EA040045B1 - X-RAY FLUORESCENCE ANALYZER WITH SEVERAL DETECTION CHANNELS AND METHOD OF PERFORMING FLUORESCENCE ANALYSIS - Google Patents

Description

Technical field

The invention relates to the field of X-ray fluorescence analysis. In particular, the invention relates to the problem of detecting relatively small amounts of fluorescent radiation from an element of interest in the presence of significant background radiation and/or fluorescent radiation from other elements.

Prior Art

X-ray fluorescence analysis can be used to detect the presence of elements of interest in a matrix of other elements and measure their concentration. For example, in the mining industry, it is important to know if a mineral or metal of interest is present in a sample and in what quantities. The method of X-ray fluorescence analysis, in order to be used in an industrial process, must be sufficiently accurate even with a relatively short exposure time, and it could be used with noise-resistant and mechanically reliable measuring devices.

A particular application of X-ray fluorescence analysis in the mining industry is the analysis of elements of interest in cuttings. By definition, slurry is an aqueous suspension of fine solids of crushed and ground ore in which the solids dry weight is less than 90%, typically 20-80% of the total mass of the sample. The fact that the sample is in the form of a slurry places special demands on the handling of the sample. For example, it is preferable to keep the sample flow turbulent so that its composition remains evenly mixed and fractions do not separate from each other. At the same time, the measurement geometry must remain as constant as possible so as not to cause undesirable geometric changes in the measurement results.

The concentrations of the elements of interest in the pulp are often very low.

For example, copper, zinc, lead, and molybdenum need to be measured at concentrations such as 0.01 percent or less, and gold concentrations that can be measured can be as low as a few ppm, such as 1 to 5 ppm. This low concentration makes measurement difficult because the intensity of the fluorescent emission from the element of interest is very low, which inevitably increases the impact of statistical errors. When the intensity is low compared to that of other emissions present, such as fluorescent emissions from other uninteresting elements, overlap with neighboring peaks causes problems. The measurement time cannot be arbitrarily long since the slurry comes as a continuous stream from the cleaning process and is an important current indicator of what is going on in the process. The X-ray fluorescence measurement must be fast enough to detect trends in pulp composition so that the measurement results can be used to control the cleaning process in real time.

The essence of the invention

The aim of the invention is to provide a device and method for performing accurate and reliable x-ray fluorescence analysis of low concentrations of elements in the pulp in difficult industrial conditions. Another object of the invention is to provide such a device and method that has the ability to adapt to different types of samples and different types of conditions.

The above and other objects are achieved by using at least two crystal diffractors and their respective detectors which detect the characteristic fluorescence emission of the same element, one of the crystal diffractors containing a pyrolytic graphite crystal.

According to the first aspect, an X-ray fluorescence analyzer is provided. The X-ray fluorescent analyzer contains an X-ray tube for emitting incident X-ray radiation in the direction of the first optical axis and a pulp processing unit configured to maintain a constant distance between the pulp sample and the specified X-ray tube in the direction of the said first optical axis. X-ray fluorescence analyzer contains the first crystal diffractor, located in the first direction from the specified pulp processing unit, and the specified first crystal diffractor contains the first crystal. The X-ray fluorescence analyzer comprises a first radiation detector configured to detect fluorescent X-ray radiation diffracted by said first crystal with a first energy resolution. The X-ray fluorescence analyzer comprises a second crystal diffractor located in the second direction from said pulp processing unit, said second crystal diffractor comprising a second crystal. The X-ray fluorescence analyzer comprises a second radiation detector configured to detect fluorescent X-ray radiation diffracted by said second crystal with a second energy resolution. The first crystal is a pyrolytic graphite crystal, and the second crystal is a material other than pyrolytic graphite. The specified first and second crystal diffractors are made with the possibility of directing the characteristic fluorescent radiation of the same element to

- 1 040045 their respective radiation detectors.

In one possible implementation of the first aspect, said second crystal is a single crystal of silicon dioxide crystal, lithium fluoride crystal, ammonium dihydrogen phosphate crystal, potassium hydrogen phthalate crystal. This provides the advantage that sharp, wavelength-dispersive diffraction can be obtained from the second crystal.

In one possible implementation of the first aspect, said first energy resolution is better than 300 eV with a reference energy of 5.9 keV. This provides the advantage that the detector can provide accurate, energy dispersive detection over a relatively wide range of wavelengths that is transmitted by the pyrolytic graphite crystal.

In one possible implementation of the first aspect, the first radiation detector is one of the following: PIN diode detector, silicon drift detector, germanium detector, germanium drift detector. This provides the advantage that the first radiation detector can combine accurate and reliable performance with compact size and solid appearance.

In one possible implementation of the first aspect, the second radiation detector is a gas filled proportional counter. This provides the advantage that a relatively good detection efficiency can be achieved at a relatively low manufacturing cost.

In one possible implementation of the first aspect, said element is gold. This has the advantage that even very low concentrations of a relatively valuable element can be detected.

In one possible implementation of the first aspect, the pulp processing unit is configured to maintain a flat surface of said pulp sample on the side facing said X-ray tube, wherein said first optical axis is at an oblique angle to said flat surface, said first crystal diffractor is located at that angle of rotation around the specified first optical axis, in which the specified flat surface of the specified sample covers the largest part of the field of view of the first crystal diffractor, and the specified second crystal diffractor is located at a different angle of rotation around the specified first optical axis. This provides the advantage that the fluorescent radiation can be collected towards the first crystal diffractor at as large a spatial angle as possible.

In one possible implementation of the first aspect, the pulp processing unit is configured to maintain a flat surface of said pulp sample on the side facing said x-ray tube, and said first optical axis is perpendicular to said flat surface. This provides the advantage that a number of measurement channels can be placed symmetrically around the X-ray tube.

In one possible implementation of the first aspect, the nominal input power of the x-ray tube is at least 400 watts. This provides the advantage that a relatively large amount of fluorescent radiation can be generated.

In one possible implementation of the first aspect, the nominal input power of the X-ray tube is at least 1 kW, preferably at least 2 kW, and more preferably at least 4 kW. This provides the advantage that even more fluorescent radiation can be generated.

In one possible implementation of the first aspect, the optical path between the x-ray tube and the pulp processing unit is direct without a diffractor located between them. This provides the advantage that most of the original incident radiation can be used and the X-ray tube can be placed very close to the sample.

In one possible implementation of the first aspect, the x-ray tube contains an anode for generating said incident x-rays, and the pulp processing unit is configured to maintain the shortest straight-line distance between said pulp sample and said anode, which is less than 50 mm, preferably less than 40 mm, and more preferably less than 30 mm. This provides the advantage that more of the original incident radiation can be used.

In one possible implementation of the first aspect, the x-ray tube is an end window x-ray tube. This provides the advantage that a small distance can be maintained between the X-ray tube and the sample while leaving enough space for the detection channels.

In one possible implementation of the first aspect, the diffractive surface of said pyrolytic graphite crystal is one of the following surfaces: singly bonded surface, curved in one direction, singly bonded surface, curved in two directions, axisymmetric nonsingly bonded surface. This provides the advantage that the most advantageous crystal shape can be selected for each application.

In one possible implementation of the first aspect, the X-ray fluorescence analyzer comprises a housing, a front wall of said housing in which an opening is made, and a holder for removably holding said pulp processing unit on the outside of said front wall and aligned with said opening in the front wall. This provides the advantage that the pulp handling unit can be easily removed for maintenance.

In one possible implementation of the first aspect, the x-ray tube and the first crystal diffractor are inside said analyzer housing on one side of said front wall. This provides the advantage that the structure is robust and good protection against accidental exposure to anything can be obtained.

In one possible implementation of the first aspect, the X-ray fluorescence analyzer comprises a filter plate located in the optical path between the X-ray tube and the pulp processing unit. This provides the advantage that the incident radiation spectrum can be suitably tuned.

In one possible implementation of the first aspect, said filter plate is located closer to the X-ray tube than to the pulp processing unit. This provides the advantage that the filter does not unnecessarily block the field of view of the detection channels.

In one possible implementation of the first aspect, the X-ray fluorescence analyzer comprises a calibration plate and an actuator capable of controlled movement of said calibration plate between at least two positions, the first of which is not in the path of the incident X-rays, and the second position is in the path of the incident X-rays. and in the field of view of the first crystal diffractor. This provides the advantage that calibration can be easily automated.

