WO2015062618A1

WO2015062618A1 - Apparatus and method for x-ray transmission analysis of a mineral or electronic waste sample

Info

Publication number: WO2015062618A1
Application number: PCT/EP2013/072471
Authority: WO
Inventors: Karl Simon Alexander HANSSON; Mikael Bergqvist; Mattias EKHOLM; Kevin REBENIUS
Original assignee: Orexplore AB
Current assignee: Orexplore AB
Priority date: 2013-10-28
Filing date: 2013-10-28
Publication date: 2015-05-07
Anticipated expiration: 2016-04-28

Abstract

An apparatus and a method for estimating a quantity of at least one predetermined element within a sample, and preferably within a mineral or electronic waste sample, is disclosed. The method comprises: irradiating the sample with X-ray radiation and measuring radiation transmitted through the sample when irradiated by the X-ray. The transmitted radiation is determined within two adjacent, narrow energy bands, which are arranged on different sides of an element specific X-ray absorption edge of the at least one predetermined element. Based on the relationship between the measured radiation transmitted within these adjacent energy bands, the quantity of the specific element within the sample is determined.

Description

Apparatus and method for X-ray transmission analysis of a mineral or electronic waste sample

Technical field

The present invention relates to an apparatus and method for X-ray transmission analysis of a sample, and in particular in a mineral or electronic waste sample.

Background of the invention

In geological applications much is to be gained from a fast and accurate method to determine the abundance of elements in a mineral sample. This was disclosed in the patent US 8 515 008 by the same applicant. Moreover, a similar analysis also has huge advantages in the recycling of electronic waste material.

Geological exploration is often done by means of drilling and subsequent analysis of the drill cores and cuttings. The time it takes before the results are delivered is often long when standard techniques such as fire assay are being used. This is mainly because the samples need to be sent to a laboratory where many steps of sample preparation are needed.

Alternative methods bases on x-ray fluorescence (XRF) are available.

However, existing XRF methods often work at low energies for which only the surface layer of the material can be analyzed. Moreover, for portable equipments it is cumbersome to study more than a small fraction of the surface of the sample. The method and system disclosed in US 8 515 008 greatly alleviates these problem. Nevertheless, XRF methods will always be influenced by the actual distribution of the element within the sample, at least to some extent. The detection limits of such methods are also ultimately limited to the speed at which the spectrometer can operate.

In general, an x-ray analysis using a transmission detector is typically much faster than x-ray spectroscopy in terms of photon count rate. Such an approach is disclosed in AU 2011 203239, in which X-ray radiation transmitted through a sample is used to determine the material composition of a sample. However, this system has problems in identifying and measuring the content of samples comprising several elements, and especially elements having low abundances.

The energy dependent attenuation coefficients for elements differ. The strongest characteristics are the so called absorption edges, which corresponds to the energy limits where ionization of electrons in the atoms are possible. X-ray absorption spectroscopy (XAS) is a means by which the elements in a sample can be probed. XAS often utilizes x-ray sources with a broad range of energies along with spectrometers, but it can also be done with a monochromatic source that is tunable in energy (read a synchrotron) along with a transmission detector. However, an XAS system is either extremely complex and costly or can only be operated with a limited count rate.

Consequently, there is still a need for a method and system being capable of providing quick and cost-efficient estimations of the quantities of elements of interest in a sample, such as e.g. gold, and to be able to detect and estimate even very small quantities. Such a system would be highly advantageous, e.g. in mineral applications, such as in analyzing drill cores and cuttings during ore exploration and other geological exploration. It would also be advantageous for optimizing mining processes, since the concentration of elements of interest often varies significantly within the area to be excavated. Thus, quick, cost-efficient and accurate analysis of drill cores and cuttings could be used to determine the quality, extension and characteristic of a finding.

Another field where such an analysis would be highly advantageous is, as mentioned, in the recycling of electronic waste material. With known elemental abundances the process of sorting can be optimized improving the efficiency of recycling. Given the large amount of material that needs to be processed a fast method of analysis for this application is desired.

Summary of the invention

In view of the above, an object of the present invention is to solve or at least reduce the problems discussed above. In particular, an object is to achieve a method and apparatus for providing a quick, accurate and cost-efficient way of estimating the quantity of one or several elements of interest within a sample, and in particular in a mineral or electronic waste sample. This object is fulfilled in accordance with a method and an apparatus as defined in the appended claims.

According to a first aspect of the invention, a method is provided for estimating a quantity of at least one predetermined element within a sample comprising the steps:

irradiating the sample with X-ray radiation;

measuring radiation transmitted through the sample when irradiated by the X- ray, wherein the transmitted radiation is determined within two adjacent energy bands, said energy bands being arranged on different sides of an element specific X- ray absorption edge of said at least one predetermined element; and estimating, based on the relationship between the measured radiation transmitted within said adjacent energy bands, the quantity of said specific element within said sample.

It has been found that this method provides a very accurate and reliable estimation of the quantity of the element(s) of interest. In experiments it has been verified that e.g. gold (Au) can be detected in quantities of 1 ppm, or even lower. The measurements can also be made in a rather short time, such as in a few minutes. Further, the apparatus does not require any spectrometer or expensive energy discriminating detectors, and can be realized with relatively low cost equipment.

Thus, the present invention makes it possible to determine the quantity of an element of interest within a sample by use of e.g. transmission detectors, instead of conventional spectrometers. Not only are such transmission detectors less costly than spectrometers, but also faster. Presently known transmission detectors have the capability of detecting more than 1000 times as many photons per second as spectrometers in the same price range. Further, spectrometers often have a rather limited resolution. Transmission detectors usually do not have any energy resolution. However, with the specific use of transmission detectors as explained in the present application, to detect energies around absorption edges, the effective resolution around the absorption edges are much higher than for spectrometers.

