********

Field of the invention

The present invention relates to a method for identifying chemical species in a sample by x-ray fluorescence and their quantification in absolute values.

Background of the invention

X-ray fluorescence has been used for the analysis of particulate matter PM10 in air since the 1970s, for a quantitative determination of metals in particular.

In this context, the document "Sources and Elemental Composition of Aerosol in Pasadena, California, by Energy-Dispersive X-ray" by Robert H. Hammerle and William R. Pierson documents the use of an automatic spectrometer for determining the amount of 9 metals (Ca, Ti, V, Mn, Fe, Ni, Zn, Br, Pb) in the particulate matter PM10 present in the air as early as 1972. An automatic device for sampling and analysing by x-ray fluorescence for the identification and measurement of lead in air is also known from patent document US 4551848. This device comprises an acquisition mechanism of the particulate matter on belt filter and meter for the amount of lead contained in the particulate matter and thus deposited on the belt. The meter includes a source of x-radiation by means of which the lead deposited on the belt is excited emitting photons that are detected by means of a photomultiplier. The emitted photons are characterised by respective spectra, from the analysis of which it is possible to determine the concentration of lead in the air. This analysis is carried out in an automatic computerised manner.

It is also known to employ techniques of the above-described type for non- automated laboratory measurements, which allow a greater number of chemical species to be detected while being contemporaneously characterised by greater sensitivity and improved control of the measurement parameters.

A methodological standard was developed in the 1990s by US-EPA for the implementation of these measures and their analysis, described in "Compendium Method IO-3.3, Determination of Metals In Ambient Particulate Matter Using X-Ray Fluorescence (XRF) Spectroscopy", U. S. Environmental Protection Agency, Cincinnati, OH 45268, June 1999.

More recently, a method, described in US patent 7539282, was proposed which employs an automatic belt sampling device that includes an automatic calibration mechanism of the measurement and control instrument and of the quality of the measurements.

The above-described prior art has the following main drawbacks:

the quantification of the absolute concentration of the particulate matter is based on a large number of parameters to be measured and employed in the calculations;

storage and cataloguing difficulty of the analysed sample. This also makes repeating the measure difficult or impossible, in particular if measurement of the same sample initially analysed is to be repeated but with different measurement devices or methods.

A further problem of known analysis methods is that the instrument calibration step is generally complex.

Summary

The aim of the present invention is to provide a new method for identifying chemical species in a sample by x-ray fluorescence, which is capable of overcoming the drawbacks reported with reference to the cited prior art.

A further aim is to provide a new method for identifying chemical species in a sample by x-ray fluorescence comprising a simple and reliable calibration step. In accordance with the invention, the aforementioned technical problem is resolved by a method having the characteristics as set out in independent claim 1. In particular, in a first aspect, the invention relates to a method of measuring chemical species in a sample by x-ray fluorescence comprising the steps of:

- arranging a sample including one or more chemical species to be identified and measured,

- exposing said the sample to x-radiation,

- measuring a spectrum associated with said chemical species, - recognising each of said chemical species by identifying the energy value of one or more peaks of said spectrum, a corresponding chemical species being associated with each of said peaks,

- calculating the mass of each of said chemical species by analysing said spectrum, said mass of each of said chemical species being calculated on the basis of the number of counts associated with each of said peaks of said spectrum,

said method further comprising a preliminary calibration step wherein a linear relation is identified between the number of counts associated with each peak of said spectrum and the mass of said corresponding chemical species associated with said peak,

said method being characterised in that in said preliminary calibration step a plurality of calibration solutions are placed on respective rough, impermeable surfaces of respective calibration supports that are identical to each other, each of said calibration solutions comprising a predefined amount of one of said chemical species.

The roughness of the calibration support allows the available surface to be increased and the surface tension of the shiny surface to be reduced, thus allowing a solution to settle on said surface occupying a greater surface with respect to a perfectly smooth surface and minimising the thickness of the solute once the solvent has evaporated.

Other advantages of the present invention are obtained by a method in accordance with the dependent claims, as better explained in the following description.

In particular, at least one sub-step is provided in the calibration step. The roughness of the calibration support is obtained by means of a plurality of micro- impressions obtained by shot peening.

Brief description of the drawings

Further characteristics and advantages of the present invention will become clearer from the following detailed description of a preferred, but non-exclusive embodiment, illustrated by way of a non-limiting example, with reference to the accompanying drawings, wherein: figure 1 is a simplified flow chart of the method according to the present invention,

figures 2 and 3 are two diagrams of two respective spectra measured with the method of the present invention,

- figure 4 is a diagram of a spectrum measured in a calibration step of the method of the present invention,

figure 5 is a diagram of a "white" spectrum measured with the method of the present invention,

- figures 6 and 7 are two diagrams obtained in a calibration step of the method of the present invention;

figure 8 is a top view of a sample analysable with the method of the present invention;

figure 9 is a side view of the detail of figure 8;

figure 10 is an enlarged detail of the detail X of figure 8.