According to a second aspect, a method for performing an x-ray fluorescence analysis is provided. The method includes irradiating a pulp sample with incident X-rays and obtaining fluorescent X-rays from the irradiated sample; separating the first and second predetermined wavelength ranges from the respective first and second portions of said received fluorescent x-ray radiation by means of the respective first and second crystal diffractors, wherein the first wavelength range and the second wavelength range include characteristic fluorescent radiation from the same element, with this first wavelength range is at least twice as wide as the second wavelength range; detecting fluorescent X-rays in said first and second selected wavelength ranges using respective first and second radiation detectors, wherein the energy resolution of said first radiation detector is better than 300 eV at a reference energy of 5.9 keV, thereby creating corresponding first and second detection results; and calculating the concentration of said element in said sample from at least one of said first and second detection results.

In one possible implementation of the second aspect of the specified calculation includes calculating the total intensity of the background radiation and fluorescent x-rays from elements other than the specified element, using at least one of the first and second detection results; subtracting from the total intensity detected in the wavelength range containing the specified characteristic peak of the fluorescent x-ray emission of the element to be detected in the specified sample, the calculated total intensity of the background radiation and fluorescent x-ray radiation from all elements, except for the specified element of interest in the specified sample; and representing the result of said subtraction as a calculated intensity of said fluorescent x-ray characteristic peak. This provides the advantage that the detection accuracy can be improved by using the detection results from the two detection channels.

In one possible implementation of the second aspect, said calculation includes analyzing said first and second detection results to see if the influence of a characteristic peak from another element on the first detection result exceeds a predetermined threshold; if said analysis shows that the effect of said characteristic peak of said other element on the first detection result exceeds said predetermined threshold, then calculating said concentration of said element in said sample based on said second detection result; and if said analysis shows that the effect of said characteristic peak of said other element on the first detection result does not exceed said predetermined threshold, then the said concentration of said element in said sample is calculated based on said first detection result. This provides the advantage that the method of processing the results can be adapted to the actual occurrence of fluorescent emission from other elements.

In one possible implementation of the second aspect, said element is gold. This provides the advantage that the presence and concentration of gold can be detected despite the presence of another element with a similar characteristic fluorescence peak.

In one possible implementation of the second aspect, said characteristic fluorescent emission includes the K or L peak of an element with 30 < Z < 92, where Z is the atomic number of said element. This provides the advantage that a large number of elements can be detected.

In one possible implementation of the second aspect, the specified sample contains the specified element in a matrix consisting mainly of elements with Z < 8, where Z is the atomic number of the element. This provides the advantage that, for example, water-based slurries can be analyzed.

Brief description of the drawings

The accompanying drawings, which are included to provide a further understanding of the invention and form part of this specification, illustrate embodiments of the invention and, together with the description, help to explain the principles of the invention. In the drawings of FIG. 1 illustrates the principle of X-ray fluorescence analysis in an industrial process;

fig. 2 shows in detail an example of an X-ray fluorescence analyzer;

fig. 3 illustrates an example of the use of a calibration plate;

fig. 4 shows an example of an X-ray fluorescence analyzer;

fig. 5 shows some structural details of an example of an X-ray fluorescence analyzer;

fig. 6 is an example of a crystal diffractor;

fig. 7 - some geometric aspects of the crystal diffractor;

fig. 8 - some forms of diffractor crystals;

fig. 9 - an example of the geometry of radiation propagation;

fig. 10 is another example of radiation propagation geometry;

fig. 11 is an example of an emission spectrum;

fig. 12 is another example of an emission spectrum;

fig. 13 is another example of an emission spectrum;

fig. 14 is another example of an emission spectrum;

fig. 15 - several detection channels;

fig. 16 is an example of emission spectra;

fig. 17 - an example of a pulp processing unit;

fig. 18 - x-ray tube, the optical axis of which is perpendicular to the surface of the sample;

fig. 19 - x-ray tube, the optical axis of which is located at an oblique angle to the surface of the sample;

fig. 20 - an example of the placement of several detection channels;

fig. 21 - an example of the placement of several detection channels;

fig. 22 illustrates the measured detection accuracy of an exemplary device;

fig. 23 is the measured detection accuracy of the exemplary device;

fig. 24 is the measured detection accuracy of the exemplary device;

fig. 25 is the measured detection accuracy of the exemplary device.

Detailed description

Fig. 1 illustrates an example of the principle of using an X-ray fluorescence analyzer in an industrial process. It is typical for industrial processes that the sample to be analyzed may come as a more or less continuous stream of material, so that there is a sample handling unit or system that transfers the sample for analysis and collects it after analysis. In the schematic illustration in Fig. 1, sample 101 enters as a material stream onto conveyor 102, which in this case is the sample handling system. An x-ray source 103 generates an incident x-ray beam 104 that hits the part of the sample 101 that is within the field of view of the beam 104. Fluorescent x-rays 105 are emitted in all directions and part of it is collected in a detection system, which in FIG. 1 includes a first slit 106, a wavelength dispersive diffraction crystal 107, a second slit 108, and a radiation detector 109. The installation may include a computer control system 110 that can control the control subsystems 111 and 112 of the conveyor 102 and the X-ray fluorescence analyzer 113, respectively.

The generation of fluorescent X-rays is inherently a random process, so any analysis that is performed on the basis of received fluorescent X-ray photons is generally more reliable the more such photons can be collected. A known way to increase the statistical reliability of X-ray fluorescence analysis is to increase the time during which the sample remains illuminated by the incident radiation. If the sample is immobile, this means simply waiting a longer time before changing the sample. However, due to the nature of the industrial process, it may be necessary for the sample to arrive as a constantly moving stream. Even then, the concept of a longer measurement time is somewhat present, since if the composition of the sample stream remains essentially constant, then the accumulation of the number of detected fluorescent X-ray photons from the moving sample stream over X minutes is essentially the same as while holding a portion of the sample material immobile during analysis for X minutes.

However, there are limitations on the longest averaging time when analyzing a constantly moving sample stream because the composition of the sample stream does change and these changes can be important and therefore should be noticed. In addition, if the sample is a slurry, there are other factors that make the situation more difficult, such as the requirement that the slurry flow remain turbulent to prevent solid-liquid separation. It is not uncommon for a pulp sample to flow through the pulp handling unit at a rate of about 20 l/min. The object of the invention is to obtain sufficiently good detection results using an averaging time on the order of minutes, for example 2 minutes or 3 to 5 minutes.

Improvements in X-ray fluorescence analysis due to factors such as measurement geometry, incident power, choice of diffractive crystal materials, choice of detector types, use of multiple detection channels, and improved utilization of detection results are discussed below.

Fig. 2 is a schematic cross-sectional view of some parts of an X-ray fluorescence analyzer. The X-ray fluorescence analyzer includes an X-ray tube for emitting incident X-rays 206 in the direction of the first optical axis 204. A window 203 for transmitting X-ray tube radiation is seen in FIG. 2. To handle a pulp sample 202, the X-ray fluorescence analyzer includes a pulp processing unit 201, which in this case includes a chamber 208 or sample cell equipped with inlet and outlet connections. The precise manner in which sample chamber 208 is formed and its inlet and outlet connections to provide turbulent flow of sample 202 within the chamber is not relevant to this specific description. For example, the principles set forth in the international patent application published as WO 2017/140938 can be followed. In any case, the pulp processing unit is configured to maintain a constant distance between the pulp sample 202 and the x-ray tube. The constant distance can be considered, for example, in the direction of the first optical axis 204.

Keeping the distance constant results in the measurement geometry not changing, at least not in terms of distance and angle of view, which have an important influence on how much of the incident X-ray 206 hits the sample 202. Thus, the device may include means for changing the distance, for example, by changing the distance at which the x-ray tube is installed. In other words, it is not necessary that the specified distance always remain the same. Simply for purposes of direct processing of the detection results, it is advantageous for the mechanical configuration of the X-ray fluorescence analyzer to be able to maintain the specified distance constant during the measurement, when necessary.

Fig. 3 shows that the pulp handling unit 201 in the wall of the sample chamber 208 includes a sample window 301 to allow X-rays to pass while retaining the pulp sample 202 in said sample chamber 208. The window 301 is a hole covered by a foil 302 made from a material that is as transparent as possible to x-rays, yet mechanically strong enough to withstand the pressure and mechanical wear caused by the flowing liquid slurry. Thus, the pulp processing unit is configured to maintain a flat surface of the pulp sample 202 on the side facing the x-ray tube. In the geometry shown in Fig. 2 and 3, the first optical axis 204 is perpendicular to said flat surface.

Also in FIG. 2 and 3, the front wall 303 of the analyzer housing and the opening in said front wall 303 are shown. Another foil 304 covers said opening in the front wall 303. Just like the foil 302 of the window 301 in the chamber 208, said other foil 304 is made of a material which is transparent to x-rays. The purpose of the other foil 304 is to protect the interior of the X-ray fluorescence analyzer device from dust, moisture, and other contaminants that may be present in large quantities in the industrial process environment.