The method and system of the present invention can be used in many different fields. However, a particularly advantageous use is in detecting and estimating the quantity (i.e. abundance) of elements of interest, such as e.g. gold, in mineral applications, such as in analyzing drill cores or cuttings during ore exploration and other geological exploration. It is of particular interest for exploration for a specific element of interest. It can also be advantageously used for optimizing mining processes, whereby determination of the quality, extension and characteristic of a finding can be used to better control the overall mining process. Hereby, time and money spent on rock and material of low interest can be avoided. This enables existing findings to be developed more efficiently.

Another field of application for the new method and system is for material sorting etc of waste. Hereby, waste can be analyzed, sorted and reused/recycled. In particular the method/system may be used for waste electrical and electronic equipment (WEEE).

Absorption and attenuation of X-ray radiation transmitted through a sample follows Beer Lambert's Law:

J fj_n mm g— Ι ^β^) where I is the intensity of beam with energy E, after passing through the sample, I₀ is the intensity of the beam from the source, 1 is the distance traveled through the sample, N is the number of particles per volume and 6 is a measure of the probability of interaction with the material, which depends on the energy, E. The intensity I may be measured as the charge collected by the transmission detector, and the intensity Io would then be the charge collected with no attenuation present.

Thus, by estimating or measuring the transmitted intensity I, as a function of the energy, E, information is achieved which is correlated to the quantities of the elements contained in the sample. Each element provides one or several absorption edge(s), where the spectrum curve (intensity in relation to energy) forms a stepwise transition. The energy absorption edge at the highest energy is referred to as the K- edge, and the second highest as the L-edge.

Absorption edge is here generally used to mean an abrupt increase in the degree of absorption of X-ray radiation by a substance as the frequency of the radiation is increased. The absorption edges are related to the sharply defined levels of energy that electrons occupy in atoms.

The specific energies at which the absorption edges occur are different and unique for each element. The K-edge for gold (Au) is e.g. found at 81 keV.

The present invention is based on the realization that by using per-se known or easily determinable specific energies corresponding to absorption edges for different elements, it is possible to determine the presence of a specific element, or several elements by successive measurements, by comparing the transmitted X-ray at narrow energy bands immediately before and after the specific energy of the absorption edge for the element(s). It has also been found that the relative intensity difference obtained at the absorption edge can provide a very reliable measure of the quantity of the element in the radiation path through the sample.

The absorption edges are specific for each element. Over 22 keV - the energy of the L-edge of uranium -the stable elements, with atomic number above 44, only have K-edges. The energies of these absorption edges are well separated, and increase in correlation with the atomic numbers. Thus, in order to avoid problems with possible L-edges, the method of the present invention is particularly suited for estimating the quantity of elements having an atomic number exceeding 44. Such elements are here referred to as heavy elements. Over 22 keV the X-ray radiation also has a penetration depth that makes it practically useable, and the penetration depth increases with increasing X-ray energy. In particular, the present invention is therefore highly suitable to determine quantities/abundances of said heavy elements, such as, but not limited to, Ag (atomic number 47) and/or ,Hf (atomic number 72), Ir (atomic number 77), Pt (atomic number 78), Au (atomic number 79). Further, it is highly suitable to determine quantities/abundances of rare earth elements, and in particular elements within the lanthanide group (atomic numbers 57-71).

One realization of the present invention involves measuring transmitted X-ray radiation through a sample three times, with different filters, and preferably transmission filters, and, based on this, to determine transmitted X-ray radiation within two narrow bands on each side of an absorption edge for the element of interest. Further, the relative intensity difference of these two bands enables a very accurate estimation of the quantity of this element within the sample. Further, as is discussed more thoroughly in the following, an even better estimate, especially for low abundances, is obtainable by considering also the continuum slope in the estimate.

The transmitted radiation is preferably measured separately in a plurality of pixels of a one or two dimensional pixel array of an X-ray detector, and wherein an estimate of the quantity of said specific element within said sample is made separately for each pixel. Hereby, it is also possible to determine not only the total quantity of the element of interest within a large part of the sample, but also to determine where and how the element is located within the sample.

The sample may also be arranged in a sample holder which is moveable in relation to the irradiation path, or alternatively, the radiation path may be moveable in relation to the sample. For example, the sample may be rotatable, and/or moveable in a linear direction. Successive measurements may then be made to measure and scan a larger part of the sample.

The radiation transmitted through the sample may be estimated within a plurality of pairs of adjacent energy bands, each of said energy band pairs being arranged on different sides of an element specific X-ray absorption edge of a corresponding plurality of elements; and wherein, based on the relationship between the measured radiation transmitted within said pairs of adjacent energy bands, the quantities of said specific elements within said sample are estimated. Hereby, measurement and quantity estimation (abundance estimation) can be made of more than one element of interest. Notably, one filter may be common between two pairs of filters, so that two pairs are obtainable with the use of three distinct filters.

The energy bands are preferably narrow, in order to avoid that more than one element is measured at each time. Preferably, the energy bands each extends over a range of less than 15 keV, preferably less than 10 keV, and most preferably less than 5 keV. However, the difference in energy between K-edges of elements decreases with decreasing energy. For example, the difference between the K-edges for Z=79 and Z=80 is 2.4keV, and the difference between the K-edges for Z=47 and Z=48 is

1.2keV. In particular, the above-discussed specific energy ranges are applicable when measuring elements having an atomic number exceeding 44.

