Combined confocal X-ray fluorescence and X-ray computerised tomographic system and method

Abstract

A correlative evaluation of a sample (104) using a combined x-ray computed tomography (CT) and x-ray fluorescence (XRF) system and the method for analyzing a sample (104) using x-ray CT and XRF is disclosed. The CT/XRF system (10) includes an x-ray CT subsystem (100) for acquisition of volume information and a confocal XRF subsystem (102) for characterization of elemental composition information. Geometrical calibration is carried out between the XRF subsystem (102) and the X-ray CT subsystem (100) such that a region of interest defined during X-ray CT acquisition can be retrieved by the XRF subsystem (102) for a subsequent XRF acquisition. The system (10) combines the sub-micrometer spatial resolution 3-D imaging capability of x-ray CT with the elemental composition analysis of confocal XRF to provide 3-D elemental composition analysis of a sample (104) with ppm level sensitivity. This is applicable to many scientific research and industrial applications, a prime example of which is the elemental identification of precious metal grains in crushed and ground ores and floatation tailings in the mining industry.

Claims

1. An x-ray computed tomography (CT)/x-ray fluorescence (XRF) system comprising: a source for generating x-rays; an x-ray CT subsystem for acquisition of volume information of the sample; a confocal XRF subsystem for acquisition of elemental composition information including a source optical element that directs the x-rays received over range of angles from the source to generate a focused beam onto a region of interest of a sample located at a confocal probing spot of the confocal XRF subsystem; and a controller in communication with the x-ray CT subsystem and the confocal XRF subsystem, wherein the controller manages the acquisition by the x-ray CT subsystem and the confocal XRF subsystem by selecting a number of points or areas predetermined from the volume information for the acquisition of elemental composition information by the XRF subsystem and placing the selected points or area at the confocal probing spot.

2. The system of claim 1, wherein the controller provides spatial calibration of the confocal XRF subsystem based on the volume information received from the x-ray CT subsystem.

3. The system of claim 1, wherein the controller combines the volume information from the x-ray CT subsystem with the elemental composition information from the confocal XRF subsystem to verify a liberation state of elements within a sample.

4. The system of claim 1, wherein the controller selects a limited number of points or small areas predetermined from the volume information for the acquisition of the elemental composition information by the confocal XRF subsystem.

5. The system of claim 1, wherein the controller performs correlation between the volume information and the elemental composition information to provide elemental contrast of a sample as a function of depth.

6. The system of claim 5, wherein in response to the correlation, the controller generates elemental distribution maps as a function of position within the sample.

7. The system of claim 1, wherein the controller: generates an interactive graphic, which enables the identification and selection of a feature of interest within the volume information acquired by the x-ray CT subsystem; creates a region of interest that includes the feature of interest, the region of interest created to translate the feature of interest from x-ray CT subsystem coordinates to confocal XRF subsystem coordinates; and accesses the region of interest with the confocal XRF subsystem for the acquisition of the elemental composition information.

8. The system of claim 1, wherein the controller generates a coordinate transfer function that translates between x-ray CT subsystem coordinates and confocal XRF subsystem coordinates that accounts for differences in resolution between the x-ray CT subsystem and the confocal XRF subsystem.

9. The system of claim 1, wherein the controller generates a coordinate transfer function for translating selected features of interest from the volume information of the x-ray CT subsystem into a region of interest for acquisition of the elemental composition information by the confocal XRF subsystem.

10. The system of claim 9, wherein the confocal XRF subsystem positions the region of interest at a confocal probing spot of the confocal XRF subsystem by referencing the coordinate transfer function.

11. The system of claim 1, wherein the controller corrects the elemental composition information acquired by the confocal XRF subsystem using absorption information of the volume information acquired by the x-ray CT subsystem.

12. The system of claim 11, wherein the absorption information is associated with voxel intensities of the volume information.

13. The system of claim 1, wherein the controller compares the elemental composition information acquired by the confocal XRF subsystem against reference elemental information to identify elements in a sample.

14. The system of claim 1, wherein the confocal XRF subsystem selectively probes a region of interest of the volume information at a sub-micrometer spatial resolution for the acquisition of the elemental composition information.

15. The system of claim 1, wherein the source optical element is an optical polycapillary element.

16. The system of claim 1, wherein the confocal XRF subsystem further comprises a collection optical element having a focus that intersects a focus of the source optical element.

17. The system of claim 16, wherein the collection optical element is an elliptical polycapillary optic.

18. The system of claim 1, wherein the source optical element is switchable to be moved out of a path of the x-rays when the system is in CT mode and acquiring the volume information of the sample.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) In the accompanying drawings, reference characters refer to the same or similar parts throughout the different views. The drawings are not necessarily to scale; emphasis has instead been placed upon illustrating the principles of the invention. Of the drawings:

(2) FIG. 1A is a perspective schematic view of an integrated x-ray CT/XRF system in CT mode according to an embodiment of the invention.

(3) FIG. 1B is a perspective schematic view of the integrated x-ray CT/XRF system from FIG. 1A in XRF mode.

(4) FIG. 2 is a perspective schematic view of an x-ray CT/XRF system including a separate x-ray CT subsystem and a separate confocal XRF subsystem that function together according to another embodiment of the invention.

(5) FIG. 3 is a top schematic view of an integrated x-ray CT/XRF system running both the CT mode and the XRF mode simultaneously or serially according to another embodiment of the invention.

