APPARATUS AND METHOD FOR COMPOSITION MEASUREMENT
Technical Field
The invention relates to the use of X- or gamma- rays to determine the composition of materials such as minerals and coals which are comprised predominantly of low atomic number elements .
Background Art Numerous techniques based on the interaction of
X- and gamma-ray photons with matter are in common use in the coal and minerals industries for the on- and off-line characterisation of materials.
The best known is X-ray fluorescence (XRF) whereby the material under study is irradiated with X- or gamma-ray photons from an X-ray tube or radioisotope source. Atoms in the material absorb these photons and re- radiate lower energy photons having characteristic energies. Measurement of the spectrum of scattered and fluorescent photons allows the elemental composition of the target material to be deduced.
For low atomic number elements, the energy of the fluorescent photons is very low and they can only travel very short distances in most materials before being absorbed. Consequently, great care must be taken to ensure that the material being measured is fine grained, uniform and has a flat surface for accurate results to be obtained. These requirements make XRF unsuitable in many applications, especially when on-line measurements are needed.
Several methods have been proposed for the determination of the relative proportions of materials that can be considered to consist of essentially two components that have different average atomic numbers. One example is the determination of ash (average atomic number of about 14) in coal (average atomic number of about 6) . These methods all rely on measuring an effective atomic
number for the binary mixture, from which the proportions of the two components can be inferred. The methods in common use include: dual energy transmission and back- scatter, measurement of the Rayleigh (elastic) to Compton (inelastic) scattering ratio and the measurement of the conversion of high energy gamma-rays to electron/positron pairs
The principal drawback to all of these methods is that they assume that the material being measured consists of two unvarying components. If the composition of either or both components varies, or if a third component is present in an appreciable concentration, then the connection between the effective atomic number measured and the material's composition is weakened or lost. Accordingly, it would be advantageous to provide an alternative technique for obtaining information about the composition of a material .
Summary of the Invention In a first aspect, the present invention provides apparatus for obtaining information about the composition of a material comprised predominantly of low atomic number elements, the apparatus including: a source of X- or gamma-ray photons that is so located with respect to the material to be examined that at least some of the photons impinge on the material; a detector capable of measuring the energy of X- or gamma-rays scattered from the material being examined; shielding means for preventing source photons from reaching the detector directly; and analysis means for determining information about the composition of the material from the shape of the energy spectrum of photons inelastically scattered from the target material . In a second aspect, the present invention provides apparatus wherein the source of X- or gamma-rays is a radioisotope source producing photons with discrete
energies .
Brief Description of Drawings
Figure 1 is a schematic drawing of a preferred embodiment of the invention;
Figure 2 illustrates a typical scattered photon energy spectrum measured by the invention.
Description of the Preferred Embodiment The preferred embodiment uses the principle of
Compton profile analysis (CPA) to perform a quantitative measurement of the composition of a material being examined, such as mineral or coal samples . Whilst not allowing a full elemental analysis to be made, the proportions of a limited number of components of a multi- component mixture can be estimated. The preferred embodiment also allows the measurement of two component mixtures where the composition of one or both of the components is subject to some variation. The principle of Compton profile analysis can be summarized as follows. To a first approximation, when X- or gamma-ray photons inelastically (Compton) scatter from a material, their loss of energy depends only on the scattering angle via the so-called Compton relation. However, the motion of atomic electrons in the target material causes a Doppler shift in the energies of the scattered photons, with the result that the Compton scattering peak is broadened. Analysis of the shape of the Compton peak allows the momentum distribution of electrons in the target material to be determined. This distribution in turn allows the sample composition to be inferred. Embodiments of the invention can be used to measure both discrete and bulk samples; examples of the latter include material travelling on a conveyor belt or flowing through a pipeline. The method is especially applicable to low atomic number (Z) materials that are not amenable to conventional X-ray fluorescence (XRF)
techniques .
A preferred embodiment of source (S) and detector (D) geometry is shown in figure 1. Photons from source (S) are directed onto the target material (M) by a collimator (C) which acts to shield the detector (D) from direct photons from the source (S) . A fraction of these photons will be scattered from the material (M) towards a detector (D) capable of measuring their energy. The dotted line shows the path of a typical photon emitted from the source (S) and scattering from the target material (M) into the detector (D) . The spectrum of measured energies is read out and stored on a computer or other suitable device. Persons skilled in the art will appreciate that other geometric arrangements can be employed that allow the detector to measure photons scattered from the target without registering unscattered photons direct from the source.
In the preferred embodiment, the source is 241Am which produces copious 59.54 keV gamma-rays. 241Am is preferred as it is relatively cheap and readily available However, the source (S) can be any radioisotope producing X- or gamma-rays with discrete energies and where there are no strong X- or gamma-ray emissions having energies above the energy of the line used to perform the Compton Profile analysis. For example, depending on the application, 109Cd or 137Cs may be appropriate. In a variation, the source (S) could be a Bremstrahlung (X-ray tube) source, rendered quasi-monochromatic by filtration or diffraction from a suitable crystal. For example, a so-called K-edge filter, which produces a gradually rising spectrum with a sharp, high energy cut-off could be used. If using a Bremstrahlung Source, persons skilled in the art will realise that a compromise has to be reached between narrowing the spectrum down and filtering out too many X-rays. The Compton profile measurement can then be performed using the X-rays close to the high energy edge with the lower energy photons constituting a background to
the measurement .
