one or more trace elements in a high purity silicon containing material, standard methods of elemental analysis may not be sufficient. Where trace element concentration is very low (e.g. in the ppb or sub-ppb range), it may be necessary to use particular methods and instruments that can perform the analysis with sufficient precision. The method of the invention involves suitable techniques for determining trace element concentration that include X-ray fluorescence, glow discharge mass spectrometry (GDMS), thermal ionisation mass spectrometry, sensitive high-resolution ion microprobe (SHRIMP), and ICP-MS. Examples of variations of ICP-MS that would be suitable for determining the concentration of trace elements include multi-collector ICP-MS (MC-ICP- MS), time of flight (ICP-ToF), or laser ablation ICP-MS (LA-ICP-MS).

Determination of trace elements of a silicon containing material is preferably performed using LA-ICP-MS. The LA-ICP-MS technique involves the irradiation of a sample with a laser (e.g. 193 nm excimer laser) causing melting and vaporisation of an irradiated portion of the sample and formation of a plasma plume containing ions of the sample. The ions are ejected from the plume into a carrier gas stream connected to an ICP-MS instrument. Elements such as Li, B, Na, Mg, Al, K, Ca, Ti, Ni, As, Rb, Sr, Sn, Sb, Cs and Ba may be measured via the LA-ICP-MS technique.

Where the silicon containing material is elemental silicon such as polysilicon, it has been found that trace elements are more concentrated at the crystallite grain boundaries. Hence, the step of determining the concentration of trace elements in a polysilicon sample may include the step of determining the concentration of trace elements at one or more grain boundaries.

Verification of the source of a silicon containing sample using an analysis of silicon isotopic ratios as a sole method may not be easily achievable due to the very small differences in isotopic ratios in samples derived from different sources. Accordingly, in some cases it may be necessary for the method of the invention to include both steps, i.e. determining trace element abundance and determining silicon isotopic ratios, in order to produce a more detailed dataset.

Known samples

The silicon isotope ratio data is compared against data for silicon containing materials of known source to verify the source of the material.

A reference dataset is obtained or prepared for samples of silicon containing materials of different known sources. The sources of samples of silicon containing materials may include a refinery, a foundry or a manufacturing facility for the silicon containing material.

For each sample of silicon containing material of known source, the ²⁸ Si : ²⁹ Si : ³⁰ Si silicon isotope ratio datapoint is obtained. Optionally, datapoints for trace element concentrations for at least one of Li, B, Na, Mg, Al, K, Ca, Ti, Ni, As, Rb, Sr, Sn, Sb, Cs and Ba are also obtained.

Where the silicon containing material comprises oxygen, such as silicon dioxide or silicate, an additional datapoint may be obtained for the ¹⁸ O/ ¹⁶ O isotope ratio. The datapoints for each sample of silicon containing material of known source are compiled into a computer-accessible library suitable for multivariate statistical analysis.

Comparison and analysis

The variations in silicon isotope ratios between silicon containing materials from different sources may be very small. Similarly, the variations in trace element concentrations in silicon containing materials from different sources may be very small.

Therefore, there may be a need for mathematical analysis of the silicon isotope data and trace element concentrations to be able to use the data to verify the source of a silicon containing material.

In the comparison of the isotope ratio and/or the trace element concentrations in the sample against data for silicon of known source, multivariate statistical models are used to generate a probability value and error rate relating to each source in the reference dataset.

Where the probability value and error rate meet a predetermined threshold for one source, the silicon of unverified source is verified as being derived from that source.

As part of the analysis, data relating to the ratio of silicon isotopes in a sample of the silicon containing material can be analysed according to kinetic models for isotope fractionation, including Rayleigh kinetic models and steady state kinetic models.

Discussion of examples

Example 1 describes a method for determining the silicon isotope ratios in samples of MGS, polysilicon and silicon wafer. The results are shown in Figures 1 and 2.

Figure 1 shows a clear clustering of samples according to their source and differentiation of samples from different sources.

Figure 2 shows show that the fingerprint appears to be conserved through the manufacturing process. MGS samples from USA Plant 1 have a similar fingerprint to Polysilicon samples from USA Plant 1.

The key findings from these data are that the two different sources have quite distinct Si isotope values and those values are preserved through the manufacturing process. Within the MGS samples, one sample was five years older than the other four, suggesting that the fingerprint is conserved over time. This would allow a baseline formed with samples collected today to still be useable and valid for audit scenarios at least five years from now.