(6) FIG. 4 is an exemplary interactive graphic provided by a controller of the x-ray CT/XRF system for a tailings sample, where the interactive graphic enables operator selection of a feature of interest within a slice selection of the CT volume dataset of the sample;

(7) FIG. 5 shows the XRF spectral data obtained by the CXRF subsystem for the region of interest in FIG. 4;

(8) FIGS. 6A-6D shows a workflow that illustrates the correlative aspects of the invention, for a feature of interest selected by an operator within a representative tailings sample;

(9) FIG. 7 is a plot of x-ray photon counts as a function of energy (keV) showing the XRF spectral data for a tailings sample that includes Platinum (Pt), for illustrating the sensitivity of the x-ray CT/XRF system;

(10) FIG. 8 is a schematic block diagram of a controller showing its interaction between the x-ray CT subsystem and the CXRF subsystem;

(11) FIG. 9 is a flowchart illustrating an x-ray CT/XRF system calibration method according to an embodiment of the invention;

(12) FIG. 10 is a flowchart illustrating a method for acquiring and preparing a mineral material for x-ray CT/XRF analysis according to an embodiment of the invention;

(13) FIG. 11A is a flowchart illustrating a method of analyzing a sample using separate x-ray CT and Confocal XRF subsystems according to an embodiment of the invention;

(14) FIG. 11B is a flowchart illustrating a method of analyzing a sample using an integrated x-ray CT/XRF system according to an embodiment of the invention;

(15) FIG. 12A is a flowchart illustrating a raster scan analysis of a sample using separate x-ray CT and Confocal XRF subsystems according to an embodiment of the invention;

(16) FIG. 12B is a flowchart illustrating a raster scan analysis of a sample using an integrated x-ray CT/XRF system according to an embodiment of the invention; and

(17) FIG. 13 is a flowchart illustrating a method of determining the composition of a sample using an x-ray CT/XRF system according to an embodiment of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(18) 3-D images delivered by x-ray CT include a cubic array of voxels with “average” attenuation values. These values are represented by “grey levels”. The differences in grey levels, or voxel intensities, correspond to differences in attenuation and are representative of the material that caused this attenuation. A histogram of these grey level values allows one to analyze and determine the various materials and their phases present in the image, provided the grey levels are suitably calibrated. These grey levels are often sufficient for many applications such as IC failure analysis or porosity analysis for oil and gas rock analysis.

(19) For tailings analysis in mining, ambiguities arise due to so-called grey-scale overlaps. These occur if materials with different elemental composition exhibit very similar or same grey levels in the x-ray CT data. For example, if a small high-density grain is found in a tailings sample, it is generally not possible to positively identify this grain as a platinum compound compared to some other high-density compounds (e.g. lead, palladium, etc.). Several hundred known platinum compounds can be present with varying density (and grey scale values) in addition to countless other high-density compounds composed of other elements, making elemental identification extremely important.

(20) The system and method here enables the probing of a grain for elemental composition analysis to give a positive identification to the element and further enables the narrowing of the precise compound. Additionally, in biological applications, the availability of trace-element sensitive spatially resolved probing is useful for toxicological studies to directly put structural changes in context with the presence of toxins.

(21) One of the proposed solutions to some of the problems with the current technology is to complement the x-ray CT scanner with a confocal XRF setup that is able to selectively probe points or small volumes of the x-ray CT volume at a spatial resolution of less than ˜(100 um).sup.3 and preferably less than ˜(50 um).sup.3, and in some cases less than ˜(20 um).sup.3 with ppm level sensitivity.

(22) The ppm sensitivity on bulk homogeneous samples is achieved with confocal XRF. For highly complex and heterogeneous samples as well as samples with very small bulk concentrations where elements to be detected are highly concentrated in very small volumes and localized in space (such as in tailings samples in mining industry), confocal XRF has significant advantages over conventional XRF due to its ability to define the volume of interest to be probed. Confocal XRF also allows for significantly improved signal to background ratios to efficiently detect localized concentrations of elements in thick samples.

(23) One of the chief reasons why the confocal XRF technique has not been commercialized is the non-practicality of scanning large volumes in a reasonable time. Although each point takes only a very short measurement time, usually measured in seconds, a small volume (50×50×50 pixels) can lead to exorbitant measurement times (173 hours for example for a point dwell time of 5 seconds). The key to make confocal XRF practical is to combine it with x-ray CT and only probe a limited number of points in the 3-D volume that are relevant.

(24) This is achieved with the combined CT/XRF system by performing an x-ray CT scan first and then selecting, based on the 3-D structural information, where to collect the elemental information data. As noted earlier, the very small concentrations of platinum are found in micron sized grains highly localized in space and for tailing samples measured in the mining industry only a small number (sometimes as low as ˜10) of grains per x-ray CT scan will be of interest for elemental identification. Such geometry lends itself naturally for the confocal XRF technique and these can be measured by confocal XRF within minutes, adding only minimal time to the CT scan times. This is due to the fact that the grains containing platinum have volumes of the order of (2 μm).sup.3 inside the confocal volumes of the order of (20 μm).sup.3 or about 1000 ppm by volume. This leads to very efficient detection of precious metals which is not possible with conventional XRF.

(25) The same argument holds true for other key applications requiring elemental identification, such as CT grey scale calibration, which generally requires probing only a very limited number of data points in order to correlate grey levels with elemental make-up. For biological applications, it is generally sufficient to probe only within the tissue of the same kind to get a representative level of trace elements. For example, in a Zebra fish sample, one would perform measurements within the organs (or a few points within each organ) and not attempt to point-scan the whole specimen.

