Method for measuring transmittance or absorbance

Abstract

To provide a method for measuring transmittance or absorbance with which the influence of stray light can be reduced which is generated before and after spectroscopy by using a new model formula.SOLUTION: A transmittance spectrum comprising the transmittance t of a sample for respective wavelengths is defined as T, a transmitted light quantity spectrum comprising the quantity of transmitted light v through the sample for respective wavelengths measured by a spectrometer is defined as V, a light quantity distribution parameter that is a matrix comprising light quantity distribution components aij characteristic to the spectrometer is defined as A, and transmittance or absorbance is measured by using a model formula represented by "AT=V". The light quantity distribution component aij is a component indicating in what degree of light quantity a light quantity for every wavelength lj (j represents a distribution wavelength before spectroscopy) before spectroscopy is distributed in each band of a measured wavelength i of detected light after spectroscopy in blank measurement using the spectrometer, and is acquired in advance.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a method for measuring the transmittance or absorbance of a sample using a dispersive spectrometer using a diffraction grating or a prism, and particularly relates to a technique for reducing the influence of stray light generated before and after spectroscopy.

Background Art

[0002] Calibration of a measuring device is an operation to make the measured value of a reference sample in the device to be calibrated fall within a specified range. There are many methods for realizing this, and roughly classified, there are calibration using a conversion formula based on a measurement principle and calibration using a conversion formula derived from the measured value of a reference sample without relying on the measurement principle. For example, the method disclosed in Patent Document 1 corresponds to the latter. Generally, since the former is based on the measurement principle, the calibration result is easy to predict and handle, but the latter is difficult to predict and handle the calibration result. In the case of calibration based on the measurement principle, the superiority or inferiority of the modeling of the measurement principle is directly connected to the superiority or inferiority of the calibration result. As the simplest model formula for absorbance and stray light (light having a wavelength component other than the measurement wavelength) based on the measurement principle, abs=-log 10 {(I0×t+α) / (I0+β)} Here, abs: measured absorbance value, I0: blank light quantity, t: reference sample transmittance at the measurement wavelength, α: stray light quantity during measurement of the reference sample, β: stray light quantity during blank measurement, is conceivable. In this model formula, since there are two variables, α and β, α and β can be determined if absorbance measurement values of two concentrations of the same reference sample are obtained.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, the applicable range of absorbance correction of the above model formula is limited to at most the range of absorbance correction degree between two concentrations of the same reference sample. The reason is that in an actual spectrometer, α is the (vector) inner product between the light quantity spectrum (vector) other than the measurement wavelength and the transmittance spectra (vectors) of two concentrations of the reference sample, and β is the sum of the light quantity spectra other than the measurement wavelength. Therefore, a more refined and excellent model needs to encompass this situation.

[0005] Against such a background, for example, in paragraph

[0038] of Patent Document 2, the total wavelength component amount (light quantity spectrum) included in the light of a certain measurement wavelength of a spectrometer is represented as a vector, and the synthesis of all measurement wavelengths is represented as the matrix sum "I + D" of the identity matrix I and the stray light distribution matrix D, and the model formula "Ymeas = (I + D) · Y" is used for the calibration of the spectrometer. Further, in paragraph

[0020] , as a method for obtaining a stray light model, it is described that "based on the detection signals obtained from each of a plurality of light receiving elements when monochromatic light is introduced into the spectroscopic element, the position of the light receiving element where the monochromatic light is incident as stray light and the ratio of the amount of stray light incident on each light receiving element to the total amount of monochromatic light incident on the light receiving element are obtained for each of a plurality of monochromatic lights having different wavelengths, and an approximate formula using, as parameters, the position of the light receiving element where the stray light is incident and the ratio of the amount of stray light obtained for the plurality of monochromatic lights is used as the stray light model", and a method for obtaining a stray light model based on the spectra of a plurality of monochromatic lights measured by the spectrometer to be calibrated is shown.

[0006] The inventors considered that there was room for improvement in the model formula "Ymeas=(I+D)·Y" in Patent Document 2, and have intensively studied a method for calibrating transmittance and absorbance using a new model formula different from this. That is, an object of the present invention is to provide a method for measuring transmittance or absorbance that can reduce the influence of stray light generated before and after spectroscopy using a new model formula. Further, in obtaining the model formula, by adopting a method different from the monochromatic method, it is intended to improve practicality.

