Method of analyzing spectral peaks

Abstract

A method of analyzing an interference peak of a sample spectrum generated using a detector of an optical spectrometer is provided. The interference peak is produced by a plurality of spectral emissions of different wavelengths. The method includes generating, based on a model of expected curve parameters of the optical spectrometer and a location of the sample peak on the detector of the optical spectrometer, interference curve parameters representing a shape of each spectral emission in the interference peak. The method further includes fitting a plurality of curves to the interference peak, each curve corresponding to one of the plurality of spectral emissions of different wavelengths that form the interference peak, wherein each curve is fitted using the interference curve parameters provided by the model of expected peak parameters. The plurality of curves are output for further analysis.

Description

Technical Field

[0001] This disclosure relates to methods for analyzing spectral peaks. More specifically, this disclosure relates to methods for analyzing spectral peaks generated using an optical spectrometer. Background Technology

[0002] Optical spectroscopy, such as optical emission spectroscopy, is an analytical technique used to analyze samples. In optical emission spectroscopy, a plasma source can be used, for example, to excite the sample. The excited atoms in the sample emit light, the wavelength of which is a characteristic of the atoms present in the sample. Thus, the light emitted by the sample comprises multiple spectral lines, each corresponding to a specific energy level transition in the atom. By detecting the presence of spectral lines at specific wavelengths, the presence of elements in the sample can be determined. Furthermore, the intensity of each spectral emission can be used to analyze the concentration of elements within the sample.

[0003] Therefore, optical spectrometers, such as optical emission spectrometers, can generate multiple spectral peaks from a sample. Part of the process of analyzing multiple spectral peaks involves identifying the peaks from the measurement data. The identification process typically involves fitting curves to the measurement data to identify the peak positions (and associated wavelengths) and peak intensities. Peak wavelengths and peak intensities can be used to determine the elements present in the sample and the relative amounts of each element.

[0004] In optical spectroscopy, the spectral peaks generated from a sample may contain two or more peaks with similar wavelengths. Therefore, when these peaks are imaged on the detector, they may overlap. In some cases, this overlap occurs because the optical spectrometer lacks sufficient resolution to distinguish between the two peaks. Due to interference between overlapping peaks, misidentification can occur.

[0005] When overlapping spectral peaks are present, users can choose not to use them for further analysis. Discarding overlapping peaks will increase the time required to analyze the sample and will require user input to view the overlapping peaks.

[0006] Alternatively, inter-elemental correction algorithms can be applied to resolve overlapping peaks. Depending on the measurements, inter-elemental correction may not be feasible and requires additional user effort. The effectiveness of inter-elemental correction algorithms also depends on the resolution of the optical spectrometer. Therefore, inter-elemental correction algorithms may not be suitable for resolving some overlapping peaks.

[0007] Therefore, this disclosure seeks to provide a method for analyzing spectral peaks that solves at least one problem associated with prior art methods, or at least provides a commercially useful alternative to them. Summary of the Invention

[0008] According to a first aspect, a method is provided for analyzing interference peaks in a sample spectrum generated using a detector of an optical spectrometer. The interference peaks are generated by multiple spectral emissions at different wavelengths. The method includes generating interference curve parameters representing the peak shape of each spectral emission in the interference peak based on a model of expected curve parameters of the optical spectrometer and the position of the sample peak on the detector of the optical spectrometer. The method also includes fitting multiple curves to the interference peak, each curve corresponding to one of the multiple spectral emissions at different wavelengths that form the interference peak, wherein each curve is fitted using interference curve parameters provided by the model of the expected peak parameters. Multiple curves are output for further analysis.

[0009] According to the method of the first aspect, sample spectra generated using an optical spectrometer, such as an optical emission spectrometer, can be analyzed. Specifically, the method of the first aspect analyzes interference peaks generated by spectral emissions from at least two different wavelengths. The method of the first aspect provides a method for generating interference curve parameters to fit multiple curves to the interference peak. The interference curve parameters can then be output for further analysis. Thus, the method of the first aspect allows characterizing the different spectral emissions from the interference peak (e.g., the intensity and peak wavelength of each spectral emission), making information from the interference peak available for further analysis. In other words, the method of the first aspect allows users to analyze a large proportion of sample spectra using a simple and efficient analytical procedure.

[0010] Specifically, the first aspect of the method generates a curve representing the peak shape of each spectral emission that forms the interfering peak. This type of curve can be used to perform baseline correction on each spectral emission, and subsequently, the concentration of the element associated with the spectral emission can be calculated based on the area under the generated curve. In other words, the method according to the first aspect allows analysis of interfering peaks to determine the concentration of the element contributing to the interfering peak.

[0011] To analyze interfering peaks, the first aspect of the method involves fitting a curve to the interfering peak. This invention recognizes that the contribution of each different spectral emission to the overall shape of the interfering peak depends at least in part on the optical aberrations introduced by the optical spectrometer due to the detector and associated optics. The degree of optical aberration depends on the location on the detector where the interfering peak is detected. The variability of optical aberrations makes accurately fitting a curve to the interfering peak challenging. That is, due to the nature of the optical aberrations introduced by the optical spectrometer, assuming that each spectral emission on the detector has the same peak shape leads to inaccurate analysis. Similarly, spectral emissions of the same wavelength can be imaged at multiple locations on the detector. Therefore, the wavelength-dependent assumption of peak shape can also lead to similar inaccuracies.