(26) The coupling of high resolution x-ray CT with confocal XRF helps mitigate and overcome the inherent limitations of the confocal XRF technique on sample sizes. In particular, the coupling overcomes such limitations as limited x-ray penetration into samples and absorption of emergent fluorescence x-rays in the sample. This is because for high resolution x-ray tomography, sample sizes are typically of the order of 1-2 mm. For optimal excitation of platinum (and other precious metals) L fluorescence lines, Mo and Ag sources are, in some embodiments, used with polycapillary optics optimized for MoKα (˜17.5 kilo electron volts (KeV)) and AgKα (˜22.2 KeV) radiation, for which the penetration depths inside tailings samples are of the order of ˜2 millimeters (mm). For platinum Lα lines (˜9.4 KeV) about 10% transmission is expected over ˜1 mm which corresponds to a total sample thickness that can be used of ˜2 mm (penetration of L x-rays are even larger). Higher transmission is expected for other L lines such as Lβ and Lγ x-ray lines. For biological applications (for example, tissues), greater x-ray penetration depths (and consequently deeper probing) is achieved, which allows for larger samples to be used. For CT grey scale calibration, exhaustive probing of points deep inside the sample is usually not required. Instead, limited probing of selected points at or near the surface of the specimen is typically sufficient.

(27) FIG. 1A shows an example of the integrated x-ray CT/XRF system 10 in CT mode, which has been constructed according to the principles of the present invention.

(28) This integrated system 10 includes an x-ray CT subsystem 100, or CT subsystem, and a confocal x-ray fluorescence subsystem 102, or CXRF subsystem.

(29) The integrated system 10 also includes a controller 128 that manages various parts of the x-ray CT subsystem 100 and the confocal x-ray fluorescence subsystem 102. The controller 128 receives data from these subsystems 100/102. Based on the data received, the controller 128 manages how these subsystems 100/102 operate accordingly and their operations coordinated.

(30) The controller 128 also includes an interactive graphic 170. The interactive graphic 170 provides interactive operator selection of features of interest 120 from the acquired volume information 150 of the sample 104, and presents the selection and display via display device 136. More detail on the controller 128 and its other components accompanies the descriptions for FIG. 4 and FIG. 8, herein below. These subsystems 100/102 are used together to determine the mineral content of a sample 104. Examples of samples 104 include mineral samples such as sandstone, bituminous sand, ore, and coal or samples containing precious metals or fluids, such as water or crude oil.

(31) In the illustrated example, the sample 104 is held within a sample tube 110. This sample tube 110 is a glass or plastic test tube, for example. In one example, the size of the sample is about 1-4 mm in diameter and ˜2-20 mm in height.

(32) In another example, the sample is a flat sample that is placed, or in a sample container other than the sample tube. Alternatively, the sample has a different configuration such as cubic or spherical. When flat samples are used, x, y, z sample stage 112 has the capability to position different regions of the flat face of the samples in the x-ray beams.

(33) The sample tube 110 is mounted on an x, y, z sample stage 112 that allows for the positioning of the sample 104 in the beam 107 generated by the x-ray source 106. In one example, the x, y, z sample stage 112 comprises three orthogonal motorized positioning arms that allow for the positioning of the sample 104 along each of the three orthogonal dimensions under the control of the controller 128.

(34) The x, y, z sample stage 112 has an integrated rotation stage 114 preferably attached underneath the x, y, z sample stage 112. The integrated rotation stage 114 enables the x, y, z sample stage 112 to rotate through 360 degrees around the vertical (z) axis. This allows the x, y, z sample stage to be rotated along with the sample tube 110 and sample 104 to any desired angle (θ) with respect to the x-ray CT subsystem 100 and/or confocal x-ray fluorescence subsystem 102.

(35) The rotation stage 114 is preferably a precision rotary stage such as an air bearing stage or a mechanical ball-bearing or roller bearing stage.

(36) The x-ray CT subsystem 100 includes an x-ray source 106 that generates the x-ray radiation beam 107. The x-ray source 106 is preferably a laboratory based source such as a sealed tube, rotating-anode, or micro-focus x-ray source. In one example, the x-ray source is a transmission x-ray micro-focus source with a source size of about 2 μm or less operated up to a maximum power of 10 Watts. The small spot size or source size allows for high resolution imaging. Use of a transmission source allows for a favorable imaging geometry by shortening the distance between the source and a detector enabling high throughput.

(37) When the sample 104 is a mineral sample, a high-energy x-ray radiation beam is selected with energy above several keV for the x-ray source 106. This is typically required to penetrate samples with tens of micrometers or greater thickness. Higher energy radiation of tens of keV is used when the sample is usually about a millimeter or greater in thickness. Generally, the range is 5-150 keV. In one embodiment, the x-ray source has a tungsten target.

(38) The x-ray CT subsystem 100 also includes a spatially resolved transmission detector 108. In one example, the detector 108 is a scintillated CCD detector with optical magnification as described in U.S. Pat. No. 7,057,187 B1, which is incorporated herein by this reference in its entirety.

(39) Specifically, the spatially resolved detector 108 records x-ray radiation 107 that passes through the sample 104. In one example, the spatially resolved detector 108 includes a scintillator, a charge-coupled device (CCD) camera, and a lens or lens system for imaging visible light from the scintillator onto the CCD camera.