Polysilicon samples from USA Plant 2 have a similar fingerprint to monocrystalline wafer samples made with polysilicon from USA Plant 2. Recycled monocrystalline samples doped with varying amounts of either phosphorus or gallium showed that the isotopic fingerprint is again retained during this process (Figure 2). Additionally, the results are also evidence that samples sourced from USA Plant 1 have a distinctly different chemical fingerprint to samples sourced from USA Plant 2 (Figure 1). The theory that Si isotopes undergo fractionation during the purification step seems to hold true. Furthermore, the fractionation appears to be specific for the different facilities.

EXAMPLES

Example 1: Determination of silicon isotope ratios

Methodology

Silicon isotopes were determined using a modification of the method described in Wille et al. (2010), Earth and Planetary Science Letters, 292, 281-289. Each solid polysilicon sample was prepared by encasing the sample within 2 snap-lock bags. Using a hammer (or equivalent), each sample was crushed to form a powder to a consistency resembling castor sugar. Powdered samples (about 0.2 g) were accurately weighed and placed in a nickel crucible and powdered NaOH (1 g) added. Each mixture of sample and NaOH was then melted by heating in a muffle furnace according to the following program: 300 °C ramp o Hold 2 hours

Increase to 500 °C over 1 hour o 675 °C ramp

Hold for 30 minutes o Cool down

Once cooled, each melt was transferred to a 50 ml DigiTube (or equivalent) and diluted to 50 mL with deionised water. The concentration of Si in each melt was determined using the method of Strickland & Parsons (1972), A Practical Handbook of Seawater Analysis, 2nd edition. Ottawa, Canada, Fisheries Research Board of Canada, 310pp. (Bulletin Fisheries Research Board of Canada, Nr. 167 (2nd ed)). DOI: http://dx.doi.org/10.25607/OBP-1791.

Each solution was purified using ion exchange columns. A column was filled with 1 ml Dowex 50 W-X8 cation exchange resin (200-400 mesh). The resin was prepared according to the following procedure: o Add 0.5 ml 8% v/v HF, leave for 1 hour and then add 0.75 ml MQ water (repeat twice more).

The clean resin was then protonated by completing the following steps twice: o Add 0.75 ml 4M HCI (do this 3 times) and then add 0.75 ml MQ (do this 3 times).

Each sample was diluted to 20 ppm Si using the results from the Si concentration analysis. Each sample (0.5 ml) was pipetted into the cleaned column and the eluent collected. The column was rinsed 4 times with MQ water (0.5 ml) collecting the eluent in the same vessel to give 4 ppm samples. o Concentrated QD HNO3 (50 pl) was added to each sample followed by dilution to 0.5 ppm (0.25 ml sample + 1.75 ml 2% HNO3).

Si isotope measurement of each sample was carried out by multi-collector inductively coupled plasma mass spectrometry. ²⁸ Si was set to the centre cup and the ²⁹ Si peak set at 28.966 to avoid NO interference.

Results

A total of 48 samples of MGS, polysilicon and wafer were obtained from two manufacturing plants, USA Plant 1 and USA Plant 2. These samples were measured for silicon isotope ratio values. The results are presented in Tables 1-3 and Figures 1 and 2 where the distribution of the samples is shown.

Stable isotope data are reported as delta values which express the deviation from an accepted standard. The standard used is NBS28 as which serves as the primary reference material for silicon 3 ^2S Si~ 3 ^3C Si measurements defining the zero point of the 3- scaie. The delta values are defined as: where SAN refers to the sample values. Results are multiplied by 1000 and reported in units of per mille (%o).

Table 1: Silicon isotope values for monocrystalline wafer samples

Table 2: Silicon isotope values for polysilicon samples

Table 3: Silicon isotope values for metallurgical grade silicon (MGS) samples

Example 2: Verification of source

The source of a sample of polysilicon of unverified source may be verified according to the following methodology. A reference dataset in computer-accessible format is prepared for several samples of polysilicon of different known sources. For each sample of polysilicon of known source, the ²⁸ Si/ ²⁹ Si/ ³⁰ Si silicon isotope ratio is obtained. Trace element concentrations for at least one of Li, B, Na, Mg, Al, K, Ca, Ti, Ni, As, Rb, Sr, Sn, Sb, Cs and Ba may also be obtained. A sample of polysilicon of unverified source is analysed to obtain its ²⁸ Si/ ²⁹ Si/ ³⁰ Si silicon isotope ratio. Trace element concentrations for at least one of Li, B, Na, Mg, Al, K, Ca, Ti, Ni, As, Rb, Sr, Sn, Sb, Cs and Ba may also be obtained.

The data for the polysilicon of unverified source and the reference dataset is analysed in two or more discriminatory and non-discriminatory multivariate statistical models to generate a probability value and error rate relating to each source in the reference dataset. Where the probability value and error rate meet a predetermined threshold for one source, the polysilicon of unverified source is verified as being derived from that source.