In FIG. 2 shows how incident x-rays 206 striking sample 202 produce fluorescent x-rays 207. These are initially omnidirectional, but what is of interest is the fluorescent x-rays 207 that exit chamber 208 through sample window 301 and can be collected into one or more detection channels. The location, geometry, and properties of such detection channels are described in more detail below.

Another feature shown in Fig. 2 and 3 is a filter plate 205 installed in the optical path between the X-ray tube and the pulp processing unit. A filter plate of this type is an additional feature. It works as a high-pass filter, attenuating in particular the lowest energy portion of the X-rays that were originally generated in the X-ray tube. The material and thickness of the filter plate 205, if used, can be chosen such that the plate transmits that X-ray that has sufficient energy to generate fluorescence in the element(s) of interest in the sample 202. The use of a filter is particularly useful. high pass instead of, for example, the main diffractor, which is a band pass filter, since the high pass filter will pass a wide spectrum of higher energy incident x-rays, which is then available to simultaneously generate fluorescent x-rays in a number of elements of interest.

If a filter plate 205 is used, it is preferable to place it closer to the x-ray tube than to the pulp processing unit. The filter plate 205 may even be attached to the x-ray tube so that it is very close to the window 203 for transmitting the radiation of the x-ray tube. If the filter plate 205 has additional dimensions in the transverse direction such that it is only slightly larger than or even no larger than the radiation window 203, then it can be ensured that the filter plate 205 will not unnecessarily obscure any field of view that otherwise it would be available for detection channels. The thickness of the filter plate 205 may be on the order of a millimeter or even less so that the use of the filter plate does not increase the overall distance between the X-ray tube and the sample to any significant extent.

Another feature shown in Fig. 2 and 3 is a calibration plate 305 that can be controlled and selectively placed in a position in which it is in the path of the incident X-rays 206 and within the field of view of the detection channels that are used to receive the fluorescent X-rays 207. The calibration plate 305 has a very well known composition, so it can be used to periodically calibrate detection channels. If the calibration process is to be automated, then the X-ray fluorescence analyzer may be equipped with an actuator that is configured to move the calibration plate 305 in a controlled manner between at least two positions, one of which is the position shown in FIG. 3 and the other at a position that is not in the path of the incident x-rays 206.

Fig. 4 illustrates an example of an X-ray fluorescence analyzer made in accordance with one embodiment of the invention. It contains a housing 401 which functions as the main supporting and protective structure. The front wall 303 of the analyzer housing is visible on the left in FIG. 4. As explained above with reference to FIG. 2 and 3, an opening is provided in the front wall 303 for the passage of the incident X-ray radiation generated by the X-ray tube 402. A holder 403 is provided to hold the pulp processing unit 201 on the outside of the front wall 303 aligned with said opening in the front wall 303.

In a preferred embodiment, the holder 403 may be configured to hold the pulp processing unit 201 on the front wall in a removable manner. The holder 403 may include, for example, hinges that allow the pulp handling unit 201 to be rotated sideways, or a bayonet mount that allows the pulp handling unit 201 to be quickly detached from the front wall 303 so that the window foil described above with reference to FIG. 2 and 3 turns out to be open. This makes it relatively easy to inspect and maintain those parts that are critical to the propagation of both incident and fluorescent X-rays. Particulate matter in the slurry can cause significant wear on the interior of the foil 302 of the sample window 301 (see FIG. 3), so it is advisable to equip the window 301 with a mechanism to allow the foil 302 to be replaced when needed.

The part of the X-ray fluorescence analyzer marked in FIG. 4 with a dashed box 404 is shown from the optical axis direction of the x-ray tube 402 shown in FIG. 5. This illustration shows an example of how an actuator 501 can be configured to move the calibration plate 305 in a controlled manner between these two positions. In the first position shown in Fig. 5, the calibration plate 305 is not in the path of the incident x-rays that exit the x-ray tube window 203. In the second position, the gauge plate 305 will be substantially concentric with the window 203 shown in FIG. 5.

In FIG. 4 and 5 also show how one or more detection channels 502 can be performed. The design and operation of the detection channel will be described in more detail later in this description. Fig. 4 and 5 illustrate the positioning principle in which each of the detection channels is positioned at a respective angle of rotation about the optical axis of the x-ray tube 402. pulp), this way of placing the detection channels allows you to organize virtually the same field of view for all detection channels.

Other features shown in FIG. 4 are blocks 405, 406 and 407 with electronics inside the analyzer housing 401 for each of the detection channels and X-ray tube 402, as well as circulating cooling water 408 for X-ray tube 402.

In FIG. 6 schematically shows some parts of what is referred to above as a detection channel. The main characteristics of the detection channel in Fig. 6 are a crystal diffractor 601 and a radiation detector 602. As the name suggests, the crystal diffractor 601 contains a crystal 603, which may be referred to as a diffractive crystal, or simply a crystal for short. Crystal 603 is a wavelength dispersive diffractor element 601. The first optical path 605 between the pulp processing unit (not shown in FIG. 6) and the crystal 603 may have a first slit 604, and the second optical path 607 between the crystal 603 and the radiation detector 602 may have a second slit 606. Since the diffraction properties crystal 603 for X-rays are highly dependent on the wavelength of the X-rays, such an arrangement can be used to extract a certain range of wavelengths from that part of the fluorescent X-rays that was originally emitted in the direction in which this particular crystal diffractor is located. Reference numeral 608 designates a shroud that encloses the crystal diffractor 601, providing structural support for all of its components.

In FIG. 7 shows an example of a crystal diffractor in axonometric projection. It is assumed that the crystal diffractor is located in the first direction from the pulp processing unit (not shown in Fig. 7), so that the first optical path 605 represents the nominal direction of that part of the fluorescent x-rays that enters this crystal diffractor. The first 604 and second 606 slits are formed between the respective parts of the limiter, and the second optical path 607 represents the nominal direction of the diffracted fluorescent x-rays, which is directed to the radiation detector (not shown in Fig. 7). The crystal diffractor is enclosed in a housing 608 defined by a first flat surface 701 and a second flat surface 702 which is parallel to the first flat surface 701.

The mechanical design described here is advantageous because the flat surfaces 701 and 702 provide a support to which the internal parts of the crystal diffractor can be attached with relative ease.

X-ray diffraction in a crystal is governed by Bragg's law, which relates the diffraction angle to the distance between the reticular planes. Conventional crystalline diffractors use crystals such as silicon dioxide, lithium fluoride, ammonium dihydrogen phosphate, or potassium hydrogen phthalate, since sufficiently large single crystal pieces of these materials can be relatively easily produced with the required accuracy in the desired shapes. However, it has been found that although the wavelength selectivity of such conventional crystals is relatively high, the efficiency with which the incoming X-rays are diffracted is relatively low.

Pyrolytic graphite is an alternative material for making a crystal for a crystal diffractor. Pyrolytic graphite is a general term that refers to materials made from organic compounds containing planar structures such as benzene rings by exposing them to high temperatures, leaving essentially only carbon atoms in the structure. The original planar molecular structures cause pyrolytic graphite to exhibit a highly colored microscopic structure, for which reason it is often referred to as HOPG (Highly Oriented Pyrolytic Graphite) or HAPG, the latter referring to a slightly different way of synthesizing the material. Pyrolytic graphite is often not single-crystal in the same sense as the more traditional crystalline materials mentioned above, but is polycrystalline. However, for consistency with common terminology in the art, in practice, a diffractive element made of pyrolytic graphite is referred to as a crystal. In the following description, the term HOPG crystal is used.

It has been found that the efficiency of the HOPG crystal as a fluorescent X-ray diffractor is significantly higher than conventional diffractive crystal materials. In other words, much more of the X-rays that hit the HOPG crystal are actually diffracted than with conventional crystalline materials. However, the polycrystalline nature of pyrolytic graphite means that not all of the reticular planes are oriented in exactly the same way as, for example, in single-crystal silicon dioxide. This, in turn, means that the wavelength selectivity of a HOPG crystal in a crystal diffractor is not very accurate: fluorescent X-rays that are diffracted in a certain direction are a range of wavelengths near the nominal wavelength, which, according to Bragg's law, will be diffracted in this direction, and this range is much wider than in X-rays diffracted on conventional crystalline materials.