(40) In CT mode, the x-ray CT subsystem 100 performs sub-micrometer spatial resolution 3-D imaging of the sample 104. The x-ray source 106 emits x-rays 107 that penetrate the sample 104. Then, the x-rays 107 are detected and recorded by the spatially resolved detector 108.

(41) The generated images are passed from the spatially resolved detector 108 to the controller 128. The controller 128 operates the x-ray source 106 and the spatially resolved detector 108 to generate a series of projections 156 of the sample 104. The controller 128 also operates the rotation stage 114. In between projections 156, the controller 128 advances the rotation stage 114 through a predetermined arc. Once enough projections 156 have been generated at different values for θ, the controller 128 combines the images or projections 156 into a 3-D tomographic image of the sample 104. In one embodiment, the sample is rotated through 360 degrees during the scan. In another embodiment, the angular range for scanning the sample is limited to 180 degrees or lower.

(42) In one example, a series of 2D projection images 156 (typically about 1600 images or more) are obtained and mathematically reconstructed to produce a 3-D volume, with grey value in each voxel representing the optical density. The conditions of acquisition of images are optimized for speed and maximization of signal to noise ratios.

(43) Source filters 157 are used in some examples, to pre-harden the beam 107 incident on the sample 104 to ensure proper beam hardening corrections to the reconstructed data. Suitable calibration can be performed and CT numbers obtained in terms of Hounsfield Units which gives a measure of linear attenuation coefficient in the material at a given location.

(44) In one embodiment, the controller 128 commands the switching of the x-ray source 106 from CT mode to XRF mode and back to CT mode. This can be accomplished with a switchable source optical element 116 such as an elliptical polycapillary optic. In one example, the switchable source optical element 116 is attached to the x-ray source 106 via a moveable arm 118. When the controller 128 directs the x-ray source 106 to switch to XRF mode, the switchable source optical element 116 is positioned into the path of the x-rays 107 and in front of the x-ray source 106 as shown in FIG. 1B. When the controller 128 directs the x-ray source 106 to switch back to CT mode, the switchable source optical element 116 detaches from the x-ray source 106 and is moved out of the path of the x-rays 107 as shown in FIG. 1A.

(45) FIG. 1B shows an example of the integrated x-ray CT/XRF system 10 in XRF mode.

(46) During XRF mode, the source optical element 116 functions to control and direct x rays over a broad range of angles and energies from the source to generate a small focused beam 109 onto a volume or region of interest 122 in the sample 104.

(47) The confocal x-ray fluorescence subsystem 102 further includes a collection optical element 124 and an x-ray fluorescence detector 126.

(48) The collection optical element 124 is preferably a half optic that can be used to minimize the total number of reflections of x-rays inside the optical element 124. In another example, the collection optical element 124 is an elliptical polycapillary optic. The collection optical element 124 gathers fluorescence x-rays 111 that are generated within the particular region of interest 122.

(49) The intersection of the foci of source optical element 116 and the collection optical element 124 defines the confocal volume which depends on the focus size and strongly depends upon the detected fluorescence x-ray energy. The confocal volume, as defined by the intersection of the foci of source optical polycapillary element 116 and the collection optical polycapillary element 124, is approximately (50 μm).sup.3 in one example or (20 μm).sup.3, or less, in another example. In a current embodiment, the source optical element 116 is optimized for AgKα (˜22.2 KeV) characteristic x-rays from the source 106. On the detection side, the half optic collection optical element 124 is used to minimize the total number of reflections of x-rays inside the capillaries and increase efficiency of detection of fluorescence x-rays from the sample.

(50) In general, the collection optical element 124 is optimized to detect fluorescence x-rays in the energy range of 9-13 KeV (suitable for detection of precious metals) most efficiently. For examples where mid Z transition elements have to be efficiently detected, the collection optical element 124 is suitably optimized for the K lines of these elements whose energies lie in the 7-10 KeV range. The length of the collection optical element 124 is optimized and the distance between the energy dispersive detector 126 and the sample surface is minimized to reduce the absorption of softer fluorescence x-rays in air such that the sensitivity of the instrument is not severely compromised.

(51) The collection optical element 124 produces or collimates a parallel beam of fluorescence x-rays 113 onto the x-ray fluorescence detector 126. The x-ray fluorescence detector 126 is an energy-dispersive x-ray spectroscopy (EDS) detector, in one example. The EDS detector converts the x-ray energy received from the collimated fluorescence x-rays 113 into voltage signals. Based on the voltage signals (characterizing x-ray energy), elemental composition information 152 of the region of interest 122 is determined.

(52) The controller accepts the elemental composition information 152, also known as an XRF spectral dataset 152, from the CXRF subsystem 102. The XRF spectral dataset 152 acquired by the CXRF subsystem 102 includes individual characteristic peaks 166 associated with individual elements.

(53) In examples, XRF spectral datasets 152 for regions of interest 122 include characteristic peaks 166 associated with different elements, such as the FIG. 7 Lα line 162 and the FIG. 7 Lβ line 164 of platinum (9.45 KeV and 11.2 KeV), and the FIG. 5 characteristic peaks 166 of lead (10.45 KeV and 12.62 KeV) and Copper (Cu). The characteristic peaks 166 identify and distinguish between the elements in the sample 104.