[0012] To accurately fit the curve to the interference peak, the first aspect of the method provides a model of the expected curve parameters of the optical spectrometer. This model provides interference curve parameters representing the peak shape of the spectral emission at a specified detector location. Therefore, the interference curve parameters can be used to model the peak shape of a single spectral emission at a detector location (corresponding to a known wavelength). For an interference peak, it can be assumed that the multiple spectral emissions constituting the interference peak have similar degrees of optical aberration. For example, spectral emissions forming the interference peak can be detected at similar locations on the detector (e.g., within 20 pixels of each other). Therefore, it can be assumed that the spectral emissions forming the interference peak have the same peak shape provided by the model of the expected curve parameters (i.e., represented by the same interference curve parameters).

[0013] In some embodiments, the interference curve parameters for each spectral emission include a first asymmetric interference curve parameter and a second asymmetric interference curve parameter, so as to fit multiple asymmetric curves to the interference peak. By using asymmetric curves, the model of the expected curve parameters can more accurately explain the peak shape changes caused by variations in the degree of optical aberrations on the detector.

[0014] In some embodiments, the first interference curve parameter and the second interference curve parameter are respectively the first double Gaussian curve parameter and the second double Gaussian curve parameter. The inventors have discovered that the peak-shaped double Gaussian model is very suitable for explaining the variation of optical aberrations of optical spectrometers with wavelength.

[0015] In some embodiments, fitting each of the multiple curves to the interfering peak includes selecting a peak wavelength and peak intensity for each of the curves to be fitted to the interfering peak. Thus, the method of the first aspect can be optimized to select a peak wavelength and peak intensity for each of the curves to be fitted to the interfering peak.

[0016] In some embodiments, the model for the expected curve parameters of the optical spectrometer includes a model that defines the relationship between the detector location of the interfering peak and the parameters of the interfering curve. In some embodiments, the detector location of the interfering peak may be represented by the number of pixels.

[0017] In some embodiments, the model for the expected curve parameters is generated based on the peak shapes and associated detector positions of multiple spectral peaks in a calibration spectrum generated from a calibration sample using an optical spectrometer. By forming a model for the expected curve parameters based on the calibration spectrum for which data is to be analyzed, the model can accurately reflect the optical aberrations of the optical spectrometer. In other embodiments, the model for the expected curve parameters can be provided as a predefined model. The predefined model can account for known optical aberrations introduced by the optics of the optical spectrometer.

[0018] In some embodiments, the method further includes determining how many curves to fit to the interfering peak. In some embodiments, the number of curves to be fitted to the interfering peak is determined based on the first and / or second derivatives of the interfering peak. Thus, the first and / or second derivatives of the sample peak can be used to determine the number of different spectral emissions that generate the interfering peak. Knowing the number of spectral emissions and therefore the number of curves to be fitted to the interfering peak allows the method to fit the curves to the interfering peak more accurately.

[0019] In some embodiments, the peak wavelengths of each curve to be fitted to the interfering peak are selected from a database of known spectral emissions. By selecting peak wavelengths for each curve to be fitted to the interfering peak using known spectral emission wavelengths, the method can fit the curve to the interfering peak without prior knowledge of the number of spectral emissions present in the peak. Specifically, this type of fitting method may be particularly suitable for interfering peaks where distinguishing the number of different spectral emissions present in these peaks is challenging.

[0020] In some embodiments, a regression-based algorithm is used to fit multiple curves to the interference peak.

[0021] In some embodiments, the sample spectrum is generated by an optical spectrometer using an echelle grating to diffract the sample light onto a two-dimensional detector. Therefore, the model of the expected curve parameters can reflect variations in optical aberrations on the two-dimensional detector. In some embodiments, it should be understood that light of the same wavelength can be diffracted to multiple locations on the detector. Therefore, the model of the expected curve parameters uses detector location rather than wavelength to more accurately predict the peak shape of each spectral emission that forms interfering peaks.

[0022] In some embodiments, the detector of the optical spectrometer is an array detector. That is, the detector includes a plurality of detection elements (e.g., pixels of a charge-coupled device) arranged in an array, wherein each detection element is configured to detect the intensity of a portion of the sample spectrum incident on the detection element. The array can be a one-dimensional array or a two-dimensional array. Thus, the position of the interfering peak on the detector (detector position) can be represented by the number of pixels or a set of coordinates. In other words, the interfering peak can be spatially distributed across a region of the detector.

[0023] In some embodiments, the optical spectrometer may be an atomic emission spectrometer. Specifically, the optical spectrometer may be an optical emission spectrometer, along with a spectrometer controller. The method of the first aspect can also be applied to other types of optical spectrometers (and associated controllers), such as X-ray fluorescence spectroscopy systems, laser-induced breakdown spectroscopy systems, or optical absorption spectrometers. Therefore, it should be understood that the method of the first aspect is applicable to any form of spectroscopic determination in which interfering peaks exist in the sample spectrum.

[0024] According to a second aspect of this disclosure, a controller is provided for analyzing interference peaks in a sample spectrum generated using a detector of an optical spectrometer. The controller is configured to receive interference peaks from the optical spectrometer, the interference peaks being generated by emission of multiple spectra at different wavelengths. The controller is further configured to:

[0025] Based on the model of the expected curve parameters of the optical spectrometer and the position of the interfering peak on the detector of the optical spectrometer, interference curve parameters representing the peak shape of each spectrum emitted in the interfering peak are generated.