However, the less accurate wavelength selectivity of the HOPG crystal is not a serious disadvantage, since it can be compensated for by other factors in the development of an X-ray fluorescence analyzer. One possible approach is to use a solid state semiconductor detector as the radiation detector 602, to which fluorescent X-rays in a selected wavelength range are directed from the HOPG crystal. The radiation detector 602 may be, for example, a PIN diode detector, a silicon drift detector, a germanium detector, or a germanium drift detector. Unlike, for example, gas-filled

- 7 040045 proportional counters, the energy resolution of solid state semiconductor detectors can be made more accurate. The energy resolution of an X-ray detector is usually expressed at a reference energy of 5.9 keV. A solid state semiconductor detector of the type mentioned above can have an energy resolution of better than 300 eV with a specified reference energy of 5.9 keV.

By combining the use of a HOPG crystal in the crystal diffractor 601 with the use of a solid state semiconductor detector as the radiation detector 602, an advantageous situation can be achieved in which the crystal diffractor 601 is configured to extract a predetermined first wavelength range from the fluorescent X-ray radiation 207 propagating in the direction in which a crystal diffractor 601 is disposed (with reference to the pulp processing unit 201), and is configured to direct fluorescent X-rays in a selected predetermined first wavelength range to a radiation detector 602, that is, a solid-state semiconductor detector. The good energy resolution of the solid state semiconductor detector is then used to obtain a measurement result that gives the energy spectrum of the fluorescent x-rays in the selected predetermined first wavelength range. Based on the indicated energy spectrum, and possibly using other measurements as confirmation, the amount of fluorescent x-rays from the element of interest can be determined with relatively good accuracy.

The geometric shape of the diffractive surface of the HOPG crystal is another factor to consider when designing an X-ray fluorescence analyzer. Fig. 8 illustrates some examples of geometric shapes. It can be noted here that, more precisely, a crystal is just a thin layer of crystalline (single-crystalline, as, for example, in the case of silicon dioxide, or polycrystalline, as, for example, in the case of HOPG) material, which forms a real diffractive surface. The crystalline diffractor contains a substrate to which the crystalline material is attached. Examples of substrate materials are, for example, glass and aluminum, but the substrate material can also be iron or any other similar material that is not prone to cause unwanted interfering fluorescent radiation by itself. The crystalline material can be attached to a properly formed surface of the substrate, for example, using van der Waals forces. Alternatively, the crystalline material may be grown directly on a properly formed substrate surface, or some other suitable attachment method such as adhesive may be used.

Together, the substrate and the crystalline material constitute a three-dimensional object, with examples of these objects shown in FIG. 8. To maintain consistency with conventional terminology in this technical field, in the present description, these objects are referred to as crystals, despite the small accuracy of this term, as explained above. The term diffractive surface refers to the outer exposed surface of a crystalline material on which X-rays are diffracted; strictly speaking, X-ray diffraction occurs in reticular planes within the crystalline material near the surface, which is referred to herein as the diffractive surface.

A feature common to die 603, die 802 and 804 shown in FIG. 8 lies in the fact that the three-dimensional geometric shape of an object consisting of a HOPG crystal and a substrate is a prism, one side face of which is cut off to form a curved diffractive surface. The imaginary shape of the prism in the top row of images of these three crystals is shown by dotted lines.

The bottom row of images of the same crystals in Fig. 8 shows how the nature of the curvilinearity of the diffractive surface differs in all three cases. In the crystal 603, the diffractive surface 801 is curved in only one direction (longitudinal direction). In other words, if an imaginary transverse line like the dotted line shown in FIG. 8, pass through the diffractive surface 801 anywhere, it will always be straight. A particular advantage of such a crystal is that it is relatively easy to manufacture. Compared to FIG. 6 and 7 that the radius of curvature of the diffractive surface 801 lies in a plane defined by the first 605 and second 606 optical paths. This plane is also parallel to flat surfaces 701 and 702.

In the crystal 802, the diffractive surface 803 is curved in two directions (longitudinal and transverse), forming part of the toroidal surface. This means that if a transverse arc is drawn anywhere through the diffractive surface 803, like the two dotted arcs shown, for example, in FIG. 8, then all these transverse arcs will be identical to one another. While this geometry may be somewhat more difficult to manufacture to the required accuracy than the surface 801 shown at left, it has the advantage of focusing diffracted x-rays more accurately.

In the crystal 804, the diffractive surface 805 is curved in two directions (longitudinal and transverse), but differently from the surface 803 shown in the middle. The diffractive surface 805 forms part of a rotationally symmetrical surface whose axis of rotation 806 is

- 8 040045 is in the plane defined by the optical paths of the incoming and diffracted X-rays. This means that if a transverse arc is drawn through the diffractive surface 805, such as the dotted arc shown in FIG. 8, then the radius of curvature of such a transverse arc will be different depending on the longitudinal position in which it is drawn. In FIG. 8, it can be seen that the dotted arc in the middle is not as noticeably curvilinear as the arcuate edges seen at the ends of the die 804. This is because the dotted arc is farther away from the axis of rotation 806 than the arcuate edges at the ends of the crystal.

From a mathematical point of view, a rotationally symmetric surface is formed when a continuous curve rotates around an axis of rotation. The shape of the specified continuous curve determines how far from the axis of rotation each surface point is and what properties it can have. One example of a curve that can be used to form diffractive surface 805 in FIG. 8 is a portion of a logarithmic spiral. Although this type of surface is more difficult to fabricate than the surfaces shown above as surfaces 801 and 803, an axisymmetric surface made from a logarithmic helix section has the inherent advantage of providing very precise focusing of diffracted x-rays.

A feature that is common to all diffractive surfaces 801, 803, and 805 depicted in FIG. 8 is that in the topological sense all these surfaces are simply connected surfaces. A simply connected surface is a linearly connected surface (i.e. any two points on the surface can be connected by a trajectory that belongs entirely to the specified surface) and, in addition, any loop-shaped contour can be continuously compressed to a point, so that also all intermediate the shapes of the compressed loop completely belong to the specified surface.

An intuitive description of a simply connected surface is that it has no holes. As such, through any of the diffractive surfaces 801, 803, or 805 shown in FIG. 8, it is possible to drill a small hole without changing their properties as diffractors, no more than the amount by which the surface area is reduced by the drilled hole. For this reason, this specification defines that the requirement of a simply connected surface in a topological sense should be interpreted as referring to the general shape of the surface: in this interpretation, a small hole in the surface does not mean that it will not be simply connected. Another definition of how to interpret the requirement of simply connectedness is as follows: if the crystal is lying on its side, as shown in Fig. 8 (i.e., the principal radius of curvature, which defines the longitudinal curvature between the ends of the crystal, is in the horizontal plane, so that the diffractive surface is normally oriented vertically), any imaginary horizontal line will intersect the diffractive surface at at most one point. A surface is simply connected if it matches at least one of these intuitive descriptions.

On the right in FIG. 8, crystal 807 is shown as a comparative example. The diffractive surface 808 of the crystal 807 is curved in two directions (longitudinal and transverse), forming a complete axisymmetric surface, the rotation axis 809 of which may be in the plane defined by the optical axes of the incoming and diffracted x-rays. The curve whose rotation about the rotation axis 809 defines the shape of the diffractive surface 808 may, for example, be a section of a logarithmic spiral. Obviously, the diffractive surface 808 is not simply connected in a topological sense, since no closed curve that goes around the surface opening can be compressed to a point. Crystals of this type are relatively difficult to manufacture, but they can be used in conjunction with suitable shields (not shown in Fig. 8) that block the propagation of direct, undiffracted X-rays to collect fluorescent radiation from a larger spatial angle than those that have a single bond surface, such as 801, 803 or 805.

The geometric shape and resulting optical properties of the diffractive surface can influence how other parts of the crystal diffractor should be made. It has been explained above that the crystal diffractor 601 may have a first slit 604 in the first optical path 605 between the pulp processing unit 201 and the (pyrolytic graphite) crystal, and that there is a second optical path 607 between the (pyrolytic graphite) crystal and the radiation detector 602. If the diffractive surface 801 of said crystal 603 (pyrolytic graphite) is curved in only one direction with a radius of curvature in a plane defined by said first 605 and second 607 optical paths, then it is preferable to make said first slit 604 a linear slit oriented perpendicular to said plane, as shown in fig. 7. If the diffractive surface 803 of said crystal 802 (pyrolytic graphite) is curved in two directions, forming part of a toroidal surface, it is preferable to make said first slit a curved slit with a radius of curvature oriented perpendicular to said first optical path. If the diffractive surface 805 of said crystal 804 (pyrolytic graphite) is curved in two directions, forming part of a rotationally symmetric surface whose axis of rotation 806 is in the plane defined by said first and second optical paths, then it is preferable to make said first slit in the form of a pinhole.