Detailed description of the invention

With reference to the accompanying figure 1 , a method of measuring chemical species in a sample by x-ray fluorescence is globally indicated by 100. The method 100 comprises a first step 1 10 of preparing a sample to be analysed, including one or more chemical species to be identified and measured. The sample to be analysed consists of an analysis support onto which the chemical species to be identified and measured are affixed.

The sample to be analysed is obtained by exposing the analysis support to a flow of air comprising the chemical species to be identified and measured in such a way that they can be affixed onto a surface layer of the analysis support.

According to one possible variant embodiment, the analysis support, per se known and conventional, is shaped like a discoidal wafer, made of a permeable material, for example quartz (Si02). The air flow passes through the analysis support depositing the chemical species to be identified and measured onto a surface layer thereof, having a thickness of between 0.1 μΐτι and 30 μηι.

According to other variant embodiments of the invention, the analysis support has a different shape, being for example constituted by a belt subjected to the flow of the mass of air containing the chemical species to be identified and measured. The method 100 comprises a second step 120 wherein the sample to be analysed is exposed to x-radiation.This radiation is emitted by per se known and conventional devices, for example an x-ray tube Molybdenum anode, and are not thus described in detail

By effect of this exposure, the sample to be analysed emits photons (fluorescence) characterised by a spectrum 200 (figure 3) of the radiation that depends on the chemical species affixed onto the surface of the sample to be analysed.

The method 100 comprises a third step 130 wherein the spectrum 200 is measured. The atomic species that were excited by the x-radiation with consequent photon emission are identifiable from the spectral peaks 200a, b,c (three spectral peaks in the example of figure 2 and 3).

Tools that are per se known and conventional are employed for the measurement of the spectrum 200 and are not thus further described in detail. For example, a Si- PIN diode produced by Amptek Inc., connected to an electronic detection chain consisting of a signal amplifier, an analogue/digital converter and a multichannel analyser MCA, can be employed as photon detector.

A raw spectrum 210 (figure 2) is obtained at the output of said measurement instrumentation, which represents a series of counts on the basis of the channels assigned by the multichannel analyser MCA.

So as to be able to identify the atomic species associated with the spectral peaks 200a, b,c it is first necessary to transform the raw spectrum 210 into the spectrum 200, by converting the values of the channels into the corresponding energy values.

To this end, the method 100 comprises a calibration step 125 that is inclusive of a first sub-step of associating an energy value to each channel of said instrument by measuring and analysing known spectra. In this first sub-step a calibration spectrum 220 (figure 4) is acquire that is determined by excitation of known materials. In the example of figure 4, a sample containing titanium, chromium, nickel, iron, zinc is employed, which respectively determine the five spectral peaks 220a,b,c,d,e. The energy values associated with these peaks being known, it is possible to perform the correlation, by Gaussian fit for example, between channels and energy values, which allows the spectrum 200 to be obtained. The method 100 comprises a fourth step 140 of recognising each of the chemical species by identifying the energy value of one or more peaks in the spectrum measured in the third step 130. Before proceeding with this identification the values measured for each sample must be cleared of ambient background radiation, by removing from the spectrum 200 a so-called "white" spectrum 230, which is obtained in a second sub-step of the calibration step 125 by measuring the white spectrum 230 associated with an analysis support, for example a so- called "virgin" wafer, i.e. not subjected to the flow of the mass of air containing the chemical species to be identified and measured.

The white spectrum is periodically acquired, normally every fifteen measurements of spectra 200 associated with respective samples.

The method 100 comprises a fifth step 150 wherein the mass of each of the chemical species present in the sample to be analysed is calculated by analysing the spectrum obtained by removing the white spectrum from each spectrum 200. The mass of each of the chemical species is calculated on the basis of the number of counts associated with each of the peaks Κ _α , and Ι_ _α , _β of the spectrum obtained from this difference.

After the fifth step 150, the method 100 comprises a sixth step 160 wherein the absolute concentration of the chemical species in this mass of air is determined as the result of the ratio of the mass calculated in the fifth step 150 to the measurement of the mass of air that passed through the measuring instrument during the third measurement step 130 of the spectrum 200.

In order to link the mass of each chemical species to the spectral counts, the method 100 provides, in advance of the first step 110, for a calibration step 101 wherein a ratio of the number of counts associated with each peak to the mass of the chemical species corresponding to the peak is identified.

The calibration step 101 comprises:

a first preparation sub-step 101 a of a calibration support 50;

a second calibration sub-step 101b of the volume of a container employed in the successive third sub-step 101c,

a third preparation sub-step 101c of a calibration solution,

a fourth preparation sub-step 101d of a calibration sample 15, a fifth measurement sub-step 101 e by means of exposure to x-radiation, a sixth, final calibration step 101f of the results of the calibration step. In the first sub-step 101a a thin sheet of impermeable plastic material is shaped so as to provide a calibration support 50 having a discoidal wafer shape.