(54) The controller 128 is in communication with the confocal x-ray fluorescence subsystem 102. Based on the results of the CT analysis in FIG. 1A, the controller 128 selects subsequent regions of interest 122 from the volume information 150 to be probed by the x-ray fluorescence subsystem 102. In more detail, the controller 128 operates the x, y, z sample stage 112/rotation stage 114 to rotate around the z-axis and position the sample tube 110 along the x, y, and z axes in relation to the confocal x-ray fluorescence subsystem 102. This provides desired positioning of the region of interest 122 with respect to the x-ray source 106 and the collection optical element 124/x-ray fluorescence detector 126. The rotation/positioning is optimized to reduce absorption of fluorescent x-rays.

(55) In XRF mode, the confocal x-ray fluorescence subsystem 102 performs elemental composition analysis of the sample 104. The x-ray source 106 emits x-rays through the switchable source optical element 116 which focuses or directs the x-rays 109 towards an analysis region or a region of interest 122 in the sample 104. Typically, the x-rays used for CT mode will not be optimal for the XRF mode. Thus, in one example, the source 106 uses a new Ag target that can be operated at greater than the energy used in CT mode, such as greater than 20 Watts or up to 50 Watts or more.

(56) In general, in this and the subsequent embodiments, the CT imaging subsystem 100 typically uses an x-ray source with a high Z target such as tungsten target. However, for specialized CT applications other targets may be used such as Mo, Rh, Ag etc. The x-ray source produces a broad spectrum of x-rays with energies up to the maximum applied voltage chosen by the user. Any or all of the spectrum may be utilized for CT imaging. In contrast, the XRF subsystem 102 typically uses lower Z x-ray source targets such as molybdenum, rhodium, and silver that produce characteristic x-rays and continuum in the energy range most suitable to excite elements useful for precious metal mining analysis. Other lower Z targets such as Cu, Ni, Fe can also be used for XRF subsystem. If need be, even high Z targets containing W and Pt may also be used. The choice of targets depends specifically upon the application.

(57) The focused x-rays 109 induce fluorescence of elements within the analysis region 122, which are collected by the collection optical element 124. The collection optical element 124 receives emitted fluorescence x-rays 111. Then, the collection optical element 124 collimates the x-rays 113 towards an x-ray florescence detector 126 for detecting the florescence x-rays 111/113 from the region of interest 122.

(58) The controller 128 receives data regarding elemental composition information 152 of the region of interest 122 from the x-ray florescence detector 126 in the form of energy spectra. These data are converted by the controller 128 into their final form, XRF spectral data 152, that provide a map of the elemental composition of the sample 104 for each of the regions of interest 122 that is analyzed.

(59) In FIG. 2, the CT/XRF system 10 is shown as two separate subsystems. CT mode or function is performed using an x-ray CT subsystem 100. Then, the sample tube 110 is transferred to a separate confocal XRF subsystem 102. These separate subsystems are used in combination to determine the mineral content, for example, of the sample 104 within the sample tube 110.

(60) This CT/XRF system 10 includes a controller 128 that manages various parts of the x-ray CT subsystem 100 and the confocal x-ray fluorescence (CXRF) subsystem 102. The controller 128 receives data from these subsystems 100/102. Based on the data received, the controller manages how these subsystems 100/102 operate.

(61) Typically, the sample tube 110 is first mounted on an x, y, z sample stage 112A of the x-ray CT subsystem 100. As described above, this x, y, z sample stage 112A has an integrated rotation stage 114A for rotating the sample tube 110 to any desired angle (θ) with respect to the x-ray CT subsystem 100. The x-ray CT subsystem 100 includes a first x-ray source 106A and a spatially resolved transmission detector 108.

(62) During CT mode, the CT subsystem 100 is activated to collect a series of projections 156 at different values for θ. The projections 156 are combined into a 3-D tomographic reconstruction 150 of the sample 104.

(63) Specifically, the first x-ray source 106A emits diverging x-rays 107 through the features of interest 120. In one example, the x-ray source 106A is a transmission x-ray micro source with source size of about 2 μm or less and preferably 1.5 μm or less operated up to a maximum power of 10 Watts. In another example, the x-ray source 106A is a reflection source having a source size or spot size of about 5 μm or less operated at a power between about 10 and 30 Watts. These x-rays 107 are captured and recorded by the spatially resolved detector 108. The spatially resolved detector 108 generates images from the detection of these x-rays 107. These images are passed to the controller 128 to develop 3-D tomographic images 150 of the sample 104. As described previously, the controller 128 operates the x-ray source 106A along with the integrated rotation stage 114A to obtain the projections 156 for the different values of θ.

(64) Next, the CT/XRF system 10 switches from CT mode to XRF mode. The sample 104 is switched from the x, y, z sample stage 112A of the CT system 100 to the x, y, z sample stage 112B of the XRF system 102. A compatible kinematic mounting system is used for both systems 100, 102 to enable reproducible sample transfer that keeps a fixed coordinate transformation between both systems. In other examples, a reference marker that can be identified by both systems is used to enable the transfer of the coordinate system between both systems.

(65) During the switch, the controller 128 provides spatial calibration of the confocal XRF subsystem 102 based on the results of CT mode. Spatial calibration is the transfer of system coordinates (x, y, z, θ) from the x-ray CT subsystem 100 to the confocal x-ray fluorescence subsystem 102. The controller 128 receives CT system coordinates 204 (x, y, z, θ).sub.CT from the x-ray CT subsystem 100 that characterize the locations of the features of interest 120. The CT system coordinates 204 (x, y, z, θ).sub.CT are converted by the controller 128 to XRF system coordinates 207 (x, y, z, θ).sub.XRF. This conversion defines a region of interest 122 for the confocal XRF subsystem 102.