Means for Solving the Problems

[0007] "Measurement of Transmittance or Absorbance Using the Model Formula of AT=V" To solve the above problems, the method for measuring transmittance or absorbance according to the present invention is Let the transmittance spectrum composed of the transmittance t of the sample for each wavelength be T, the transmittance light amount spectrum composed of the transmittance light amount v of the sample for each wavelength measured by the spectrometer be V, and the light amount distribution component a ij of the spectrometer-specific matrix be A. It is a method for measuring transmittance or absorbance using the model formula represented by "AT=V", wherein the light amount distribution component a ij is a component indicating how much light is distributed in each band of the measurement wavelength i of the detected light after spectroscopy by the light amount l j (j indicates the distribution wavelength before spectroscopy). It is characterized in that it is obtained in advance.

[0008] "Estimation of Parameter A (Using Multiple Types of Reference Samples)" Here, for each input value of the transmittance spectrum matrix T REF composed of the known transmittance spectra of multiple types of reference samples, and the transmittance light amount spectrum matrix V REF composed of the transmittance light amount spectra of the multiple types of reference samples measured by the spectrometer, the light amount distribution component a REF satisfying the model formula "AT REF =V ij " is preferably estimated in advance using a computer.

[0009] "Estimation of Parameter A (using a single reference sample at multiple concentrations)" Or, using a single reference sample at multiple concentrations, a transmittance spectrum matrix T composed of the known transmittance spectra of the reference sample for each concentration REF , and a transmitted light quantity spectrum matrix V composed of the transmitted light quantity spectra of the reference sample for each concentration measured by the spectrometer REF For each input value of REF AT REF = V ij ", it is preferable to estimate the light quantity distribution component a

[0010] "Estimation of Parameter A (using multiple reference samples, each at multiple concentrations)" Or, using multiple reference samples, each at multiple concentrations, a transmittance spectrum matrix T composed of the known transmittance spectra of the reference sample for each type and each concentration REF , and a transmitted light quantity spectrum matrix V composed of the transmitted light quantity spectra of the reference sample for each type and each concentration measured by the spectrometer REF For each input value of REF AT REF = V ij ", it is preferable to estimate the light quantity distribution component a

[0011] "Estimation of Parameter A by quadratic programming method" The above estimation of the light quantity distribution component a ij is performed for each measurement wavelength i For the measurement wavelength i, the component (a REF ) that minimizes the magnitude of the vector represented by "AT REF - V ij " is preferably obtained by optimizing within the setting range of the solution of the component, by the quadratic programming method

[0012] Or, the above light quantity distribution component a ij is estimated for each measurement wavelength i, and the component (a sum )T REF -(V REF / v Bi )」 is optimized within the setting range of the solution of the component (a ij ) so that the magnitude of the vector represented by becomes minimum. It is preferably obtained by the quadratic programming method. Here, a sum is the sum (j = 1 to n) of the components (a ij ) at the measurement wavelength i, and v Bi is the measured light quantity of the blank measurement at the measurement wavelength i.

[0013] Or, the above light quantity distribution component a ij is estimated for each measurement wavelength i, and for the reference sample with a relatively low absorbance at the measurement wavelength i, the difference represented by 「-log 10 {(A / a sum )T REF}+log 10 (V REF / v Bi )」 is optimized within the setting range of the solution of the component (a ij ) so that the magnitude of the vector represented by becomes minimum. It is preferably obtained by the quadratic programming method.

[0014] Alternatively, the above light quantity distribution component a ij is estimated for each measurement wavelength i, For the reference sample with a relatively low absorbance at the measurement wavelength i, the difference represented by 「(T REF T a ij / a sum )-(v is / v Bi )」 is calculated for the measurement wavelength i, and for the reference sample with a relatively high absorbance at the measurement wavelength i, 「-log 10 (T REF T a ij / a sum )+log 10 (vis / v Bi ) to calculate the difference represented by and optimize the component (a ij ) so that the sum of the squares of these differences is minimized within the setting range of the solution of the component, preferably by the quadratic programming method. Here, T REF T is the transposed matrix of T REF , and s is the number of reference samples (s = 1 to p).

[0015] Alternatively, the estimation of the above light quantity distribution component a ij is performed for each measurement wavelength i, and for the measurement wavelength i, calculate the difference represented by "-log 10 (T REF T a ij / a sum ) + log 10 (v is / v Bi )", perform weighting on the difference according to the absorbance of the reference sample at the measurement wavelength i, and preferably obtain the component (a ij ) so that the sum of the squares of the weighted differences is minimized within the setting range of the solution of the component, by the quadratic programming method.

[0016] "Determination of reference concentration in the low absorbance region" In the estimation of the above light quantity distribution component a ij using the reference samples of the above plurality of concentrations, select a wavelength region with relatively low absorbance from the wavelength range of the transmitted light quantity spectrum of the reference sample for each measured concentration, and determine the concentration of the reference sample based on the absorbance in the selected wavelength region as the reference concentration. Based on the determined reference concentration, it is preferable to determine the correspondence between the transmitted light quantity spectrum of the reference sample for each measured concentration and the transmittance spectrum of the reference sample for each known concentration.