If the second slot 606 is used in the second optical path 607, then similar considerations can be applied. However, it should be noted that the second slit is not always necessary: its use is associated with the attenuation of the background and scattered radiation, especially with diffractive crystals, which have a high wavelength selectivity. Considering that the wavelength selectivity of HOPG is not so high, the additional benefit obtained with the second slit is relatively small.

If a second slit is used in the second optical path 607 between the (pyrolytic graphite) crystal 603, 802, 804 and the first radiation detector, then the geometry of the crystal diffractor may follow, for example, the principle of Johann's geometry or Johanson's geometry. This is shown, respectively, in FIG. 9 and 10. In FIG. 9, the central point 902 of said diffractive surface, the first slit 604 and the second slit 606 are located on a Rowland circle whose radius is R. The radius of curvature of said diffractive surface in the plane defined by said first and second optical paths is 2R, and the radius of curvature of the reticular planes 901 in this crystal is equal to 2R. This means that the first crystal diffractor has a Johann geometry. In FIG. 10, the center point 1002 of said diffractive surface, the first slit 604, and the second slit 606 are likewise located on a Rowland circle whose radius is R. However, here, the radius of curvature of said diffractive surface in the plane defined by said first and second optical paths is R, and the radius the curvature of the reticular planes 1001 in said crystal is 2R, so that the first crystal diffractor has Johanson geometry.

In order to maintain the compact size of the crystal diffractor, it is preferable that the value of R remain relatively small. For example, R can have a value of no more than 40 cm.

Fig. 11-14 are schematic illustrations of fluorescent x-ray spectra in some cases. Spectra are usually expressed as the number of counts recorded at each photon energy. In practice, the sampling detector has a certain energy resolution, which determines how close the energies of two photons can be to each other so that the detector can generate two different types of output signals. Signal processing is used to classify the received X-ray photons into energy cells of finite width, and the number of counts is output per energy cell. The more accurate the resolution of the detector, the narrower (in energy units) the energy intervals can be made.

In FIG. 11 plot 1101 is smooth without any visible peaks or spectral holes. Such a spectrum is rarely obtained in practice, but it illustrates the situation when only background and randomly scattered radiation is received, without any characteristic peaks of the elements of interest. In FIG. 12, plot 1201 is otherwise the same, but there is a characteristic peak 1202 of the element of interest. The problem is that the concentration of the element of interest in the measured sample is so low that the height of the characteristic peak 1202 is small compared to the overall level of the spectrum in the same energy range. Thus, even if a relatively large number of photons are observed in this energy range, only a relatively small number of them are actually fluorescent photons from the element of interest.

The energy of a photon is inversely proportional to its wavelength, so when the wavelength selectivity of various diffractive crystals was considered above, the energy selectivity could also be taken into account. In FIG. 13 schematically shows what a crystal diffractor radiation detector equipped with a HOPG crystal can receive. The energy range 1301 of the fluorescent X-rays that the HOPG crystal will direct to said radiation detector is relatively wide, which is a direct result of the relatively moderate wavelength selectivity of the HOPG crystal. However, at the same time, the diffraction efficiency of the HOPG crystal is relatively good. Thus, the radiation detector will receive a significant fraction of the photons that fall in the two shaded areas in FIG. 13. Of these, the photons belonging to the first shaded region 1302 are background and scattered photons, while the photons belonging to the second shaded region 1303 are the actual fluorescent photons from the element of interest.

In FIG. 14 schematically shows what a crystal diffractor radiation detector equipped with a silicon dioxide crystal (or other conventional crystal) could obtain in the same situation. The energy range 1401 of the fluorescent X-rays that a conventional crystal directs to its radiation detector is relatively narrow, which is a direct result of the conventional crystal's relatively good wavelength selectivity. However, at the same time, the diffraction efficiency of a conventional crystal is lower than that of a HOPG crystal. Thus, the radiation detector will only receive a limited portion of the photons that actually come from the element of interest in the sample (see shaded area 1303 in FIG. 13). Small peak 1402 in FIG. 14 denotes this fluorescent X-ray, which in this case will actually be detected.

- 10 040045

One of the factors to consider when developing an X-ray fluorescence analyzer is the possibility of using differently equipped detection channels.

Here, the term differently equipped means, first of all, the choice of a diffractive crystal and the choice of a radiation detector.

In FIG. 15 schematically shows how a commercial X-ray fluorescence analyzer for analyzing pulp samples can contain a large number of detection channels. The detection channels are shown in Fig. 15 in a straight line because the image is schematic. In practice, they can be arranged, for example, axisymmetrically around the X-ray tube, as shown in FIG. 4 and 5, each with a field of view directed towards the pulp processing unit of the X-ray fluorescence analyzer.

The X-ray fluorescence analyzer comprises a first crystal diffractor 1501 located in a first direction from said pulp processing unit, the first crystal diffractor 1501 comprising a first crystal. The first radiation detector 1505 is configured to detect fluorescent X-rays diffracted by the first crystal 1502 with a first energy resolution. The X-ray fluorescence analyzer comprises a second crystal diffractor 1511 located in the second direction from said pulp processing unit, the second crystal diffractor comprising a second crystal 1512. The second radiation detector 1515 is configured to detect fluorescent X-rays diffracted by the second crystal 1512 with a second energy resolution.

As a first assumption, the first crystal 1502 is a pyrolytic graphite (HOPG) crystal and the second crystal 1512 is made of a material other than pyrolytic graphite, such as silicon dioxide, lithium fluoride, ammonium dihydrogen phosphate, or potassium hydrogen phthalate. Also, as a first assumption, it can be assumed that the first and second crystal diffractors are configured to direct the characteristic fluorescent radiation of the same element to their respective radiation detectors. In other words, the two detection channels are equipped differently, but both of them are aimed at detecting the presence of the same element in the pulp sample and determining its concentration.

Thus, setting up a crystal diffractor to direct the characteristic fluorescent radiation of a particular element to its radiation detector is usually done by: 1) choosing a crystal with a certain distance between its reticular planes, 2) choosing the curvature of the crystal and reticular planes, and 3) choosing angle and distance values crystal and slit(s) such that only a specific wavelength range of X-rays reaches the detector, said specific wavelength range including the required characteristic peak of the element of interest. An element of interest may have multiple characteristic peaks, so stating that two detection channels are capable of measuring the characteristic fluorescent emission of the same element does not necessarily mean that they will be set to detect the same characteristic peak, although this is not the case. excluded.

If these two detection channels are configured to measure the same characteristic peak, then the measurement results they produce may resemble those shown in FIG. 13 (for the HOPG channel) and in FIG. 14 (for a channel with a different crystal). The task of determining the actual concentration of an element of interest can be described as a method, for example, as follows.

The method is aimed at performing x-ray fluorescence analysis and includes irradiating the pulp sample with incident x-rays and obtaining fluorescent x-rays from the irradiated sample. Due to the geometry of the measurement, the first part of the fluorescent x-ray will be directed to the first detection channel, and the second part of the fluorescent x-ray will be directed to the second detection channel. The method includes separating the first 1301 and second 1401 specified wavelength ranges from the respective first and second parts of the specified received fluorescent x-ray radiation using the respective first 1501 and second 1511 crystal diffractors. Both the first wavelength range 1301 and the second wavelength range 1401 include the characteristic fluorescent emission of the same element. In addition, the first wavelength range 1301 is at least twice as wide as the second wavelength range 1401.

The method includes detecting fluorescent x-ray radiation in the specified first and second selected wavelength ranges using the respective first 1505 and second 1515 radiation detectors. The energy resolution of said first radiation detector 1505 is better than 300 eV with a reference energy of 5.9 keV. Thus, the method includes obtaining the respective results of the first and second detections. The method includes calculating the concentration of said element in said sample from at least one of said first and second detection results.

As used herein, the term at least one emphasizes the fact that not all detection results are best treated in the same way. A lot depends on the test. In some samples, the concentration of the element of interest may be relatively large, resulting in a relatively large number of detected fluorescent photons even in the second radiation detector 1515, despite the moderate diffraction efficiency of the second crystal 1512. In some other cases, the concentration of the element of interest may be so low that that in the second wavelength range 1401 only a very small and indistinct peak is visible. In some cases, the first wavelength band 1301 may appear relatively clear of any interfering radiation, while some other samples may contain significant amounts of some other element whose characteristic peak is so close that it becomes visible and even dominates the first band. 1301 wavelengths, but not in the second range of 1401 wavelengths.