The collimator (C) is made of a suitable material strongly absorbing of the gamma-rays produced by the source. Normally, a high atomic number material such as lead or tungsten would be used, but a lighter material would suffice for lower energy photons. The use of a narrow and fairly long collimator not only stops source photons from reaching the detector but is also believed to reduce the effects of multiple scattering. In the preferred embodiment, the target material is raw, crushed or ground material travelling along a conveyor belt. In a variation, the target material would be conveyed through a pipeline as a slurry or powder. In a second variation, the material would be a mineral or coal sample, placed in a container chosen to be largely transparent to the X- or gamma-rays used.
By way of example, the technique can be applied in order to:
■ measure the amount of ash and sulphur in coal ■ measure the solid loading of a slurry
■ measure the salt content of an oil/water/gas mixture
■ measure one or more components present in a multi-component mineral mixture. It has been found that unlike conventional techniques, the technique of the preferred embodiment is particularly suitable to separating out low atomic number (organic) elements from higher atomic number elements, whilst being minimally sensitive to just which high atomic number elements are present.
In the preferred embodiment, the detector (D) has an energy resolution significantly less than the width of the Compton scattering peak. Examples would include high- resolution semiconductor devices, made of silicon, hyper- pure germanium, cadmium zinc telluride (CZT) or mercuric iodide (Hgl) . In a variation, a lower resolution scintillator or proportional counter detector could be
used with filters to select different energy bands.
A typical photon energy spectrum collected with the proposed instrument is shown schematically in figure 2. The larger peak labeled (C) is due to Compton scattered photons. The smaller peak (R) is due to elastically or Rayleigh scattered photons. Multiple scattering in the target material broadens the portion of the Compton peak to the left of the dashed line (the "low" side of the Compton peak) . This effect tends to reduce the utility of this part of the spectrum for composition determination. Consequently, in the preferred embodiment, the portion of the photon spectrum between the centre of the Compton peak (dashed line) and the Rayleigh peak (R) is analysed (the "high" side of the Compton peak) . The shape of this portion is analysed to infer information about the composition of the material being measured.
More specifically, the energy spectrum of the Compton scattered photons is recorded in the form of a histogram. The width of each bin in the histogram is small compared to both the width of the Compton peak and the resolution of the detector used to measure it. This means that a relatively large number of bins are present across the width of the Compton peak and guarantees that maximum information about the shape of the Compton profile is preserved.
To relate the measured Compton profile to parameters of interest such as ash fraction in coal or solids loading in a slurry, the right hand side of the Compton peak is divided into several (typically 3-5) windows. Each window consists of a contiguous group of histogram bins. For each window, the number of counts (recorded photons) falling in all bins that form the window is formed into a sum. This process is repeated for a suite of samples (typically 50 or so) have different compositions covering the range of interest.
A linear regression is then performed between the window count rates (and an additional constant term) and the parameter of interest.
The number of windows used and the window boundaries are adjusted to achieve the best (smallest) average squared error.
This analysis technique of the preferred embodiment extracts at least two parameters from the shape of the peak to enable measurement of materials which are quasibinary mixtures of two components. That is, where the composites of either or both components varies or a third component is present in an appreciable concentration. Example 1. A prototype Compton profile analyser (CPA) for the determination of ash in coal was constructed. Fifty synthetic samples were prepared simulating different coals with variable ash, sulphur and moisture contents. The chemical composition of both the coal (organic matter) and ash constituents were also highly variable.
The energy spectrum of scattered 59.54 keV gamma- rays from a 241Am source was measured for each sample using a high purity germanium detector. The shapes of the right- hand sides of the inelastic scattering peaks were analysed to deduce an ash calibration using the linear regression technique referred to above. An accuracy of 0.58 wt% ash was achieved. For comparison, the theoretical error using a standard dual energy transmission (DUET) analyser for the same set of coals would be 1.4 wt% . The improvement is due to the considerably reduced sensitivity of the CPA technique to changes in ash and coal composition.
The reduced sensitivity to ash composition is summarised in table 1 below. The table shows the apparent changes in ash content in a sample containing 10% ash by weight that would be measured when the calcium or iron content of the ash is changed by 1%. The three columns show the sensitivity to ash composition of dual energy
transmission (DUET) , pair production (PAIR) and Compton profile analysis (CPA) gauges. The CPA technique is typically 2-4 times less sensitive to ash composition than a pair production gauge and 7-8 times less sensitive than a DUET gauge.
Error in ash (wt%)
Change DUET PAIR CPA
+1% CaO in ash 0.24 0.11 0.03
+1% Fe203 in ash 0.57 0.16 0.08
Table 1. Comparison of sensitivities to ash composition of DUET, pair production and CPA coal analysers.