The energy bands on each side of the absorption edge of the element to be estimated, with atomic number Z, preferably has a width ranging from the corresponding absorption edge of element Z-a to the corresponding absorption edge of element Z+b, wherein a and b are 5 or less, and preferably 3 or less, and more preferably 2 or less, and most preferably 1.

The transmitted radiation is preferably sequentially determined within said two adjacent energy bands. However, sequential measurements could be made in parallel at different subsamples, in which case simultaneous measurements could be made in both/all energy bands, but sequentially in each sub-sample.

The measuring of radiation transmitted within specific energy bands is preferably made by filtering the radiation, either before or after passing the sample, and measuring the transmitted radiation when using at least two different filters, and preferably at least three different filters. In case three filters are used, each filter may be arranged to essentially prohibit transmission of radiation above or below a certain threshold value, wherein differences between measurements with said filters provide estimates of the transmitted radiation within each energy band. Thus, the filters here provide relatively broad, overlapping measuring ranges. However, by subtracting the measurements from one range from another, estimations of the transmitted radiation within narrow bands outside the overlap are determinable.

Other ways of obtaining measurement in narrow energy bands are feasible. For example, filters may be provided in or inside the X-ray source, between the X-ray source and the sample, between the sample and the detector, or even in the detector(s).

The X-ray irradiation is preferably provided within a broad polychromatic X- ray range, and preferably within a range of 1-300 keV, and most preferably within the range 10-200 keV. However, the optimal ranges depend on what element(s) that are to be measured.

The estimation of the quantity of the specific element within the sample may further comprise the step of correcting the estimate based on the relationship between the measured radiation transmitted within the adjacent energy bands with a correction factor being dependent on the matrix material of the sample. Knowledge or estimation of the matrix material, which typically primarily comprises light elements (Z<30), can be obtained in various ways. However, preferably the composition etc of the matrix material is determined by additional X-ray fluorescence measurements. Such additional measurements can be obtained simultaneously with the transmission measurements, by simultaneous detection of fluorescent radiation obtained during the transmission measurements. However, it may also be obtained in a separate step, before or after the transmission detection.

According to another aspect of the invention, an apparatus is provided for estimating a quantity of at least one predetermined element within a sample comprising:

an X-ray source for generating an X-ray beam to irradiate the sample;

at least one X-ray detector to measure radiation transmitted through the sample when irradiated by the X-ray, wherein the transmitted radiation is determined within two adjacent energy bands, said energy bands being arranged on different sides of an element specific X-ray absorption edge of said at least one predetermined element; and

a processing unit for estimating, based on the relationship between the measured radiation transmitted within said adjacent energy bands, the quantity of said specific element within said sample.

Hereby, similar advantages and specific features and embodiments as discussed above in relation to the first aspect are achievable and useable.

The X-ray detectors may contain filters, or alternatively filters may be arranged between the X-ray source and the detectors. In the latter case, the filters may preferably be arranged between the sample and the detectors, but alternatively the filters may be arranged between the X-ray source and the sample.

Preferably, at least two, and most preferably at least three, filters are provided, each filter prohibiting radiation above or below a certain threshold energy.

The filters can preferably be placed on an automatic filter changing mechanism, e.g. in the form of a wheel or other rotational or translational

arrangement, enabling fast filter switching.

The filters may be provided in or inside the X-ray source, between the X-ray source and the sample, between the sample and the detector, or even in the detector(s) The X-ray detector preferably comprises a one- or two-dimensional pixel array of X-ray detectors positioned with respect to the X-ray source and adapted to detect, in respect of each pixel in the array, two or more, preferably three, measures of X-ray transmission through the sample, each pair of measures corresponding to a distinct X-ray energy range.

The processor unit is preferably arranged to determine an equivalent thickness of said at least one predetermined element in respect of each pixel. From this the quantity of the predetermined element(s) is in turn determinable. The X-ray detector may be a two dimensional pixel array of X-ray detectors in the form of at least one of: a charge coupled device (CCD), a pixellated solid state detector, such as a CMOS detector, a detector fitted with a scintillator screen, a direct converting detector, such as CdTe, a single X-ray quantum counting detector system capable of discriminating the energy of individual X-ray photons, or any

combinations of these. Other objectives, features and advantages will appear from and be further elucidated by the following detailed disclosure, from the attached dependent claims as well as from the drawings. Brief description of the drawings

The above, as well as additional objects, features and advantages of the present invention, will be better understood through the following illustrative and non-limiting detailed description of preferred embodiments of the present invention, with reference to the appended drawings, where the same reference numerals will be used for similar elements, wherein:

Fig. 1 illustrates a simulated curve illustrating number of detected photons as a function of energy when passed through a foil of gold;

Fig. 2 illustrates a schematic view of an apparatus in accordance with an embodiment of the present invention;

Fig. 3 illustrates the transmission through two exemplary filters;

Fig. 4 illustrates the transmission through exemplary narrow band

transmission filters;

Fig. 5 illustrates the use of pairs of transmission filters to measure narrow energy bands around the K-edge for an element of interest;

Fig. 6 illustrates the difference between adjacent narrow energy bands in dependence of the equivalent thickness of the element of interest;

Fig. 7 illustrates the difference (quotient) between the interaction probability 6 at 79 keV and at 82 keV, in dependence on the atomic number; and

Figs. 8a-d are transmission images obtained in an experimental test of a method and system according to an embodiment of the present invention.