In general, said calculation may include calculating the total intensity of background radiation and fluorescent x-ray radiation from all but the indicated element using at least one of the first and second detection results. The method may then include subtracting from the total intensity detected in the wavelength range containing the specified characteristic peak of the fluorescent X-ray emission of the element to be measured in the specified sample, the calculated total intensity of the background radiation and fluorescent X-ray emission from other elements, excluding the specified element of interest in the specified sample. The method may then include presenting the result of said subtraction as a calculated intensity of said characteristic x-ray fluorescent peak.

Said calculation may include analyzing, from said first and second detection results, whether the influence of the characteristic peak from another element on the first detection result exceeds a predetermined threshold. If said analysis shows that the effect of said characteristic peak of said other element on the first detection result is greater than said predetermined threshold, then the method may include calculating said concentration of said element in said sample based on said second detection result. If, on the other hand, said analysis shows that the effect of said characteristic peak of said other element on the first detection result does not exceed said predetermined threshold, then the method may include calculating said concentration of said element in said sample from said first detection result.

Another possibility is to generate specific models for each measuring channel in each sampling line using calibration samples. The measurement channel is then selected to be used for the actual measurements of that sampling line, based on which one gives the most accurate calibration.

The element of interest may be gold because gold is valuable and because there are reasonably efficient ways to extract it even from sludge streams where it appears in very low concentrations. There are other elements whose interference characteristic peaks may or may not be present and may appear very close to that of gold. If a significant amount of these interfering elements is present in the sample, then the detection channel with a HOPG crystal can give relatively inaccurate and unreliable results, at least when used alone.

It is possible to imagine intermediate forms of these two extreme cases, in which the contribution of the first and second detection results is taken into account in different ways. The decision on which calculation method to choose can be made, for example, by using an artificial intelligence algorithm that compares the first and second detection results with previously obtained comparable results and with some evaluation data on how the various available calculation methods performed with the indicated comparable results. .

Fig. 16 schematically illustrates a fluorescent X-ray spectrum that contains two distinct peaks 1601 and 1602. In such a case, the calculation method chosen may depend on whether both peaks 1601 and 1602 are characteristic peaks of the same element of interest, or whether one of them is the characteristic peak of some interfering element. Smaller peaks closer to the energy axis represent the expected detection result that a detection channel equipped with a conventional crystal (eg, silicon dioxide) could obtain from these two peaks.

An interesting case is that both peaks 1601 and 1602 are peaks of the element of interest. It is of particular interest if one of them (in this document: peak 1601) is more intense, for the measurement of which a detection channel equipped with SiO ₂ is provided. In such a case, the best characteristics of both channels can be used: the precise wavelength selectivity of the silica crystal can be used to separate a well-defined wavelength range 1401 that includes only the required characteristic peak, so that the relatively high intensity of this peak still gives a sufficient amount of counts in the corresponding detector in a relatively short time. At the same time, the good diffraction efficiency of the HOPG crystal can be used to highlight a wider wavelength range 1301 that includes a different, lower characteristic peak. The concentration of the element of interest can be calculated from the detection results provided by the two detectors when the overall characteristics of the two detection channels are known from calibration measurements.

A method of the type described above may be applicable in many cases where the characteristic fluorescent emission includes the K or L peak of an element with 30 < Z < 92, where Z is the atomic number of said element. The adaptability of the method is well suited for measuring samples that contain one or more elements of interest in a matrix consisting primarily of elements with Z < 8, where Z is the atomic number. These are, for example, water-based pulps.

The principles discussed above regarding the use of two detection channels can be generalized to apply to the use of three or more detection channels. The form factor of the detection channel as described above, i.e., one in which each crystal diffractor 601 is enclosed by a first flat surface 701 and a second flat surface 702 that is parallel to said first flat surface 701, allows multiple detection channels to be distributed in the form cassettes, for example, in an axisymmetric manner around the x-ray tube. Detection results from detection channels configured to measure the characteristic fluorescent emission of the same element may be combined in various ways as described above. A large number of detection channels makes it possible to simultaneously calculate the concentrations of two or more elements of interest in a sample, if the detection channels are configured to detect the characteristic fluorescent emission of such two or more elements of interest. It is also possible to cross-correlate the detection results of channels configured to detect different elements. For example, if one element has two characteristic peaks, one of which is measured using a dedicated first detection channel and the other approaches the characteristic peak of another element of interest, the detection results from the first channel can be used to correct the detection results from the channel that is performed with the ability to measure the characteristic peak of another element.

One factor to consider when designing an industrial X-ray fluorescence analyzer for pulp sample analysis is the power of the X-ray tube and the geometry and dimensions of the area between the X-ray tube and the pulp processing unit.

Fig. 17 illustrates the possibility of using the so-called transmission geometry. A window 203 for transmitting X-ray tube radiation is seen on the right in FIG. 17, the incident x-rays are emitted in the direction of the optical axis 204 through the main filter plate 205. The pulp processing unit 201 includes a chamber 1701 with an exit slot 1702 from which the sample 202 flows out as a curtain and falls under the action of gravity. Incident X-rays generate fluorescent X-rays in a relatively thin curtain of falling pulp. Reference number 1703 indicates fluorescent x-rays that are directed obliquely backward and that can be detected with detection channels (not shown in FIG. 17) arranged in the same manner as in the geometries described above with reference to FIG. 2, 3, 4, and 5. Item number 1704 indicates fluorescent x-rays that are directed in other directions, especially in directions that are on the other side of the sample flow. They can be detected using detection channels (not shown in Fig. 17) located on that side. This can be a particularly advantageous way of arranging the detection channels, since they can have a better field of view and hence a better spatial collection efficiency of the fluorescent x-rays. It can also cause the X-ray tube to get very close to the sample. However, it should be noted that proper geometric radiation protection measures must be taken to prevent any incident X-rays from entering the detection channels.

Fig. 18 is a partial cross-sectional view of the exit portion of an x-ray tube 402. The x-ray tube includes an anode 1801 for generating incident x-rays. Incident x-rays are emitted in the direction of the optical axis 204 to the sample 202, which is shown here only schematically without the pulp processing unit for clarity. However, it is assumed that the pulp handling unit is configured to maintain a flat surface 1802 of the pulp sample 202 on the side facing the x-ray tube 402. As explained above, this can be done, for example, by providing a sample window with a foil of a material that is transparent for x-rays. A sample window may be provided in the wall of the sample chamber so that X-rays can pass through while the pulp sample is retained within the sample chamber.

The remaining parts of the X-ray tube, shown schematically in FIG. 18 are a cooling water circuit 1803, an annular cathode 1804 for emitting accelerated electrons, and a window

- 13 040045

203 for radiation transmission.

When the goal is to produce such a large fluorescent emission that even very small concentrations of elements of interest can be detected, it is preferable that as many photons (of sufficient energy) of the incident radiation as possible can enter the sample 202. One way to achieve this is using a very powerful x-ray tube. In accordance with one implementation, the nominal input power of the x-ray tube 402 is at least 400 watts. Even more powerful x-ray tubes can be used: according to other embodiments, the nominal input power of x-ray tube 402 can be at least 1 kW, or at least 2 kW, or even at least 4 kW. Even if only a fraction of the power claimed as the rated input power of the x-ray tube ends up as random x-rays generated, the rated input power is nevertheless an important indicator of the x-ray tube's ability to produce an intense incident x-ray beam.

The use of x-ray tubes with higher power than before means that radiation shielding needs to be re-evaluated compared to known lower power x-ray sources. According to one embodiment, thicker radiation shielding plates and denser radiation shielding materials can be used to ensure that ionizing radiation does not penetrate areas where it could be hazardous.

Another way to ensure that a very intense incident X-ray beam hits sample 202 is to make the distance between anode 1801 and sample 202 as small as possible. The pulp processing unit may be configured to maintain the shortest linear distance d between the anode 1801 and the sample 202 such that d is less than 50 mm. In yet another embodiment, d may be less than 40 mm. In yet another embodiment, d may be less than 30 mm.

However, it should be noted that, as a general rule, the closer the anode 1801 of the x-ray tube 402 approaches the sample 202, the greater the spatial angle around the sample 202 is blocked by the x-ray tube structures. This is an important factor to consider because X-ray tube 402 designs can block the field of view of the detection channels. One way to solve this problem is to use an X-ray tube with a so-called end window rather than an X-ray tube with a side window. Fig. 18 and 19 can be seen as illustrating the use of an end window X-ray tube. In this type of X-ray tube, the radiation window 203 is usually located at one end of the overall tubular structure, which leaves a relatively large amount of free space around said tubular structure for accommodating detection channels. Another possibility would be to use an X-ray tube with a side window and place the detection channels on one or both sides of the X-ray tube.