(66) The confocal XRF subsystem 102 includes a second x-ray source 106B in communication with the controller 128. The controller 128 adjusts this second x-ray source 106B accordingly during XRF mode. In one example, the second x-ray source 106B is a laboratory micro focus source with Ag target that is operated at greater than 20 Watts or 50 Watts or more. Generally, however, the source's target as a Z of less than 50, as discussed previously.

(67) The confocal XRF subsystem 102 also has a second rotation stage 114B and a second x, y, z sample stage 112B for receiving and holding the sample tube 110 from the x-ray CT subsystem 100. The controller 128 directs the rotation of the sample tube 110 via the second x, y, z sample stage 112B/rotation stage 114B based on the XRF system coordinates 207.

(68) The confocal XRF subsystem 102 includes the source optical element 116 such as an elliptical polycapillary optic for focusing the x-rays 109. The source optical element 116 can be attached to the second x-ray source 106B or placed in front of the second x-ray source 106B.

(69) The confocal XRF subsystem 102 is able to collect fluorescence radiation with the collection optical element 124 and detect that radiation with the x-ray fluorescence detector 126 as described above.

(70) In XRF mode, the separate confocal x-ray fluorescence subsystem 102 performs elemental composition analysis of the sample 104, which is typically limited to regions of interest 122 identified in the CT analysis. The second x-ray source 106B emits x-rays 107 through the source optical element 116 which focuses the x-rays 109 towards a region of interest 122 in the sample 104. In the illustrated example, the controller operates the rotation stage 114B and the x, y, z sample stage 112B of the XRF subsystem 102 to place the desired regions of interest 122 at the focal point of the source optical element 116 and the collection optical element 124. The focused x-rays 109 induce fluorescence of elements within the region of interest 122. Fluorescence x-rays 111 are received and guided by the collection optical element 124 towards an x-ray florescence detector 126 for detecting the florescence x-rays 111/113 from the region or volume of interest 122.

(71) The controller 128 receives data regarding energy of the x-rays detected by the fluorescence detector 126 and thus the elemental composition information 152 of the region of interest 122. The XRF spectral datasets 152 provide a map of the elemental composition of the sample for the analyzed regions of interest 122. After performing correction of the XRF spectral 152 using depth information of the volume information, the controller matches the XRF spectral datasets 152 to reference elemental information 168.

(72) Using the reference elemental information 168, the controller converts the XRF spectral data into element-specific estimates of concentrations present in the sample, identifying characteristic peaks 166 of the XRF spectral datasets 152. The controller 128 then displays the XRF spectral datasets 152 onto display device 136.

(73) In addition, the controller correlates information from the CT volume dataset/volume information 150 and the elemental composition information/XRF spectral datasets 152 for the region of interest 122. More detail for the correlation capability of the controller 128 accompanies the description for FIG. 6A-6D and FIG. 8.

(74) FIG. 3 illustrates another example of a CT/XRF system 10. This is a top view of an integrated CT/XRF system 10 that has the capability of running both CT analysis and XRF analysis simultaneously or sequentially without removing the sample from the system.

(75) In this example, two x-ray sources 106A/106B are used. The first x-ray source 106A is for CT analysis. The second x-ray source 106B includes an optical element 116 attachment for XRF analysis.

(76) In simultaneous CT/XRF mode, the first x-ray source 106A emits a dispersive group of x-rays 107 across features of interest 120 in the sample 104. The second x-ray source 106B utilizes the source optical element 116 to emit a focused set of x-rays 109 at a region of interest 122 in the sample 104.

(77) The CT/XRF system 10 can simultaneously develop 3-D CT images as well as XRF elemental composition images during CT/XRF mode. The x-rays 107 are captured and detected by the spatially resolved detector 108. The focused x-rays 109 are collected and guided by the collection optical element 124. The collection optical element 124 captures the fluorescence x-rays 111 and directs the rays 113 as a parallel beam towards an x-ray florescence detector 126 for detecting the fluorescence rays 111/113 from the region of interest 122.

(78) The CT/XRF system 10 can alternatively run CT mode or XRF mode at different times from one another. For example, the CT/XRF system 10 is run in CT mode first to determine desired features of interest 120 in the sample 104. Then, the system 10 is run in XRF mode to analyze the regions of interest 122 that surround the desired features of interest 120 to thereby determine the elemental composition.

(79) FIG. 4 displays an exemplary interactive graphic 170 generated by the controller 128, for a representative tailings sample 104. Preferably, the interactive graphic 170 displays volume information 150 generated by the x-ray CT subsystem 100 from the projections for the sample 104. Using a slice selector 140, the operator selects a synthetic slice 142 within the volume information 150, and the interactive graphic 170 displays a 2-D image of the slice 142 in response to the selection.

(80) The synthetic slice 142 includes clusters 144 of pixels associated with grains of elements and compounds of the sample 104. In tailings samples, the clusters 144 typically include trace metals such as Iron (Fe), Platinum (Pt), Copper (Cu), and Chromium (Cr).