[0026] Multiple curves are fitted to the interference peak, each curve corresponding to one of multiple spectral emissions at different wavelengths that form the interference peak. Each curve is fitted using interference curve parameters provided by a model of the expected peak parameters; and

[0027] Output multiple curves for further analysis.

[0028] Therefore, the controller of the second aspect can be used to perform the method of the first aspect of this disclosure. It should therefore be understood that the controller of the second aspect can incorporate any optional features and associated advantages of the first aspect discussed above.

[0029] The controller for the second aspect can be provided using a controller for an optical emission spectrometer. In some embodiments, the controller may include a processor, a microprocessor, etc. In some embodiments, the controller may be a controller for an optical emission spectrometer or an optical emission spectrometer system.

[0030] According to a third aspect of this disclosure, an optical spectroscopy measurement system is provided. The optical spectroscopy measurement system includes an optical spectrometer configured to generate a sample spectrum from a sample using a detector of the optical spectrometer, and a controller configured to analyze interfering peaks in the sample spectrum, said interfering peaks being generated by multiple spectral emissions of different wavelengths. The controller is configured to:

[0031] Based on the model of the expected curve parameters of the optical spectrometer and the position of the interfering peak on the detector of the optical spectrometer, interference curve parameters representing the peak shape of each spectrum emitted in the interfering peak are generated.

[0032] Multiple curves are fitted to the interference peak, each curve corresponding to one of multiple spectral emissions at different wavelengths that form the interference peak. Each curve is fitted using interference curve parameters provided by a model of the expected peak parameters; and

[0033] Output multiple curves for further analysis.

[0034] Therefore, the optical spectroscopy system may include the controller of the second aspect. The optical spectroscopy system may be configured to perform the method of the first aspect. Therefore, it should be understood that the optical spectroscopy system of the third aspect may combine any optional features and associated advantages of the first or second aspects discussed above.

[0035] In some embodiments, the optical spectroscopy system may be an optical emission spectroscopy system including an optical emission spectrometer.

[0036] According to a fourth aspect of this disclosure, a computer program is provided that includes instructions for causing the optical spectroscopy system of the third aspect to perform the steps of the method of the first aspect. Therefore, it should be understood that the computer program of the fourth aspect can incorporate any optional features and associated advantages of the first, second, or third aspects discussed above.

[0037] According to a fifth aspect of this disclosure, a computer-readable medium is provided having a computer program of the fourth aspect stored thereon. Therefore, it should be understood that the computer-readable medium of the fifth aspect may incorporate any optional features and associated advantages of the first, second, third, or fourth aspects discussed above. Attached Figure Description

[0038] This invention can be practiced in many ways, and specific embodiments will now be described by way of example only and with reference to the accompanying drawings, in which:

[0039] - Figure 1 A schematic diagram of an optical spectroscopy measurement system according to an embodiment of the present disclosure is shown;

[0040] - Figure 2 A schematic diagram of the detector of an optical spectroscopy measurement system according to an embodiment of the present disclosure is shown;

[0041] - Figure 3 A block diagram illustrating a method for analyzing spectral peaks of a sample spectrum according to an embodiment of the present disclosure;

[0042] - Figure 4 An example of an interfering peak generated by two spectral emissions is shown, where the peak amplitude of the interfering peak does not correspond to the peak wavelength of the two fundamental spectral emissions (in terms of wavelength);

[0043] - Figure 5 A diagram illustrating the double Gaussian distribution is provided.

[0044] - Figure 6 An image of a calibration spectrum recorded by an optical spectroscopy system according to an embodiment of the present disclosure is shown;

[0045] - Figure 7 Show Figure 6A graph showing the calibrated spectrum region of the image;

[0046] - Figure 8 An example of interference peaks generated by three spectral emissions is shown;

[0047] - Figure 9 Another example of a method for analyzing interference peaks according to embodiments of the present disclosure is shown; and

[0048] - Figure 10 Showing the fit to Figure 9 Multiple curves of interference curves. Detailed Implementation

[0049] According to embodiments of the present disclosure, an optical spectroscopy system 10 is provided. The optical spectroscopy system 10 is configured to perform an optical spectroscopy method on a sample to generate a sample spectrum. The optical spectroscopy system 10 can also analyze sample peaks in the sample spectrum according to the methods of the present disclosure. Figure 1 A schematic diagram of the optical spectroscopy measurement system 10 is shown. (As shown) Figure 1 As shown, the optical spectroscopy system 10 includes a light source 11, an optical arrangement 12, a detector 13, a processor (μP) 14, a memory 15, and an input / output (I / O) unit 16.

[0050] exist Figure 1 In this embodiment, the light source 11 is a plasma source, such as an inductively coupled plasma (ICP) source. Therefore, Figure 1 The optical spectroscopy system 10 can be an optical emission spectroscopy system 10. In other embodiments, the light source 11 can be a furnace or any other high-temperature light source that generates an excited state of matter suitable for optical emission spectroscopy. Alternatively, other optical spectroscopy systems 10 can provide a light source 11 suitable for the optical spectroscopy method being performed. The light source 11 can be configured to receive a sample to be analyzed using the optical spectroscopy system 10. For example, in the case where the light source 11 is a plasma source, the sample can be introduced into the plasma, where the sample interacts with the plasma. Samples in aqueous solution form can be introduced directly into the plasma source, while solid samples can be introduced using, for example, laser ablation or vaporization.