Detailed description of preferred embodiments

Fig. 1 is an illustration of a characteristic absorption edge for gold. The graph illustrates the number of photons - corresponding to the particle flux - detected when passing through a thin foil of gold (30 micron) in relation to the energy of the transmitted radiation. As is clearly determinable from the graph, there is a significant stepwise reduction in the intensity at 81 keV, at the absorption edge for gold. Fig. 2 is a schematic illustration of an apparatus for estimating a quantity of at least one predetermined element, such as gold, within a sample. The apparatus comprises an X-ray source 1 for generating an X-ray beam, to irradiate a sample 2. At least one X-ray detector 3 is provided to measure radiation transmitted through the sample when irradiated by the X-ray. Further, a filter arrangement 4 is provided, to delimit the energy band of the X-ray radiation. The apparatus may further comprise an outer casing (not shown) enclosing all or some of the other parts. The housing may be closeable by means of a lid, a door or the like.

A controller 5, including a processing unit, is arranged to receive data about the transmitted radiation within two adjacent energy bands, arranged on different sides of an element specific X-ray absorption edge of at least one predetermined element, and to, based on the relationship between the measured radiation transmitted within said adjacent energy bands, estimate the quantity of said specific element within said sample.

The X-ray source 1 may be a conventional X-ray tube, and may be equipped e.g. with a Soller collimator, comprising a stack of metal plates spaced a few tenths off a millimeter apart, to form a collimated beam. The X-ray source 1 generates one or several X-ray beam(s) having a principal direction parallel to the plane in which the X-ray source 1 and the detector(s) 3 are arranged. However, a collimated beam is not necessary. For example, the beams emitted from the X-ray source may be fan- or cone-shaped, originating from the tube anode, in one or two dimensions. The X-ray radiation should have an energy exceeding the characteristic absorption edge energy for the element(s) of interest. Preferably, a polychromatic X-ray source is used, preferably generating energies within the range of about 1 keV to about 300 keV, and more preferably within the range of about 10 keV to 200 keV. The X-ray source may operate at a peak voltage of between 30kVp and 300kVp, preferably between 40kVp and 200k Vp. The narrow energy bands to be measured are preferably obtained by a suitable filter arrangement.

Preferably, the X-ray source provides a relatively focused and narrow focal spot size, which enables a high resolution for improved detection of particles or clusters of particles. Preferably, the beam focal spot size, i.e. the area of the surface upon which the beam is impinged, has a diameter (or full width-half-maximum in case of a spot with Gaussian profile) within the range of 1-200 microns, and preferably within the range 10-150 microns, and more preferably within the range 20- 120 microns, such as 30, 50 or 100 microns.

The X-ray source may be equipped with a filter in order to make the original X-ray spectrum harder. The filter may be in form of a bowtie filter, or the like, in order to flatten the intensity distribution reaching the detector, thereby also

prolonging the lifetime of the detector.

The sample 2 may be contained in a sample container arranged on a sample holder, or be arranged directly on a sample holder 6, depending on whether the sample is in powder or slurry form, or the like, or in solid form. The sample holder 6 preferably enables movement of the sample in at least one direction. The sample may e.g. be rotatable about its longitudinal axis, or another axis not being parallel to the X- ray beam direction. It can also preferably be translatable along its longitudinal axis, or any other direction not being parallel to the X-ray beam direction. Hereby, the entire sample may be measured. Movement of the sample and sample holder is preferably controlled by the controller 5.

Thus, separate measurements are preferably made at different angles, generated by rotating the sample about an axis that is substantially not parallel to the X-ray beam direction. Additionally or alternatively, separate measurements may be obtained by translating the sample between measurements. A small number of unique angles, for instance from 2 to 20, or any number in between, may be used for each distinct energy band measurement. More angles provide more information, but makes the analysis more complicated.

Based on such detection, the controller may be arranged to compute a tomographic reconstruction of particles of the element of interest, using the obtained measurements at different angles. If tomography is used, the number of angles is preferably even higher, such as exceeding 50, or exceeding 60, or even higher.

However, alternatively, the measurements can be made at only one angle.

The X-ray detector 3 preferably comprises a one- or two-dimensional pixel array of X-ray detectors. A two dimensional pixel array of X-ray detectors may e.g. be in the form of a charge coupled device (CCD), or a pixellated solid state detector. The detectors may be fitted with a scintillator screen. Further, a direct converting detector, such as CdTe may be used. Further, a single X-ray quantum counting detector system capable of discriminating the energy of individual X-ray photons may be used. Still further, various combinations of such detectors may be used.

The pixel size of the detector arrays may be in the range of a few microns to hundreds of microns. However, both smaller and larger pixels can be used as well. By placing the sample between the source and detector, the image of the sample on the detector is magnified. The magnification can be adjusted by altering the

source/sample and source/detector distances to obtain the desired resolution.

In a preferred embodiment, the detector comprises a TDI (time delay and integration) scanner or a image detector being sensitive to X-ray within the energy range of interest. The detection of the X-ray radiation can be either direct, as in a CdTe detector, or indirect, as in a scintillator based detector.

The detector can further be provided with a collimator to prevent detection of scattered radiation. A collimator may also be arranged between the sample and the detector.

The transmission detector(s) 3 may preferably be located along the principal direction of the X-ray beam in such a way that it faces the X-ray source 1, to best measure transmission of X-rays in the irradiation path through the sample. During analysis the transmission detector measures the X-rays passing through the sample.

The filter arrangement 4 preferably comprises a filter wheel, or other arrangement, with at least two filters, and most preferably at least three filters per element to be analyzed. The filter arrangement is preferably controllable by the controller 5, e.g. by being connected to a servo. The filters are preferably large enough to cover the entire cross-section of the X-ray beams. The filters preferably comprise different materials, and with predetermined thicknesses, in order to provide adequate filtering characteristics. This will be exemplified further in the following.