(81) The gray levels of pixels in a slice 142, or voxels within the full CT volume dataset 150, correspond to attenuation of x-rays for features within the sample 104 in response to the CT scan. The brightness of the grey levels for each pixel/voxel reflect the proportion of X-rays scattered or absorbed as they pass through the sample. The brightest clusters 144 correspond to metallic elements and compounds, which typically absorb the highest proportion of incident x-rays within a sample 104.

(82) Within the selected slice 142, the operator then selects a feature of interest 120, such as one of the brighter clusters 144 associated with metallic elements and compounds. The controller 128 then calculates a region of interest 122 that surrounds the feature of interest 120. The controller provides the region of interest 122 to the CXRF subsystem 102 to use as an acquisition target for performing elemental compositional analysis. More detail and supporting information for the creation of the region of interest 122 by the controller 128 accompanies the description for FIG. 8 in this section.

(83) FIG. 5 shows an XRF spectral dataset 152 for the slice 142 and region of interest 122 selected in FIG. 4. The XRF spectral dataset 152 has been compared to reference elemental information 168. In one example, the reference elemental information 168 is included in a database 154 that communicates with the controller 128. As a result, the XRF spectral dataset 152 includes characteristic peaks 166 of elements that uniquely identify the elemental composition of the region of interest 122 of the sample 104.

(84) FIG. 6A-6D shows a workflow that illustrates the correlative aspects of the CT/XRF system 10, for a feature of interest 120 selected by an operator within a tailings sample 104. As in FIG. 4, an operator uses the interactive graphic in FIG. 6A to select a feature of interest 120 within a slice 142. The controller 128 calculates a region of interest 122 in response to the selection. The region of interest 122 includes the feature of interest 120.

(85) FIG. 6B shows the XRF spectral data 152 obtained by the CXRF subsystem 102 for the region of interest 122 of FIG. 6A. The controller 128 generates the XRF spectral data 152 as outlined in the description for FIG. 2 and FIG. 5.

(86) The controller 128 then correlates the volume information 150 with the elemental composition information 152 for the region of interest 122 to create and display elemental composition maps, such as the maps of FIGS. 6C and 6D. FIG. 6C shows a map of the elemental distribution for the region of interest 122 as a function of depth, and FIG. 6D shows a contour map of elemental iron (Fe) for the selected feature of interest 120.

(87) FIG. 7 displays the XRF spectral data 152 for a tailings sample 104 that includes Platinum (Pt) to show the sensitivity of the x-ray CT/XRF system 10. The XRF spectral data 152 includes characteristic peaks 166 of different elements, such as metallic elements Iron (Fe), Chromium (Cr), Nickel (Ni), and Copper (Cu). Specifically, the XRF spectral data 152 includes characteristic peaks 166 of the Pt Lα line 162 and the Pt Lβ line 164, the display of which is not possible using currently available XRF-only systems.

(88) FIG. 8 shows more detail for the controller 128. In examples, the controller 128 is a separate card of a computer system, or a stand-alone device. The controller 128, in addition to other functions, manages the acquisition of volume information 150 on the x-ray CT subsystem 100, and the acquisition of elemental composition information on the confocal XRF subsystem.

(89) The controller 128 accepts the projection images 156 and calculates the volume information 150 from the x-ray CT subsystem 100, and includes the interactive graphic 170 that enables operator selection of a synthetic slice 142 and feature of interest 120. The controller 128 creates a region of interest 122 that includes the feature of interest 120. The creation of the region of interest 122 depends on whether the x-ray CT subsystem 100 and CXRF subsystem 102 are part of an integrated system, or exist as separate, non-integrated systems.

(90) In general, the resolution of the volume information/CT volume dataset 150 generated by the x-ray CT subsystem 100 is typically on the order of a few microns or less. In contrast, the resolution limit of current XRF systems is typically ˜(20 μm).sup.3. For the x-ray CT/XRF systems 10, the controller 128 must account for the differences in resolution provided by the x-ray CT subsystem 100 and the CXRF subsystem 102. As a result, the controller 128 creates a region of interest 122 of the volume information 150 that includes the selected feature of interest 120, and additional voxels of the volume information surrounding the selected feature of interest 120.

(91) The controller 128 generates a coordinate transfer function 158 to convert CT system coordinates 204 (x, y, z, θ).sub.CT from the x-ray CT subsystem 100 to XRF system coordinates 207 (x, y, z, θ).sub.XRF. This conversion creates the region of interest 122 for the confocal XRF subsystem 102. This conversion is also referred to as spatial calibration of the confocal XRF subsystem 102 based on the results of CT mode.

(92) Once the controller 128 determines the region of interest 122, the controller 128 provides the positions of the region of interest 122 to the CXRF subsystem 102 for acquisition of the elemental composition information 151. The controller 128 generates the XRF spectral dataset 152 as previously outlined in the description for FIG. 2 and FIG. 5.

(93) In a preferred embodiment, the controller 128 includes applications 138 that perform “offline” correlations between the volume information 150 and the XRF spectral datasets 152. This enables generation of such correlated information as the FIG. 6C 3-D elemental compositional plot as a function of depth, and the FIG. 6D 2-D contour plot of specific elements. The controller 128 displays the XRF spectral datasets 152 and the correlated information to the display device 136.

(94) FIG. 9 illustrates an example of the steps for a calibration process of a CT/XRF system according to an embodiment of the invention.