[0051] exist Figure 1 In some embodiments, the optical arrangement 12 may include an echelle grating and a prism (and / or additional gratings) to generate a two-dimensional image of the light produced by the light source 11 (and a sample, if present). The two-dimensional image is formed on the detector 13. In such arrangements, it should be understood that the optical arrangement 12 is configured to direct radiation from the light source 11 to the detector such that the radiation is suitable for detection by the detector 13.

[0052] exist Figure 1In some embodiments, detector 13 may be a CCD (charge-coupled device) array. A typical CCD array may have at least approximately 1024 × 1024 pixels (1 megapixel). The CCD array may be arranged to generate spectral intensity values ​​corresponding to the measured light quantity of a mid-echelon grating spectrum and to transmit the spectral values ​​to processor 14. Thus, detector 13 may be a multi-channel detector configured to detect multiple different wavelengths. Detector 13 (as in...) Figure 1 In one embodiment, detector 13 can be configured to detect a two-dimensional spectrum. In other embodiments, detector 13 can be a CMOS or CID detector.

[0053] Processor 14 (controller) may include a commercially available microprocessor or the like. Memory 15 may be a suitable semiconductor memory and may be used to store instructions that allow processor 14 to execute embodiments of the methods according to this disclosure. Processor 14 and memory 15 may be configured to control optical spectrometry system 10 to execute the methods according to embodiments of this disclosure. Thus, memory 15 may include instructions that, when executed by processor 14, cause optical spectrometry system 10 to execute the methods according to embodiments of this disclosure.

[0054] The optical spectroscopy system 10 can be configured to generate a sample spectrum by introducing a sample into a light source 11. The light generated by the light source 11 interacts with the sample, where spectral emission, a characteristic of the sample, is emitted by the sample. The spectral emission from the light source 11 and the sample is guided to a detector 13 by an optical arrangement 12. An echelle grating in the optical arrangement 12 diffracts spectral emission of different wavelengths in varying amounts, such that peaks associated with each spectral emission are detected at different locations on the detector 13.

[0055] Figure 2 Show Figure 1 A schematic diagram of the two-dimensional detector 13 in an embodiment. Figure 2 The two-dimensional detector 13 is formed by a pixel array, but each pixel is not in Figure 2 It is displayed separately in the text. Figure 2 A schematic representation (dashed lines) of stages 20 containing light diffracted by an echelle grating and a prism and imaged on detector 13. Each stage 20 corresponds to a different wavelength range, and the wavelength varies laterally along each stage. For example, in Figure 2 In this embodiment, the wavelength of the diffracted light can increase from left to right along each order. The starting wavelength can also increase from order a) to order i). Figure 2Four detailed views of example single-spectral emissions are also shown, which are imaged by pixel groups of the detector at different locations on the detector. It should be understood that the peak shape of each spectral emission differs based on the optical aberrations of the optical arrangement 12. In some embodiments, the optical arrangement 12 may cause light of a certain wavelength to be diffracted to a single location or multiple locations on the detector 13. Thus, in some embodiments, spectral emissions may appear at multiple locations on the detector 13.

[0056] When two spectral emissions have similar wavelengths, the peaks associated with each emission can be directed to similar regions of detector 13. When the regions of detector 13 where two spectral emissions are directed such that at least a portion of one emission peak overlaps with another peak of the other emission, resolving each peak individually can be challenging. These peaks are referred to as interference peaks.

[0057] Detector 13 is configured to output the intensity of each pixel recorded by detector 13 to processor 14 for further analysis.

[0058] Next, we will refer to Figure 3 Methods for describing and analyzing spectral peaks in sample spectra 100. Figure 3 A block diagram of method 100 is shown. Method 100 can be executed by the processor of optical spectrometry system 10. Alternatively, method 100 can be executed by any other processor that provides a sample spectrum generated by optical spectrometry system 10.

[0059] In step 102 of method 100, processor 14 determines whether a sample peak in the sample spectrum is an interfering peak. The sample spectrum may include multiple peaks generated by the spectral emission of the optical spectroscopy system 10. An interfering peak is the result of two or more spectral emissions incident on the same region of the detector. That is, peaks from two or more spectral emissions fall within approximately 20 pixels of each other (on the detector), such that at least a portion of the peak associated with each spectral emission overlaps with one or more other peaks of other spectral emissions.

[0060] Figure 4 Examples of interference peaks that can be analyzed according to embodiments of the present disclosure are shown. Figure 4 An example of an interfering peak generated by two fundamental spectral emissions is shown, wherein the peak amplitude of the interfering peak does not correspond to the peak wavelength of the two fundamental spectral emissions (in terms of wavelength). The peak intensity of the interfering peak also differs from the peak intensity of the two fundamental spectral emissions. Therefore, the fundamental spectral emissions may have significantly different shapes from the interfering peak. The method according to this disclosure aims to provide a robust method for analyzing interfering peaks, capable of analyzing a variety of different interfering peak shapes.

[0061] Figure 4It is a graph of the intensity values ​​measured along a level by the pixels of the detector. Figure 4 The measurements shown have been interpolated. Interpolation can include cubic spline interpolation, preferably cubic Hermite interpolation or Gaussian interpolation. In some embodiments, bicubic spline interpolation can be used. The cubic spline interpolation curve is also... Figure 4 As shown in [the image]. Although in [the image] Figure 4 Spline interpolation was used in the example, but it should be understood that other interpolation methods known to those skilled in the art may also be appropriate.