The filters 4 may be arranged between the X-ray source 1 and the sample 2, as illustrated with solid lines in Fig. 2. Additionally or alternatively, the filters 4' may be arranged between the sample 2 and the detector 3, as illustrated in dashed lines in Fig. 2. Arrangement of the filter arrangement close to the X-ray source can have smaller dimensions, and still cover the entire cross-section of the radiated beams. Smaller filters are easier and less costly to produce. Further, since filters will generate some scattered radiation, filters arranged closer to the X-ray source will also generate less scattered radiation that may be detected by the detectors than filters arranged closer to the detectors. On the other hand, filters arranged after the sample will not attenuate the radiation prior to the sample, which e.g. is an advantage if fluorescence measurements are also to be carried out (see below).

The transmission through a filter is dependent both on the material of the filter and the thickness of the filter. The absorption edges of the material(s) of the filter play a significant role in defining the characteristics of the filter. Two elements being adjacent to each other in the periodic system (atomic number Z and Z+l) will have similar filter properties. The thickness of two filters comprising element Z and Z+l, respectively, can easily be adjusted to make the difference between them essentially zero, except in the area between their absorption edges.

As an example, the transmission through a filter comprising primarily, and preferably essentially only Au, and having a thickness of 400 microns, and a filter comprising primarily, and preferably essentially only Pt, and having a thickness of 376 microns, are illustrated in Fig. 3. Apart from the range extending between the absorption edges, the characteristics of these filters are very similar, and here, the difference between these filters is less than 1/1000 for all energies outside the range extending between the absorption edges. The filters could additionally be fine tuned by adding very small amounts of other materials in order to fine tune the differences outside the energies between the primary K-edges to differences less than 1/1000.

Thus, having a pair of filters of elements being close in the periodic system, and a suitably chosen thickness, the difference will be found essentially only in the narrow energy range between the absorption edges of these elements. Such a pair of filters may consequently be used to determine the transmission within the narrow energy band between these absorption edges. The exact thicknesses to be used depend on the element used in the filter. For measuring Au, it has been found that the relative filter thicknesses of Au - 100%; Pt - 94% and Hg - 137.8% work very well. Such thicknesses minimize the difference between the filters outside the area close to the K-edges of these elements. Further, a thicker filter provides a narrower wavelength range through which radiation is transmitted. This reduces the number of photons of uninteresting energies. However, the number of photons in energy ranges of interest, i.e. within the measured energy bands, will also be reduced. However, all elements have different filter characteristics, and the thicknesses are consequently strongly element dependent. Thus, the differences between various materials are significant, and also depending on the density and attenuation of the filter material. The thickness needs to be adapted to the filter density, attenuation and material in question, and the element to be measured. For a filter composed of a single element, useful thicknesses could preferably range from about 10 microns up to about 3000 microns.

To provide measures of the radiation transmitted through the sample within two adjacent energy bands being arranged on different sides of an element specific X- ray absorption edge of a predetermined element of interest, it is preferred to use two pairs of such transmission filters wherein the difference between the absorption edges of these transmission filters provide the narrow energy bands. Preferably, one of the filters is common between the pairs, so that three filters are used to form two pairs. When the element of interest has atomic number Z, the first filter pair preferably comprises an element with atomic number Z and an element with atomic number Z-a, respectively, and the second filter pair preferably comprises an element with atomic number Z and an element with atomic number Z+b, respectively. Here, a and b are preferably 3 or less, and preferably 2 or less, and most preferably 1. The thicknesses of the filters are preferably selected to obtain an essentially identical transmission outside the range between the absorption edges. For example, filters comprising essentially only Pt, Au and Hg may be used to form pairs - Pt + Au, and Au + Hg, respectively - to measure narrow energy ranges to estimate the quantity of Au in a sample. In Fig. 5, this is illustrate schematically. The graph illustrates the transmission through 30 microns Au, and the left hand band illustrates the transmission difference between Pt and Au, whereas the right hand band illustrates the transmission difference between Au and Hg.

However, instead of Z+l and Z-l, it is also possible to use e.g. Z+2 and/or Z- 2, or Z+3 and/or Z-3, with essentially the same reliability of the estimation. For Z-a, a value of "a" larger than 1 would mean that a wider energy band below the K-edge is measured. However, in most practical situations, this would have a very limited impact on the accuracy of the estimation. For Z+b, a value of "b" larger than 1 would include the K-edge of one or more other elements, having atomic number above Z but below Z+b. In case there is a significant quantity of such other elements, this may lead to problems. However, in many practical situations, the presence of e.g. elements Z+l and Z+2, when measuring element Z, would be very limited, with a low abundance. Thus, inclusion of elements of low abundance compared to the element of interest in the measured band can often be accepted.

For practical reasons, it may e.g. be easier to use a Pb filter (Z=82) when detecting Au, than Hg (Z=80), since Hg is toxic. Further, the intermediate elements Hg (Z=80) and Tl (Z=81) are very rare in most situations.

A relative intensity difference a close to the absorption edge of the element of interest corresponds to the quantity of the element of interest that has been passed. Fig. 6 illustrates, as an example, the difference a for Au in dependence of the equivalent thickness, i.e. the thickness of Au. For smaller quantities, the difference a is logarithmically proportional to the equivalent thickness through which the radiation has passed. For larger quantities, the proportionality is not linear, but may still be used to estimate the exact quantity with good accuracy, especially if combined with the total intensity of the transmitted radiation. Thus, the precise information obtained of the transmission close to an absorption edge of an element of interest can consequently be used to estimate, with great accuracy, the equivalent thickness and quantity of the element of interest.