(95) In step 400, a series of projections are acquired covering the reference sample at different angles. In one example, the reference sample includes at least three fiducial markers, also known as fiducials, in three different planes. This is required for unambiguous coordinate transfer. Preferably, the fiducial markers have high contrast in CT mode and are identifiable in XRF mode. In one example, high density spheres, such as 1-10 μm diameter copper or gold spheres, function as fiducials. These fiducials are embedded in an epoxy matrix plug that functions as the reference sample. In another example, a cross-shaped fiducial is deposited on a membrane, that functions as the reference sample or target.

(96) Then, in step 402, CT volume data of the reference sample is generated from the projections. Next, in step 404, the one or more fiducials are found in the reference sample using the CT volume data.

(97) Depending on the type of CT/XRF system, the reference sample may need to be moved to the XRF system in step 406, using a common kinematic mounting system. In step 408, the sample is scanned to generate XRF data. In step 410, fiducials are found in the reference sample using the XRF spectral data. The reference sample fiducials from the CT are compared with the reference sample fiducials of the XRF in step 412. In step 414, based on the step 412 comparison, a coordinate transfer function 158 is determined between CT system coordinates (x, y, z, θ).sub.CT and XRF system coordinates (x, y, z, θ).sub.XRF.

(98) Before a mineral sample 104 is placed into a sample tube 110 for analysis by the CT/XRF system 10, a pack of mineral material needs to be acquired and prepared.

(99) FIG. 10 illustrates an example of the steps involved in converting mineral material to a mineral sample for analysis.

(100) In step 500, mineral material is mined. Then, in step 502, the mineral material is crushed and ground into a powder. This mineral powder is run through a flotation process in step 504. Flotation includes mixing the powder with detergent in an aqueous solution in order to separate mineral particles from the powder into the concentrate and tailings. Next, in step 506, the raw mineral material and/or concentrate and/or tailings are packed into a sample tube 110.

(101) In step 508, x-ray CT and confocal XRF analysis are performed on this mineral sample/concentrate/tailing 104 in the sample tube 110. Based on the results of the x-ray CT/confocal XRF analysis, the mining/flotation process is modified and/or optimized in step 510.

(102) The analysis of a sample using a CT/XRF system is a process that can be run through separate CT and XRF systems as described in the method of FIG. 11A, or through an integrated CT/XRF system as illustrated in the method of FIG. 11B.

(103) This process starts with a step 600/620 of acquiring a series of projections covering the sample 104 within the sample tube 110 at different angles. In step 602/622, CT volume data (x, y, z, θ).sub.CT of the sample (e.g. mineral sample) are generated from the projections. This is accomplished using tomographic reconstruction algorithms that are executed by the controller 128, for example.

(104) Next, in step 604/624, a user or computer algorithm executing on the controller 128 selects features of interest (x, y, z, θ).sub.CT in the CT volume data.

(105) In steps 606/626, regions or volumes of interest are defined from the selected features of interest based on the CT volume data of the sample. In general, the resolution of the CT volume data is a few microns, or less. In contrast, the resolution limit of the XRF system under current technologies is ˜(20 μm).sup.3. Thus, for each feature of interest that is identified in the CT volume data, a region of interest surrounding that feature is defined for the analysis executed by the XRF system.

(106) If the analysis is performed using separate CT and XRF systems, step 608 includes the transfer of a sample tube to the XRF system after the regions of interest are defined. In step 610/628, the selected regions of interest are positioned at the confocal probing spot of the XRF system by reference to the coordinate transfer function 158. In step 612/630, XRF spectral data from the regions of interest are acquired. The XRF spectral data are corrected using absorption information from the CT volume data in step 614/632. Then, in step 616/634, the XRF spectral data are matched with reference elemental information that correlates with elemental composition. Based on this match, the controller is able to identify the elemental composition of the regions of interest.

(107) A raster scan analysis of the sample with a CT/XRF system can also be performed using separate CT and XRF systems as illustrated in FIG. 12A, or an integrated CT/XRF system as illustrated in FIG. 12B.

(108) This process starts with step 700/720 of acquiring a series of projections covering the sample 104 within the sample tube 110 at different angles.

(109) In step 702/722, CT volume data (x, y, z, θ).sub.CT of the sample (e.g. mineral sample) is generated from the projections. The sample is segmented from the CT volume data S(x, y, z, θ) in step 704/724. If the analysis is using separate CT and XRF systems, in step 706 the sample tube 110 is transferred to the XRF system after the sample is segmented in step 704. Next, in step 708/726 surface points are raster scanned using segmentation with optimized rotation θ. The sub-surface of the sample is raster scanned using segmentation in step 710/728. In step 712/730, the raster scan sub-surface step 710/728 is repeated to penetration limit.

(110) The CT volume data are used to correct the XRF spectral data at depth using the transfer function to correlate CT volume data with XRF spectral data in step 714/732.

(111) FIG. 13 illustrates the process of determining the composition of the sample using CT and XRF systems.

(112) In step 800, the process starts with acquiring a series of projections covering the sample 104 within the sample tube 110 at different angles. CT volume data (x, y, z, θ).sub.CT of the sample (e.g. mineral sample) is generated from the projections in step 802. Next, in step 804, the CT volume data are segmented into classes based on gray level. In step 806, regions within each segmentation class are selected. Selected regions are positioned at the confocal probing spot of the XRF system in step 808. In step 810, the composition of the selected regions is determined. Then, in step 812, a phase map (x, y, z, θ).sub.CT of the sample is generated and the composition of the sample is determined using the composition information that is generated for each gray level in the CT volume data.

(113) While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.