[0062] In method 100, the first derivative of the sample spectrum can be analyzed to... Figure 4 The interference peak shown is distinguishable from other peaks in the sample spectrum. Figure 4 In the example, the first derivative of the interpolated spectrum was analyzed. Sample peaks in a sample spectrum generated by a single spectral emission can be distinguished from interfering peaks based on, for example, the number of inflection points (i.e., fixed points) within a specified range of the detector (e.g., 15, 20, or 25 pixels, but the actual number of pixels within the specified range will depend on the pixel size and the optical characteristics of the optical spectrometer system). Based on the number of inflection points present, the method can determine the number of different spectral emissions that form interfering peaks. In some embodiments, the second derivative of the sample spectrum can also be analyzed to determine whether a peak in the sample spectrum is an interfering peak. Figure 4 In one embodiment, the second derivative of the interpolated peak, along with its first derivative, can be analyzed to determine whether the peak is an interference peak. The determination of interference peaks can be based on the number of peaks (extremes) indicated by the second derivative and whether the peaks (extremes) are positive or negative.

[0063] For each distinct spectral emission present in the interfering peak, method 100 attempts to fit a curve to the interfering peak representing the spectral emission. Thus, if the first derivative (and optionally the second derivative) of the sample spectrum indicates that two distinct spectral emissions are present in the interfering peak, method 100 then fits two curves to the interfering peak.

[0064] In other embodiments of this disclosure, the method may proceed to fitting a curve to the interfering peak using an iterative process. Therefore, the method according to this disclosure does not require determining the amount of spectral emission present in the interfering peak prior to analysis.

[0065] If an interfering peak is detected, method 100 proceeds to step 104, where an interfering curve parameter is generated for the interfering peak using a model of the expected curve parameters. The model of the expected curve parameters generates the interfering curve parameters, which define the shape (e.g., deviation) of each curve to be fitted to the interfering peak. The invention recognizes that detector position-dependent optical aberrations introduced, for example, by the optical device arrangement 14, cause the peak shape of the spectral peak to vary depending on the detector position of the interfering peak to be analyzed. Therefore, to improve the accuracy of curve fitting, the model of the expected curve parameters provides the interfering curve parameters, which account for the detector position-dependent optical aberrations of the optical spectrometry system 10.

[0066] exist Figure 4 In this embodiment, the model for the expected curve parameters assumes that each spectral emission detected by the optical spectrometer 10 has a double Gaussian peak shape. An example of a double Gaussian peak is... Figure 5 As shown in the image. Figure 5 As shown, the double Gaussian peak can be defined by the following equations 1) and 2):

[0067]

[0068]

[0069] In equations 1) and 2) above, y is the intensity of spectral emission, y0 is the background intensity, H is the amplitude of the spectral peak, and x c The detector position (corresponding to the wavelength) is where the peak's maximum value occurs. The interference curve parameters w1 and w2 define the deviation of the double Gaussian distribution on either side of the peak. Therefore, the interference curve parameters w1 and w2 determine the shape of the double Gaussian peak.

[0070] As noted above, the optical arrangement 12 introduces optical aberrations that affect the shape of each spectral peak detected by the detector 13. For Figure 1 In this embodiment, optical aberrations introduced by the optical arrangement 12 cause variations in the shape of the spectral peaks on the two-dimensional detector 13. Therefore, a model of the expected curve parameters can define the values ​​of the interference curve parameters w1 and w2 for a range of different detector locations (with associated wavelengths). For example, in the case where the detector is a two-dimensional detector, the model of the expected curve parameters can define the interference curve parameters w1 and w2 for each pixel (or pixel region) of the detector. Thus, the model of the expected curve parameters can provide the wavelength (x) with the interference peak (e.g., the wavelength at the point of maximum intensity of the interference peak) and return the interference curve parameters w1 and w2 for fitting a curve to the interference peak.

[0071] As discussed above, the model of the expected curve parameters provides interference curve parameters w1 and w2, which can be used to fit the double Gaussian curve to the interference peak. The interference curve parameters vary with wavelength / detector position to account for optical aberrations present in the optical spectrometry system 10. To more accurately account for optical aberrations, in some embodiments, the model of the expected curve parameters can be generated based on the peak shapes of multiple spectral peaks in the calibration spectrum generated from the calibration sample using an optical spectrometer.

[0072] The calibration spectrum can be generated by analyzing the calibration solution using the optical spectrometer system 10. The calibration solution may include one or more known elements with known spectral emissions. In some embodiments, the calibration solution may be selected to provide multiple non-interfering peaks well distributed across the detector. For example, the calibration solution may include one or more transition metal elements that produce multiple spectral emissions across a wide wavelength range.

[0073] Figure 6 An image of the calibration spectrum generated using detector 13 of the optical spectroscopy system 10 is shown. Figure 2 The diagram shown is similar, comprising orders of light diffracted by a mid-echelon grating and prisms imaged on a detector. Each order corresponds to a different wavelength range, and the wavelength varies laterally along each order. Figure 6 The detailed view shown is a magnified portion of the image. Along a level n in the magnified view, several non-interference peaks are identified and labeled 1, 2, 3, 4, 5, 6, and 7.