Such pairs of suitably selected transmission filters provide extremely precise and narrow energy bands. However, it is also possible to use other types of narrow band filters instead. Such narrow band transmission and diffraction filters are per se known. For example, Fig. 4 illustrates the transmission through two exemplary narrow band transmission filters is illustrated. The filters are a 2000 microns Au filter, and a 1740 microns Pt filter. Even though such filters may be used to provide the narrow energy band measurements, a disadvantage is the relatively large overlap between the filters, and the low transmission obtained at e.g. 80 keV, compared to the pair of transmission filters illustrated in Fig. 3.

The controller 5 is preferably connected to and arranged to control the operation of at least one, and preferably all, of: the X-ray source, the filter arrangement, the sample holder and the detector(s). Further, the controller receives detection data from the detector(s). In respect of the X-ray source, the controller may control and adjust the X-ray tube voltage in accordance with e.g. the element to be measured. The filter arrangement may be controlled to automatically switch filter. The sample holder may be controlled to move the sample during measurement, either intermittently or continuously.

The processing unit of the controller may be a conventional CPU, on which runs software in order to process input data to obtain the resulting X-ray analysis. A multichannel analyzer (MCA) can also be provided between the detectors and the processing unit.

In many practical situations, the X-ray source will have some fluctuations in the provided intensity. If the intensity varies between measurements with different filters, this may affect the accuracy of the resulting estimations. To this end, part of the radiation emitted from the X-ray source may be detected directly, without passing the sample, to detect such fluctuations. Further, if fluctuations occur, the controller may use the detected information of such fluctuations to provide corrections and compensations in the estimates made, or the like.

Further, the controller is preferably connected to an external display (not shown) in order to display the results of the X-ray analysis to a user, a printer for printing the results, and/or other suitable user interfaces.

Based on the measurement data provided by the transmission detectors, as measured at different energy ranges and/or with different filters, the controller calculates the transmission at two narrow energy bands, arranged on both sides of an absorption edge of the element of interest. The calculation can be made on predetermined correlations between the difference between these energy bands and the equivalent thickness of the element of interest present in the path of the radiation. Such predetermined data may be determined experimentally, by computer simulations, and the like.

However, the intensity of the transmitted radiation, and the difference in intensity between the energy bands, also to some extent depends on the radiation continuum, and how this varies over the energy range in question. In particular, this is the case when there are low quantities of the element to be measured. Added material in the path of the radiation will always lower the intensity in the lower energy range more significantly than in the upper energy range, except for the element of interest.

When the matrix material is homogeneous, the slope of the continuum will depend solely on the amount of matrix material, which is determinable from the attenuation. Thus, measurement of the overall transmitted intensity may be used to determine the slope of the continuum, and thereby provide additional means for correcting the estimates of the quantity of the element to be measured.

However, when the matrix material is inhomogeneous, additional corrective measures may be taken, since different materials behave differently in this respect. Fig. 7 illustrates the difference (quotient) between the interaction probability 6 at 79 keV and at 82 keV, in dependence on the atomic number. Gold (Z=79), where

is ^not included. These two energy levels are between the absorption edges of Pt, Au and Hg. As can clearly be derived from this, elements having similar atomic numbers behave similarly in this respect, whereas there is much difference in particular between light elements (Z<30) and heavier elements (Z>30).

When the element of interest is not present, the influence of the slope of the continuum on the two energy bands may generally be expressed as:

where ii and i₂ are the intensity of the first and the second energy band, respectively, and δ is a correction factor. For homogeneous matrix materials, δ can be determined by measuring the intensity of any of the energy bands. However, for non- homogeneous materials, the correspondence between δ and the intensity of a single energy band is not unique. Accordingly, additional information may be provided, in order to improve the correction and the final estimation even further.

For example, it is possible to provide a better correction by making additional transmission measurements, e.g. by using additional filters, by using different settings for the X-ray source, etc. By such additional measurements, further information of the appearance of the spectrum will be obtained, which may be used to determine δ more exactly.

Additionally or alternatively, the total thickness of the material in the radiation path, t, may be determined, either by measuring the thickness mechanically or optically, by estimating the thickness by computer tomography, by pre-knowledge (since drill cores often have standardized dimensions), or in other ways, as are per se known in the art. The thickness is a parameter which has a great impact on δ, and in which range it could possibly be. Thus, estimation/measurement of the thickness may be used to improve the estimation of the correction factor, and, thus, to improve the overall estimate of the quantity of the element to be measured. Tomography also has the advantage that the absorption and abundances can be determined in smaller cells, and thus alleviating the risk of mixing of materials in the analysis.

The parameter δ is also, as mentioned above, highly dependent on the material composition of the matrix material. Thus, additional information regarding the material composition may be used to determine δ more exactly. For example, it is possible to obtain such information by performing an additional X-ray fluorescence (xrf) measurement. Such a measurement can be made sequentially before or after measurement of the transmitted X-ray, but may also be made in parallel. The X-ray fluorescence detector and the transmission detector(s) are preferably located apart from each other. This minimizes the overlap in the measured signals. The

transmission detector(s) is preferably arranged directly opposite to the X-ray source, whereas the fluorescence detector(s) may e.g. be arranged in a lateral position in relation to the primary radiation path. In particular, it is preferred that the

fluorescence detector is arranged at an angled position within a range of 45-170 degrees relative to the primary radiation path. The differential cross-section for Compton radiation has a radiation minimum at 90 degrees. Consequently, such an arrangement reduces the background radiation below the fluorescence peaks.

However, an angle at about 125 degrees from the source to target vector is often preferred, since the X-ray illuminated side of the sample will shine more brightly in fluorescence radiation.