In all of the drawings described above, the optical path between X-ray tube 402 and sample 202 is also straight, which means that there are no diffractors between them. This is another way to ensure that the maximum number of incident X-ray photons can enter the sample. First, the presence of a diffractor between them would inevitably mean a greater distance between the anode 1801 and the sample 202, because some space must be reserved for the diffractor. Secondly, the very nature of the diffractor is to extract only a certain range of wavelengths from the original radiation spectrum, which necessarily means fewer incident X-ray photons hitting the sample. Other beneficial consequences of not using any so-called main diffractor between the x-ray tube 402 and the sample 202 are that incident x-rays are simultaneously delivered to excite characteristic peaks of a number of elements in the sample, and that fewer features can block the field of view of the detection channels.

In FIG. 18, the optical axis 204 of the X-ray tube 402 is perpendicular to the flat surface 1802 of the sample 202. While this arrangement provides excellent rotational symmetry for the detection channels located around the X-ray tube 402, this is not the only possibility. In FIG. 19 shows an alternative embodiment in which the optical axis 204 of the x-ray tube 402 is at an oblique angle to said flat surface. This arrangement can help to make the shortest linear distance d between anode 1801 and sample 202 even shorter, while leaving a sufficiently free field of view for detection channels on at least some sides of X-ray tube 402. This principle is discussed in more detail below with reference to FIG. 20 and 21.

In FIG. 20 shows the x-ray tube 402 and five detection channels as viewed from the sample side. The radiation window 203 of the x-ray tube 402 is visible in the middle of the drawing. The entrance window of each detection channel for receiving fluorescent radiation is located at the proximal end of the corresponding crystal diffractor; input window 2001 is shown as an example. In order for as much of the generated fluorescent radiation as possible to enter the channel

- 14 040045 detection, it is advisable to place these input windows as close as possible to the sample, and also so that the input window is visible on the sample surface at the largest spatial angle possible. Each of said plurality of crystal diffractors is positioned at a respective angle of rotation about the optical axis of the x-ray tube 402. Each of said crystal diffractors is configured to extract a predetermined range of wavelengths from fluorescent x-rays that propagate in a respective direction, and is configured to direct the fluorescent x-rays in the corresponding selected predetermined first wavelength range to the corresponding radiation detector.

In FIG. 21 is a side view showing an X-ray tube 402 and two detection channels. An exemplary window 301 is shown schematically in FIG. 21: this illustrates the area where the pulp handler can maintain a flat surface of the pulp sample on the side facing the X-ray tube 402. Thus, this area is the area that must be within the field of view of the X-ray tube 402 in order for the incident X-ray radiation to got on trial. This also illustrates the area that should cover as large a spatial angle as possible in the field of view of the detection channels in order to collect as much fluorescent X-rays as possible.

The optical axis 204 of the x-ray tube 402 is located at an oblique angle to the specified flat surface. The first crystal diffractor 1501 is positioned at that angle of rotation about said optical axis 204 at which said flat surface of said sample covers the largest portion of the field of view of the first crystal diffractor 1501. Assuming that no other designs block any portion of the available field of view, in practice this means that the first crystal diffractor 1501 is located opposite the X-ray tube, that is, in the direction in which an imaginary light beam would be reflected along the optical axis 204 if the sample surface were a mirror.

The second crystal diffractor 1511 is positioned at a different angle of rotation around said optical axis 204. In FIG. 21, the second crystal diffractor 1511 is positioned at what could be described as the worst possible angle of rotation because its view of the sample surface is limited to that edge of the x-ray tube 402 that is closest to the sample window 301. If said other angle of rotation is less than 180° different from the angle at which the first crystal diffractor 1501 is located, then the second crystal diffractor 1511 may be positioned rather as one of several other crystal diffractors in FIG. 20. In such a case, the flat surface of the sample in the sample window 301 will cover the part of the field of view of the second crystal diffractor 1511 that is between the two extreme points shown in FIG. 21.

According to one embodiment, the first crystal diffractor 1501, which is positioned at the optimum angle of rotation (in terms of field of view) in FIG. 20 and 21 is a crystal in which the diffraction crystal is a HOPG crystal and the radiation detector is a solid state semiconductor detector. Considering the good diffraction efficiency of the HOPG crystal, this placement of the first crystal diffractor helps to ensure that the maximum amount of X-ray fluorescent photons actually reaches the detector. If there is some prior knowledge of the expected levels of concentrations of various elements in the samples to be measured, then it may be advantageous to install this crystal diffractor with the most optimal angle of rotation, which is configured to separate and direct to the appropriate detector the characteristic fluorescence emission of that element of interest, the concentration of which is expected to be the smallest.

One of the factors to consider when designing a commercial X-ray fluorescence analyzer for the analysis of slurry samples is the choice of radiation detectors in those channels that have diffractive crystals from materials other than pyrolytic graphite. The wavelength selectivity of conventional diffractive crystalline materials such as silicon dioxide is relatively good, which can be interpreted to mean that there is not as much need for precise energy resolution in the radiation detector as if a HOPG crystal were used. A gas-filled proportional counter can provide quite satisfactory detection results in a detection channel that uses a crystal other than HOPG as the diffraction crystal at a much lower manufacturing cost than a solid state semiconductor detector.

However, nothing said above should be construed against the selection of a solid state semiconductor detector also for detection channels which have other than HOPG crystals as diffraction crystal. Similarly, the use of a solid state semiconductor detector in a detection channel equipped with a HOPG crystal is not a requirement if the energy resolution of another type of radiation detector is sufficient.

Fig. 22-25 illustrate calibration measurements, with the vertical axis in these drawings representing the concentrations measured with a single test device detection channel.

- 15 040045 which was an industrial X-ray fluorescence analyzer for the analysis of pulp samples, made in accordance with one embodiment. The horizontal axis represents the concentrations in the same samples but measured over long periods with laboratory equipment in order to obtain the most accurate and reliable results possible. For laboratory measurements, pulp samples were dried and homogenized, with the amount of water removed being compensated by calculation to make the laboratory measurements comparable to industrial type measurements. If this type of calibration measurement shows that the points are along a straight line, then two different devices give very similar results, which means that the device under test is very accurate. Deviations from a straight line indicate that the device under test is not giving accurate results.

The scales are arbitrary, but the scales in FIG. 22 and 23 are the same, and the scales in FIG. 24 and 25 are the same. In all dimensions, the element of interest was gold. Fig. 22 and 23 depict measurements of samples in which the concentration of the interfering element was below 300 ppm, while in the measurements presented in FIG. 24 and 25, the concentration varied from 0 to 2%. An interfering element is an element that has a characteristic fluorescence peak close to at least one characteristic fluorescence peak of the element of interest.

Fig. 22 and 24 are cases in which measurement with the device under test was performed using a detection channel that had a silicon dioxide crystal in a crystal diffractor and a gas-filled proportional counter as a radiation detector. Fig. 23 and 25 are cases in which measurement with the device under test was performed using a detection channel that had a HOPG crystal in a crystal diffractor and a solid state semiconductor detector as a radiation detector.

Comparison of FIG. 22 and 23 shows that when the interfering element concentration is low, the detection channel with HOPG crystal and solid state semiconductor detector gives more accurate detection results than the detection channel with silicon dioxide crystal and gas-filled proportional counter. The average error between concentrations measured with the DUT's HOPG channel and concentrations measured in the laboratory was +/- 0.24 ppm, while the comparable average error with a silica crystal and a gas-filled proportional counter was +/- 0.56 ppm.

Comparison of FIG. 24 and 25 shows that when the interfering element concentration is significant, the detection channel with HOPG crystal and solid state semiconductor detector gives less accurate detection results than the detection channel with silicon dioxide crystal and gas-filled proportional counter. The average error between the concentrations measured with the DUT's HOPG channel and the concentrations measured in the laboratory was +/- 1.62 ppm, while the comparable average error with a silica crystal and a gas-filled proportional counter was +/- 0.42 ppm.

The results shown in FIG. 22-25 can be used in a variety of ways. For example, an industrial X-ray fluorescence analyzer for analyzing pulp samples may be equipped with first, second and third detection channels, of which the first and second detection channels are equipped with crystal diffractors configured to separate and direct to their respective detectors the characteristic fluorescent X-rays of the same element such as gold. The first detection channel may be a channel with a HOPG crystal and a solid state semiconductor detector, and the second detection channel may be a channel with a silicon dioxide crystal and a gas-filled proportional counter. The third detection channel may be equipped with a crystal diffractor capable of separating and directing the characteristic fluorescent x-rays from the interfering element to a suitable detector. Then, the detection results of all three detection channels can be analyzed. If the detection results obtained using the third detection channel indicate that there is a significant concentration of the interfering element in the sample, then the calculation of the gold concentration will more emphasize (or even exclusively use) the detection results of the second detection channel. Accordingly, if the detection results obtained by the third detection channel indicate that there is only a negligible concentration of the interfering element in the sample, then the calculation of the gold concentration will more enhance (or even exclusively use) the detection results of the first detection channel.