[0074] Figure 7 Showing pixels along Figure 6 The graph depicts the intensity values ​​of the level records. Figure 7 The peaks in the text are marked with numbers to indicate... Figure 6 The corresponding intensity peaks are shown in the image. The peak labeled 2 is not shown. Figure 7 It is shown in full because its intensity amplitude is greater than Figure 7 The other peaks shown are much larger. When the calibration solution contains one or more known elements, a list of expected non-interfering peaks can be provided, allowing selection / identification of non-interfering peaks from the calibration spectrum. Alternatively, the processor can identify non-interfering peaks by analyzing the first and second derivatives of the calibration spectrum.

[0075] like Figure 7As shown, multiple calibration peaks are depicted as a continuous spectrum, as indicated by the continuous line. The processor can fit a double Gaussian curve to each non-interference peak using a regression-based algorithm (e.g., minimizing the squared error). The processor can then record the shape parameters of each calibration peak (in this example, the fitted curve is a double Gaussian curve, and therefore the shape parameters are w1 and w2) and the corresponding detector position of the calibration peak. Thus, models of the expected curve parameters, independent of the peak intensities in the calibration spectrum, can be generated for different detector positions.

[0076] Therefore, processor 14 generates a model of the expected curve parameters based on the peak shapes of multiple spectral peaks in the calibration spectrum generated from the calibration sample using an optical spectrometer. That is, processor 14 calculates the w1 and w2 parameters for each non-interference peak in the calibration spectrum and records the detector position associated with each non-interference peak. Calibration information can be recorded in a database, lookup table, etc. Thus, a database can be generated that records the expected curve parameters w1 and w2 for a given detector position based on the non-interference peaks of the calibration spectrum.

[0077] It should be understood that calibration spectra do not provide expected curve parameters for every pixel value. Therefore, in some embodiments, a database can be used to find the nearest pixel for which calibration information has been recorded (the nearest calibration pixel). Alternatively, interpolation can be used to generate values ​​for w1 and w2 for each pixel. Preferably, the model of the expected curve parameters includes calibration peaks within a distance of no more than 100 pixels from the detector location of interest. More preferably, the model of the expected curve parameters should provide calibration peaks within a distance of no more than 70, 50, 30, 25, or 20 pixels from the detector location of interest. Thus, the model of the expected curve parameters can provide an accurate estimate of the optical aberrations introduced by the optical spectrometry system 10 (through interpolation or the nearest calibration pixel), thereby allowing analysis of interfering peaks.

[0078] Therefore, when processor 14 analyzes the interference peak, the model of the expected curve parameters provides interference curve parameters w1 and w2 based on the detector position or wavelength of the interference peak. For example, interference curve parameters w1 and w2 can be selected based on the peak wavelength (the pixel with the highest intensity) of the interference peak. Alternatively, the interference curve parameters can be selected based on the center pixel of the interference peak. Since the detector detects multiple curves to be fitted to the interference peak at similar locations (i.e., within, for example, about 20 pixels apart), it can be assumed that the peaks associated with each of the multiple spectral emissions have the same peak shape. Thus, the same interference curve parameters w1 and w2 can be used for each curve to be fitted to the interference peak.

[0079] When fitting a double Gaussian curve to the interference peak, the model of the expected curve parameters provides parameters w1 and w2 to the processor based on the wavelength of the interference peak. The method then moves to step 106, where the processor 14 fits multiple curves to the interference peak. The number of curves to be fitted corresponds to the number of spectral emissions detected by the processor when analyzing whether the sample curve is an interference curve in step 102.

[0080] In step 106 of method 100, each curve to be fitted to the interference peak is a double Gaussian curve with the general shape discussed above in equations 1) and 2 (see also...). Figure 7 ).

[0081] When the curve is fitted to the interfering peak, the background (y0) can be removed (assuming the optical spectrometer system 10 is properly calibrated). Parameters w1 and w2 are provided by the model. Therefore, the processor 14 is designed to process the remaining parameters (H and x). c The interference peaks of each spectral emission are fitted to the existing spectrum.

[0082] Various methods can be used to fit curves to interference peaks. For example, regression-based fitting algorithms can be used to fit multiple curves to interference peaks. Figure 4 The curves shown are fitted using a least-squares minimization algorithm. Therefore, the fitting algorithm aims to select H and x for each curve to be fitted. c The value of is determined in order to minimize the squared error between the sum of the fitted curves and the interference peaks.

[0083] exist Figure 4 In this example, processor 14 has previously determined that two curves should be fitted to the interfering peak, each curve corresponding to a different spectral emission. Therefore, processor 14 attempts to fit the two curves to the interpolated interfering curve. Fitting the curves to the interpolated interfering curve can further improve the accuracy of the curve fitting process. The processor retrieves the interfering curve parameters from the model of the expected curve parameters based on the detector position of the interfering peak. Figure 4 In one embodiment, the detector location of the interference peak can be determined based on the pixel with the highest intensity. In other embodiments, the pixel corresponding to the center of the interference peak can be used as the detector location of the interference peak. Therefore, in Figure 4 In the example, the model of the expected curve parameters generates interference curve parameters w1 = 1.43434 and w2 = 1.23361 to fit the curve to the interference peak.

[0084] like Figure 4 As shown, the first and second curves are fitted to the interference peak. Figure 4As shown, each of the first and second curves has a double Gaussian shape. As explained above, the same interference curve parameters w1 = 1.43434 and w2 = 1.23361 are used for both the first and second curves. The fitting algorithm fits where x... c The first curve with H = 17.34896 and H = 1866.9748, and x c The second curve has H = 12.58096 and H = 117.54216. In Figure 6 In the diagram, the horizontal axis represents the nominal number of pixels of the detector. It should be understood that the number of pixels represents the wavelength, and therefore the peak wavelength associated with each of the first and second curves can then be calculated by the processor 14.