Determination of a material composition by xrf may e.g. made with a system and method as disclosed in US 8 515 008 by the same applicant, said document hereby being incorporated in its entirety by reference. In particular, a combination of xrf and transmission measurement may be used, as disclosed in said patent, to determine the elements and material composition of a sample. This information may then be used to provide an improved estimate of the correction parameter δ.

However, other means of determining or estimating the material composition of the matrix material are feasible. For example, the general compositions of the matrix materials are already known to a fairly high degree at most mining facilities and the like.

It has been found by the present inventors that the radiation detected by the detectors in the above-discussed set-up will in most cases primarily be transmitted radiation, and that scattered radiation, based on Rayleigh and Compton scattering, is negligible. Under normal conditions, the scattered radiation will be less than 1%. This has been verified using a collimated source and a line scanner. For a less collimated source, e.g. when using a 2D detector, the fraction of scattered radiation could be somewhat larger. However, when there is a large attenuation, e.g. due to a large quantity of the element to be measured, corrective measures may be adopted, to improve the estimations even further. For example, by using multiple detectors/pixels, a picture, or absorption map, of the irradiated sample may be generated. This may further be used to form a model of the sample. This model is preferably a scattering model of the sample, which provides corrections in order to remove the component which is related to scattering. Still further, a part of the detector, or a separate detector, may be used to measure radiation from the X-ray source which has not passed through the sample, which may consequently be used to determine the attenuation provided by the sample, and also to measure the amount of scattered radiation incident on the detector.

If the attenuation is determined to be very high, causing low signal to noise in the transmitted images, this may be used to generate an alarm, to mark these measurements with a warning, or the like.

Further, filters may sometimes have imperfections, such as variations in thickness. This may also be compensated by the controller. For example, calibration images may be generated with a filter present but when there is no sample in the radiation paths. If the sensors have already been calibrated, without any filter present, any variations occurring when the filter is present is due to the filter imperfections. These variations may be measured, and later used to generate pixel-to-pixel corrections to be used on the transmission images to be generated when the sample is measured.

Still a further correction that may be contemplated is to correct for movement of the sample. Each part of the sample would typically be radiated and measured sequentially, each time with a different filter. In some environments, it is possible that the sample may move inadvertently between these measurements, due to vibrations, shocks etc. Such small displacements may be corrected by the controller by identifying artifacts or other distinguishing features of each image, and correlating the images related to different filters based on these artifacts/distinguishing features. Typically, sharp intensity gradients would be distinguishable, and may be used to correlate the images, and identify possible inadvertent movement that has occurred between the measurements. In particular, it is noted that a 2D transmission detector provides images having more position information than ID images, measured by the use of a line scanner. In addition to the above-discussed corrections, the accuracy of the estimates is also related to the resolution. The resolution is inter alia dependent on the size of the focal point of the X-ray source radiation, the pixel size of the detector, and the distances between the sample, X-ray source and detector. A high resolution is particularly advantageous for determining the quantity, size and shape of large and arbitrarily shaped particles of the element to be measured.

In an exemplary method for estimating the quantity of an element of interest within a sample, the following steps may be executed:

a) Calibration of the transmission detector by providing irradiation from the X-ray source without any filter or sample;

b) Reception of transmission images with three different filters, e.g. Z-l, Z and Z+1, and preferably at least one additional filter, Yl, or at least one additional setting of the X-ray source, Y2;

c) Calibration of the images to correct variations in filter thicknesses; d) Calibration of the images to correct variations in the X-ray source intensity;

e) Correlate the different images in relation to determined position deviations;

f) Calculate measurement data for Z-(Z-l) and (Z+l)-Z for each pixel, thereby generating two images on each side of the absorption edge of the element to be measured;

g) Correct, if necessary, for scattered radiation and fluorescence;

h) Correct, if necessary, for limited resolution;

i) Find a model for each pixel in a reference library by matching intensities for Z+1 and Yl and/or Y2;

j) Translate the information about the absorption edge (K-edge) of each pixel, preferably from f), to an equivalent thickness of the element to be measured, preferably based on i);

k) Calculate the mass of the element to be measured based on a geometrical model of the sample;

1) Calculate the mass of the matrix material from Z-l and Yl and/or Y2 - such calculations are per se known; and

m) Calculate the quantity, measured as ppm, of the element to be measured, based on k) and/or 1).

However, all the above-discussed steps are often not necessary, and can be omitted, even though a more thorough analysis improves the results even further. For example, the calibration steps, such as steps a, c and d, may be omitted if e.g. calibration has already been made previously, and the circumstances are still similar. Further, several of the correction steps, such as steps e, g, and h, may often be omitted, dependent on the circumstances. Further, several of the steps may be performed differently, as has already been discussed thoroughly in the foregoing.

In an experimental test, a sample was prepared by adding small parts of Sn, Pb and Au to a container of rock dust. For the measurement, an X-ray tube having 120 kV anode voltage, and a tube current of 5 mA was used. Three filters were used: an Au filter, having a thickness of 400 microns, a Pt filter having a thickness of 475 microns, and a Pb filter, having a thickness of 623 microns, and measurements were obtained using each of said filter sequentially. A scintillator based line scanner having 256 pixels was used as the transmission detector. The sample was arranged 24 cm from the X-ray source, and 13 cm from the detector. The measurement lasted for only a few minutes. The same measurement was also made on a calibration sample, having known properties, and without any Au, Sn or Pb. The average value of Au-Pt = ii and Pb-Au = i₂ were calculated, and generated an estimate of δ. These correction factors were then subtracted from each measurement value/pixel for Pb-Au in the sample containing Au. Where the difference between Au-Pt and Pb-Au are significant, this is an indication of the presence of Au in the sample.