The sheet material is selected from materials having low atomic number, so as to eliminate the relative fluorescence component during the measurements consequent to excitation by x-radiation. In one variant embodiment of the present invention, the calibration support material 50 exclusively contains carbon, hydrogen, oxygen, hydrogen and nitrogen, being for example constituted by a polyamide film, such as Kapton®.

It is generally required that the calibration support material 50 contain chemical elements having atomic number less than or equal to 14.

Each calibration support 50 is provided with a rough surface 55. In order to obtain the surface 55, micro-impressions with average diameter not greater than 10 μΐη and having a depth not greater than 5 microns are thus produced on the calibration support 50. The roughness generated on the calibration supports 50 allows the available surface area to be increased and the tension with respect to the original plastic sheet to be reduced. The micro-impressions are, for example, generatable by shot peening, i.e. high speed impact of plastic microspheres having an average size of 200 μιτι or by another method that allows the same result to be obtained. The calibration support 50 is subsequently washed in the ultrasound bath filled with demineralised water.

The first sub-step 101a is repeated several times in order to obtain a plurality of calibration supports 50 that are identical to each other.

In the second sub-step 101 b, the volume of a container for liquids, for example constituted by 3.0 pL pipette, is accurately determined by performing a plurality of repeated measurements and subsequently carrying out the statistical analysis of these measurements.

In the third sub-step 101c, a predefined amount of the chemical species for which the calibration is to be carried out is prepared by weighing a salt containing the chemical species with a micro-balance and dissolving it in a solvent, for example water or alcohol, until the saturated solution condition is obtained. In the fourth sub-step 101d, the saturated solution is diluted to different concentrations generating a plurality of calibration solutions having different and known concentration. The calibration solutions are deposited, by means of the container which volume was accurately determined in the second sub-step 101 b, onto a plurality of respective calibration supports 50 prepared in the first sub-step 101a, constituting a plurality of respective calibration samples 15. The roughness of the calibration supports 50 allows the calibration solutions to settle on the surface 55 occupying a greater surface with respect to a perfectly smooth surface and minimising the thickness of the solute once the solvent has evaporated. The thickness of the solute remaining on the surface 55 is between 0.1 μιη and 30 μΐη, being thus equal to the thickness of the layer of the chemical species on the analysis support employed in the first step 1 10 of the method 100. The impermeable nature of the calibration supports 50 allows the solute not to penetrate the calibration supports.

The calibration samples 15 are subsequently dried by irradiation to obtain evaporation of the solvent.

The calibration solutions being of different concentrations, the chemical species deposited on each calibration support will be different and known. The mass of the chemical species present on the surface of each calibration sample 15 will consequently also be known.

In the fifth sub-step 101e, each calibration sample 15 obtained in the previous sub- step 101d is exposed to x-radiation. By effect of this exposure, each sample 15 emits photons (fluorescence) characterised by a respective spectrum 300, which is acquired and subsequently analysed in the successive sixth sub-step 101f.

In the final sub-step 101f, the spectrum 300 is analysed to associate the number of counts relating to visible fluorescence peaks to the value of the mass of material deposited on the calibration sample 15.

The chemical species employed in the calibration step 101 and the relative fluorescence peaks are those indicated in table 1.

A fit procedure is then performed for the spectrum 300 (figure 6), by means of a "Savitzky-Golay" algorithm for example, to obtain a curve 301 wherein the fluctuations are attenuated and, contemporaneously, the calculation of the so- called "Continuum", i.e. a numerical approximation 302 of the spectrum obtained as the difference between the 300 and its peaks.

The most visible fluorescence peak of the spectrum having been selected, the number of N _tot counts in a predetermined time is measured (for example 1 hour), by subtracting the value Ni of the curve 301 from the value N ₂ of the Continuum 302 at this peak, according to the following formula:

The count frequency υ is then calculated according to the formula: where t _a is the acquisition time of the spectrum 300,

It is observed (figure 7) that the values of υ obtained for a plurality of calibration samples 15, onto which different amounts of the same chemical species were respectively deposited, linearly depend on the value of the mass of the above- mentioned chemical species present in the calibration sample 5.

The two respective intercept and slope values that characterise the linear regression 400 (figure 7) which links mass and counts for each of the chemical species in table 1 are thus identified in sub-step 101f, according to known statistical methods.

The described technical solutions allow the predetermined object and aims to be achieved with reference to the cited prior art, achieving as a further advantage the fact of making the calculation of the absolute mass of the chemical species analysed according to the present method particularly fast and easily automated. The method of the present invention is particularly suitable for performing automatic analyses in the field.