[0085] Figure 4 The cumulative plot of the first and second curves is also shown. It should be understood that the cumulative plot provides an accurate fit to the original data, indicating that the first and second curves are accurate representations of the two spectral lines that form the interfering peaks. For example, as... Figure 4 As shown, the adjusted R of the curve fit 2 The value is 0.99943, indicating that the two curves accurately fit the original data.

[0086] Figure 8 This illustrates another example of interference peaks generated by emissions from three different spectra. For example... Figure 8 As shown, the interference peak was Figure 8 Pixels 10 to 25 (approximately) are detected. Since the interference peak extends across less than approximately 20 pixels, it can be assumed that the spectral emission has approximately the same optical aberrations, and therefore the same interference curve parameters w1 = 1.43434 and w2 = 1.22261 can be used for each of the three curves to be fitted. According to embodiments of this disclosure, the interference curve parameters are provided by a model of the expected curve parameters based on the wavelength (pixel position) of the interference peak. Figure 8 As shown, the first, second, and third curves are fitted to the interference peaks, with each curve corresponding to the spectral emission present in the interference peak. The first, second, and third curves can be used as described above regarding... Figure 6 The regression-based algorithm discussed is used for fitting. For example... Figure 8 As shown, the fitting algorithm fits x. c The first curve with x = 17.6344 and H = 2528.6722, where x c The second curve with H = 14.3721 and H = 853.04439, and the x in it c The third curve with H = 21.84255 and H = 145.94758. Adjusted R² for curve fitting. 2 The value is 0.99948, indicating that the three curves accurately fit the original data.

[0087] Once the curve is fitted to the data, in step 108, method 100 can output the fitted curve for further analysis. For example, parameters H and x can be output. c This allows for the identification and quantification of individual spectral peaks of interfering peaks.

[0088] Figure 4-7 and Figure 8 Examples of methods for analyzing interference peaks are provided, wherein the number of interference spectral emissions is known or determined prior to fitting analysis. In some embodiments of this disclosure, the method can iteratively fit a curve to the interference peak.

[0089] Figure 9 The graph shows a spectral peak detected by the optical spectrometer system 10. At first glance, the spectral peak does seem to possess any characteristics that indicate it is an interfering peak, such as inflection points (fixed points). However, when attempting to generate curves using the expected curve parameters corresponding to the detector position to analyze the spectral peak, it can be seen that the shape of the spectral peak differs from the expected shape of the non-interfering peak. That is, the peak is actually an interfering peak, where two or more peaks emitted by the interfering spectrum have not yet been resolved by the optical spectrometer system 10. In this case, it may be difficult to determine the x-axis of each curve to be generated for the interfering peak. c value.

[0090] According to a second embodiment of this disclosure, the peak wavelength of each curve to be fitted to the interfering peak is selected from a database of known spectral emission. Therefore, the database can provide peak wavelengths (xi) of curves that can be used to generate the interfering peak. c (The choice of )

[0091] Databases can be used to provide x-axis values ​​for one or more curves to be fitted to interference peaks. c Value. In some embodiments, the x-value to be used can be selected by the user before the curve generation process. c Values. For example, the user can know in advance the possible spectral emissions present in the sample used to generate interfering peaks (thus allowing some spectral emissions to be targeted while others are ignored). Alternatively, the processor can select the x-value for each curve to be generated. c The value of x c The value provides the closest match to the interfering peak. Therefore, in some embodiments, the processor can iteratively fit the curve to the interfering peak by referencing a database of known spectral peaks. An example of a suitable database of known spectral peaks is the National Institute of Standards and Technology (NIST) Atomic Spectroscopy Database, NIST Standard Reference Database #78.

[0092] An example of this fitting process is in Figure 9 and 10 As shown in the image. Figure 9 and 10 Graphs of interference peaks generated by samples including iron (Fe) and carbon (C) are shown separately. Therefore, the interference peaks are essentially products of the 273.074 nm spectral emission from iron and the 273.063 nm spectral emission from C (i.e., other spectral emissions may be present in the peaks at low levels).

[0093] like Figure 9 and 10 As shown, the method iteratively generates curves to fit the interfering peaks using interference curve parameters and peak wavelengths selected from a database of known spectral emission. As discussed above, the method can select the interference curve parameters to be fitted based on the detector position of the interfering peak (i.e., the detector position corresponding to the peak intensity value of the interfering peak). For each fitted curve, the method attempts to minimize the error between the sum of the interfering peak and the fitted curve. Table 1 below shows a portion of the database of spectral peaks that can be used to perform this process. Figure 9 and 10 The curve shown is generated.

[0094] Analytes Peak wavelength (nm) Li 273.055 OS 273.061 C 273.063 Hf 273.071 Ta 273.073 Fe 273.074 Ti 273.085 Ru 273.093

[0095] Table 1

[0096] Based on the database of spectral peaks shown in Table 1, processor 14 can select the peak corresponding to Fe -273.074 nm as x. c Value, the x c The value represents the minimum error between the curve and the interference peak for a single curve. Processor 14 then scales the intensity of the first curve to minimize the error between the first curve and the interference peak. Figure 9 The graph shows the first curve generated as the interference peak.