In fig. 8 a-d, transmission images obtained in the experiment are shown. Fig. 8a is a transmission image of the quotient between the transmission obtained with the Pt filter and the transmission obtained with the Au filter. Similarly, fig. 8b is a transmission image illustrating the quotient of a Pb filter and the Au filter. Fig. 8c is a transmission image combining the measurement data obtained with the Pt, Au and Pb filters. Fig. 8d finally is the same transmission image as in Fig. 8c, but where areas containing Au, as identifiable by the present invention, has been marked in a different nuance/color. .

The detected findings were later verified, by comparing it to known data about the structure and composition of the pre-prepared samples. However, the one pixel detection could either be a real detection or a spurious one. Notably, the uppermost detected gold particle weighs only 1.7 mg.

The present invention has now been disclosed with reference to certain embodiments. However, as would be readily acknowledged by a person skilled in the art, other embodiments than the ones disclosed above are equally possible. For example, the narrow energy bands may be provided by other types of narrow band filters, instead of the preferred use of pairs of filters. Further, the method/system may be arranged to measure only one element of interest, or a plurality of different elements of interest. Still further, different steps of calibration, corrections, measurement and calculations may be performed in many different ways, and may also be combined differently. Still further, the result of the measurements may be presented or used in various ways, such as being presented to the user on a display or the like, forwarded to other devices, etc. Still further, the sample may be rotated and translated in various fashions, either during the irradiation or intermittently between different measurement steps. In an alternative embodiment the X-ray source may also be translated and/or rotated while the sample container is held still, thereby providing the same relative motion as when the sample container is moved. Further, the sample container may take many different shapes and dimensions. Such and other modifications of the above-discussed embodiments must be considered to be encompassed by the invention as defined by the appended claims.

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.

Claims

CLAIMS 1. A method for estimating a quantity of at least one predetermined element within a sample, and preferably within a mineral or electronic waste sample, comprising the steps:

irradiating the sample with X-ray radiation;

measuring radiation transmitted through the sample when irradiated by the X- ray, wherein the transmitted radiation is determined within two adjacent energy bands, said energy bands being arranged on different sides of an element specific X- ray absorption edge of said at least one predetermined element; and

estimating, based on the relationship between the measured radiation transmitted within said adjacent energy bands, the quantity of said specific element within said sample.

2. The method of claim 1, wherein the transmitted radiation is measured separately in a plurality of pixels of a one or two dimensional pixel array of X-ray detectors, and wherein an estimate of the quantity of said specific element within said sample is made separately for each pixel.

3. The method of claim 1 or 2, wherein radiation transmitted through the sample is estimated within a plurality of pairs of adjacent energy bands, each of said energy band pairs being arranged on different sides of an element specific X-ray absorption edge of a corresponding plurality of elements; and wherein, based on the relationship between the measured radiation transmitted within said pairs of adjacent energy bands, the quantities of said specific elements within said sample are estimated.

4. The method of any one of the preceding claims, wherein the energy bands each extends over a range of less than 15 keV, and preferably less than 10 keV, and most preferably less than 5 keV, and in particular for measurement of elements having an atomic number exceeding 44.

5. The method of any one of the preceding claims, wherein the energy bands are narrow.

6. The method of any one of the preceding claims, wherein the energy bands on each side of the absorption edge of the element to be estimated, with atomic number Z, has a width ranging from the corresponding absorption edge of element Z- a to the corresponding absorption edge of element Z+b, wherein a and b are 5 or less, and preferably 3 or less, and more preferably 2 or less, and most preferably 1.

7. The method of any one of the preceding claims, wherein the transmitted radiation is sequentially determined within said two adjacent energy bands.

8. The method of any one of the preceding claims, wherein measuring of radiation transmitted within specific energy bands is made by filtering the radiation, either before or after passing the sample, and measuring the transmitted radiation when using at least two different filters, and preferably at least three different filters.

9. The method of claim 8, wherein three filters are used, each filter being arranged to essentially prohibit transmission of radiation above or below a certain threshold value, wherein differences between measurements with said filters provide estimates of the transmitted radiation within each energy band.

10. The method of any one of the preceding claims, wherein the X-ray irradiation is provided within a polychromatic X-ray range, and preferably within a range of 1-300 keV, and most preferably within a range of 10-200 keV.

11. The method of any one of the preceding claims, wherein the estimation of the quantity of the specific element within the sample further comprises the step of correcting the estimate based on the relationship between the measured radiation transmitted within the adjacent energy bands with a correction factor being dependent on the matrix material of the sample, the matrix material preferably being determined by additional X-ray fluorescence measurements.

12. An apparatus for estimating a quantity of at least one predetermined element within a sample comprising:

an X-ray source for generating an X-ray beam to irradiate the sample;

13. The apparatus of claim 12, further comprising at least two, and preferably at least three, filters, each filter prohibiting radiation above or below a certain threshold energy.

14. The apparatus of any one of the claims 12-13, wherein the X-ray detector comprises a one- or two-dimensional pixel array of X-ray detectors positioned with respect to the X-ray source and adapted to detect, in respect of each pixel in the array, two or more measures of X-ray transmission through the sample, each measure corresponding to a distinct X-ray energy range.

15. The apparatus of any one of the claims 12-14, wherein the X-ray detector is a two dimensional pixel array of X-ray detectors in the form of at least one of: a charge coupled device (CCD), a pixellated solid state detector, such as a CMOS detector, a detector fitted with a scintillator screen, a direct converting detector, such as CdTe, a single X-ray quantum counting detector system capable of discriminating the energy of individual X-ray photons, or any combinations of these.