The many advantages of an industrial X-ray fluorescence analyzer for the analysis of pulp samples have been described above. After all, they all serve a common goal, which is to make reliable measurements of even very low concentrations of the elements of interest in various types of slurries, at a reasonable cost and under the harsh conditions that can occur in an industrial environment: short measurement times, extreme temperatures, frequent occurrence of moisture, dust and dirt, long service intervals, and the like. Advantageous characteristics can be combined with each other in several ways, so that the most advantageous combination may depend on the particular case and its particular boundary conditions.

For a person skilled in the art it is obvious that with the development of technology the main idea of the invention can be implemented in various ways. Thus, the invention and its embodiments are not limited to the examples described above, instead they may vary within the scope of the claims. For example, even though gold has often been mentioned above as a typical element of interest, the same principles apply to measurements of other elements of interest. Examples of such other elements of interest are, for example, copper, silver, platinum group metals and uranium.

Claims (25)

CLAIM 1. X-ray fluorescent analyzer containing x-ray tube (402) for emitting incident x-rays (206) in the direction of the first optical axis (204); a pulp processing unit (201) configured to maintain a constant distance between the pulp sample (202) and the x-ray tube; the first crystal diffractor (601, 1501) located in the first direction from the block (201) pulp processing, and the first crystal diffractor (601, 1501) contains the first crystal (60z, 1502); a first radiation detector (602, 1505) configured to detect fluorescent x-rays diffracted by said first crystal (603, 1502) with a first energy resolution; a second crystal diffractor (1511) located in a second direction from the pulp treatment unit (201), the second crystal diffractor (1511) comprising a second crystal (1512); a second radiation detector (1515) configured to detect fluorescent x-rays diffracted by said second crystal (1512) with a second energy resolution, characterized in that the first crystal (603, 1502) is a pyrolytic graphite crystal; the second crystal (1512) is made of a material other than pyrolytic graphite; and the first and second crystal diffractors are configured to direct characteristic fluorescent radiation of the same element to their respective radiation detectors. 2. The X-ray fluorescence analyzer according to claim 1, wherein the second crystal (1512) is one of the following crystals: silicon dioxide crystal, lithium fluoride crystal, ammonium dihydrogen phosphate crystal, potassium hydrogen phthalate crystal. 3. An X-ray fluorescence analyzer according to claim 1 or 2, wherein said first energy resolution is better than 300 eV at a reference energy of 5.9 keV. 4. X-ray fluorescence analyzer according to any of the preceding claims, wherein the first radiation detector (1505) is one of the following detectors: a PIN diode detector, a silicon drift detector, a germanium detector, a germanium drift detector. 5. An X-ray fluorescence analyzer according to any one of the preceding claims, wherein the second radiation detector (1515) is a gas filled proportional counter. 6. X-ray fluorescence analyzer according to any one of the preceding claims, wherein said element is gold. 7. X-ray fluorescence analyzer according to any one of the preceding claims, in which block (201) pulp processing is configured to maintain a flat surface of the specified sample (202) pulp on the side facing the x-ray tube (402); the specified first optical axis (204) is located at an oblique angle to the specified flat surface; the first crystal diffractor (601, 1501) is located at such an angle of rotation around the specified first optical axis (204), in which the specified flat sample surface (202) covers the largest part of the field of view of the first crystal diffractor (601, 1501); and the specified second crystal diffractor (1511) is located at a different angle of rotation around the specified first optical axis (204). 8. X-ray fluorescence analyzer according to any one of claims 1 to 6, in which the pulp processing unit (201) is configured to maintain a flat surface of said pulp sample (202) on the side facing the x-ray tube (402) and said first optical axis (204) is perpendicular to the specified flat surface. 9. An X-ray fluorescence analyzer according to any one of the preceding claims, wherein the nominal input power of the X-ray tube (402) is at least 400 watts. - 17 040045 10. The X-ray fluorescence analyzer of claim 9, wherein the nominal input power of the X-ray tube (402) is at least 1 kW, preferably at least 2 kW, and more preferably at least 4 kW. 11. An X-ray fluorescence analyzer according to any one of the preceding claims, wherein the optical path between the X-ray tube (402) and the pulp processing unit (201) is straight without a diffractor located therebetween. 12. An X-ray fluorescence analyzer according to any one of the preceding claims, wherein the X-ray tube (402) comprises an anode for generating said incident X-ray radiation (206) and a pulp processing unit (201) configured to maintain the shortest straight-line distance between said pulp sample (202) and said anode which is less than 50 mm, preferably less than 40 mm and more preferably less than 30 mm. 13. The X-ray fluorescence analyzer according to claim 12, wherein the X-ray tube (402) is an end-window X-ray tube. 14. An X-ray fluorescence analyzer according to any one of the preceding claims, wherein the diffractive surface of said pyrolytic graphite crystal (603) is one of the following surfaces: singly bonded surface, curved in one direction, singly bonded surface, curved in two directions, axisymmetric nonsingly bonded surface. 15. X-ray fluorescent analyzer according to any one of the preceding paragraphs, containing a housing (401); the front wall (303) of said body (401); an opening in said front wall (303); and a holder (403) for removably holding said pulp treatment unit (201) on the outer side of said front wall (303) aligned with said opening in said front wall (303). 16. An X-ray fluorescence analyzer according to claim 15, wherein both the X-ray tube (402) and the first crystal diffractor (601) are inside said analyzer housing on one side of said front wall (303). 17. X-ray fluorescence analyzer according to any one of the preceding claims, comprising a filter plate (205) located in the optical path between the x-ray tube (402) and the pulp processing unit (201). 18. X-ray fluorescence analyzer according to claim 17, wherein said filter plate (205) is located closer to the x-ray tube (402) than to the pulp processing unit (201). 19. An X-ray fluorescence analyzer according to any one of the preceding claims, comprising a calibration plate (305) and an actuator (501) configured to controllably move the calibration plate (305) between at least two positions, of which the first position is not in the path of the incident X-ray radiation (206), and the second position is in the path of the incident X-rays (206) and in the field of view of the first crystal diffractor (601). 20. The method of performing x-ray fluorescence analysis using the device according to any one of claims 1 to 19, including irradiating the sample (202) of the pulp with incident x-rays and obtaining fluorescent x-rays from the irradiated sample (202); extraction of the first and second predetermined wavelength ranges (1301, 1401) from the respective first and second parts of the specified received fluorescent X-ray radiation using the respective first and second crystal diffractors (1501, 1511), while both the first wavelength range (1301) and and the second wavelength range (1401) includes the characteristic fluorescent radiation of the same element, and the first wavelength range (1301) is at least twice as wide as the second wavelength range (1401); detection of fluorescent X-ray radiation in the specified first and second selected wavelength ranges by the respective first and second radiation detectors (1505, 1515), thus providing the corresponding first and second detection results, while the energy resolution of the first radiation detector (1505) is better than 300 eV at a reference energy of 5.9 keV, calculating the concentration of said element in said sample (202) from at least one of said first and second detection results. 21. The method of claim 20, wherein said calculation calculates the total intensity of background radiation and fluorescent x-ray radiation from elements other than said element using at least one of the first and second detection results; subtract from the total intensity detected in the wavelength range containing the specified characteristic peak of the fluorescent X-ray emission of the element to be measured in the specified sample, the calculated total intensity of the background radiation and fluorescent X-ray emission from other elements in the specified sample, other than the specified representative element interest; and presenting the result of said subtraction as a calculated intensity of said characteristic x-ray fluorescent peak. 22. The method according to claim 20 or 21, wherein said calculation analyzes said first and second detection results in excess of the influence of a characteristic peak from another element on the first detection result of a predetermined threshold; and if said analysis shows that the effect of said characteristic peak of said other element on the first detection result exceeds said predetermined threshold, calculating the concentration of said element in said sample (202) from the second detection result; and if said analysis shows that the effect of said characteristic peak of said other element on the first detection result does not exceed said predetermined threshold, the concentration of said element in said sample (202) is calculated from the first detection result. 23. The method according to any one of claims 20-22, wherein said element is gold. 24. A method according to any one of claims 20-22, wherein said characteristic fluorescent emission contains a K or L peak of an element with 30 < Z < 92, where Z is the atomic number of said element. 25. The method according to any one of claims 20-24, wherein said sample (202) contains said element in a matrix consisting primarily of elements with Z < 8, where Z is an atomic number.