[0097] Figure 10 The graphs show the first and second curves fitted to the interference peak. For the second peak, the processor can iteratively fit the curve based on different peak wavelengths in the database. Figure 10 As shown, a curve corresponding to C-273.063 nm can be generated as a second curve. Therefore, the method iterates through the database of spectral peaks from beginning to end to achieve an optimal fit (error minimization) for each curve fitted to the interfering peak. The method can use regression-based fitting algorithms similar to those discussed above to generate curves for the interfering peak.

[0098] Therefore, as Figure 10 As shown, the contributions of Fe and C to the interference peaks can be analyzed separately, thus allowing for further analysis of the interference peaks.

[0099] Therefore, according to embodiments of this disclosure, a method for analyzing sample peaks in a sample spectrum is provided. The method can be performed by an optical spectrometer system to analyze interfering peaks detected by the optical spectrometer system. Specifically, the method can be performed by an optical emission spectrometer system. Thus, the method and system of this disclosure allow characterization of the spectral emission from interfering peaks, making the information from the interfering peaks available for further analysis. Therefore, the method and system of this disclosure allow users to analyze a large proportion of sample spectra using a simple and efficient analytical process.

Claims

1. A method for analyzing interference peaks in a sample spectrum generated by a detector of an optical spectrometer, the interference peaks being generated by multiple spectral peaks of different wavelengths, the method comprising: Based on the model of the expected curve parameters of the optical spectrometer and the position of the interfering peak on the detector of the optical spectrometer, interference curve parameters representing the peak shape of each spectral emission in the interfering peak are generated. Multiple curves are fitted to the interference peak by selecting a peak wavelength and peak intensity for each curve to be fitted to the interference peak, each curve corresponding to one of multiple spectral emissions at different wavelengths forming the interference peak, wherein each curve is fitted using interference curve parameters provided by a model of the expected curve parameters; and Output the multiple curves for further analysis.

2. The method according to claim 1, wherein The interference curve parameters include a first asymmetric interference curve parameter and a second asymmetric interference curve parameter, so as to fit multiple asymmetric curves to the interference peak.

3. The method according to claim 2, wherein the interference curve parameters are first double Gaussian curve parameters and second double Gaussian curve parameters.

4. The method according to any one of claims 1-3, wherein the model of the expected curve parameters of the optical spectrometer includes defining the relationship between the center wavelength of the interference peak and the interference curve parameters.

5. The method according to any one of claims 1-3, further comprising: The model for the expected curve parameters is generated based on the peak shapes and associated detector positions of multiple spectral peaks in the calibration spectrum generated from the calibration sample using the optical spectrometer.

6. The method according to any one of claims 1-3, wherein The sample spectrum is generated by the optical spectrometer using an echelle grating to diffract the sample light onto a two-dimensional detector.

7. The method according to any one of claims 1-3, further comprising: Determine how many curves should be fitted to the interference peak.

8. The method according to claim 7, wherein The number of curves to be fitted to the interference peak is determined based on the first and / or second derivatives of the interference peak.

9. The method according to any one of claims 1 to 3, wherein The peak wavelengths to which each curve is fitted to the interference peak are selected from a database of known spectral emission.

10. The method according to any one of claims 1-3, wherein The multiple curves to be fitted to the interference peak are fitted using a regression-based algorithm.

11. The method according to any one of claims 1-3, wherein The detector of the optical spectrometer is an array detector.

12. A controller for analyzing interfering peaks in a sample spectrum generated by a detector of an optical spectrometer, wherein the controller is configured to receive the interfering peaks from the optical spectrometer, the interfering peaks being generated by emission of multiple spectra at different wavelengths, the controller being further configured to: Based on the model of the expected curve parameters of the optical spectrometer and the position of the interfering peak on the detector of the optical spectrometer, interference curve parameters representing the peak shape of each spectral emission in the interfering peak are generated. Multiple curves are fitted to the interference peak by selecting a peak wavelength and peak intensity for each curve to be fitted to the interference peak. Each curve corresponds to one of multiple spectral emissions at different wavelengths that form the interference peak, wherein each curve is fitted using the interference curve parameters provided by a model of the expected curve parameters; and Output the multiple curves for further analysis.

13. An optical spectroscopy measurement system, comprising: An optical spectrometer configured to generate a sample spectrum from a sample using the detector of the optical spectrometer; A controller configured to analyze interference peaks in the sample spectrum, the interference peaks being generated by multiple spectral emissions of different wavelengths, the controller being configured to: Based on the model of the expected curve parameters of the optical spectrometer and the position of the interfering peak on the detector of the optical spectrometer, interference curve parameters representing the peak shape of each spectral emission in the interfering peak are generated. Multiple curves are fitted to the interference peak by selecting a peak wavelength and peak intensity for each curve to be fitted to the interference peak. Each curve corresponds to one of multiple spectral emissions at different wavelengths that form the interference peak, wherein each curve is fitted using the interference curve parameters provided by a model of the expected curve parameters; and Output the multiple curves for further analysis.

14. The optical spectroscopy system according to claim 13, wherein... The optical spectrometer includes a plasma source.

15. A computer program product comprising instructions that cause the controller according to claim 12 or the optical spectrometer according to claim 13 or 14 to perform the steps of the method according to any one of claims 1 to 11.

16. A computer-readable medium having instructions stored thereon that cause the controller according to claim 12 or the optical spectrometer according to claim 13 or 14 to perform the steps of the method according to any one of claims 1 to 11.

Images