[0017] "Update of parameter A" Here, the method for measuring the transmittance or absorbance according to the present invention includes the method for updating the light quantity distribution parameter A, and the updating method is as follows: In the blank measurement of the spectrometer, the light quantity l for each wavelength before spectroscopy j (l j = a 1j + a 2j + ··· + a mj ) is used to form the light quantity spectrum L before spectroscopy, and the light quantity l for each wavelength before spectroscopy j indicates the light quantity ratio distribution component d of how much the light quantity is distributed in each band of the measurement wavelength i of the detected light after spectroscopy ij The light quantity ratio distribution parameter composed of is defined as matrix D. From the existing light quantity distribution component a ij to the light quantity ratio distribution component d ij (d ij = a ij / l j Here, "a ij / l j " means dividing each element in the j -th column of the light quantity distribution component a ij by the light quantity l for each wavelength j ). Calculate Execute the blank measurement of the spectrometer to obtain the blank light quantity spectrum V B . For each input value of the light quantity ratio distribution parameter D and the blank light quantity spectrum V B , use the computer to calculate a new light quantity spectrum L before spectroscopy that satisfies the model formula "DL = V B ". NEW Using the light quantity ratio distribution parameter D and the new light quantity spectrum L before spectroscopy NEW , calculate the updated value of the light quantity distribution parameter A NEW (a ij = d ij l j ). It is characterized by including the above.

[0018] "Discrimination of light quantity fluctuation" Also, the method for measuring the transmittance or absorbance according to the present invention furthermore, the blank light quantity spectrum V obtained by performing a blank measurement using the spectrometer Band the past blank light quantity spectrum V stored in the memory B are compared, and the variation of the blank light quantity spectrum V B is determined, and when there is a variation, the update of the light quantity distribution parameter A is automatically executed, which is preferably included.

[0019] "Measurement of transmittance or absorbance using the existing parameter A" Here, the method for measuring transmittance or absorbance according to the present invention is in the blank measurement of the spectrometer, the light quantity spectrum before spectroscopy l for each wavelength j (l j =a 1j +a 2j +···+a mj ) is defined as the light quantity spectrum L before spectroscopy, and the light quantity ratio distribution component d indicating to what extent the light quantity l for each wavelength before spectroscopy j is distributed in each band of the measurement wavelength i of the detected light after spectroscopy ij is defined as the light quantity ratio distribution parameter D composed of, and from the existing light quantity distribution component a ij the light quantity ratio distribution component d ij (d ij =a ij / l j ) is calculated, the blank measurement of the spectrometer is executed to obtain the blank light quantity spectrum V B , the spectrometer is used to measure the sample to obtain the transmitted light quantity spectrum V, a diagonal matrix having an arbitrary value greater than 0 as diagonal components is defined as L DIAGONAL and in the formula "T={(DL DIAGONAL ) -1 V} / {(DL DIAGONAL ) -1 V B}", each value of the light quantity ratio distribution parameter D, the blank light quantity spectrum V B and the transmitted light quantity spectrum V of the sample are input, and the transmittance spectrum T of the sample is calculated by the computer, or a diagonal matrix having an arbitrary value greater than 0 as diagonal components is defined as LDIAGONAL as, "S = -log 10 ((DL DIAGONAL ) -1 V) + log 10 ((DL DIAGONAL ) -1 V B ", substitute the values of the light quantity ratio distribution parameter D, the blank light quantity spectrum V B and the transmitted light quantity spectrum V of the sample into the formula, and it is preferable to calculate the absorbance spectrum S of the sample by the computer.

[0020] "Selection of using the model formula" In addition, the method for measuring transmittance or absorbance according to the present invention further includes selecting whether to execute or not execute a measurement method using the model formula represented by "AT = V" during the measurement with the spectrometer.

[0021] "Measurement program" In addition, the measurement program for transmittance or absorbance according to the present invention is for causing a computer to execute the procedure of the above method for measuring transmittance or absorbance.

[0022] "Spectrometer" In addition, the spectrometer according to the present invention is characterized by including a computer capable of executing the above measurement program for transmittance or absorbance.

Advantages of the Invention

[0023] <Method for Measuring Transmittance or Absorbance Using the New Model Formula "AT = V"> In the method for measuring transmittance or absorbance of the present invention, based on the measurement principle of the spectrometer, using the new model formula "AT = V", the transmittance or absorbance of a sample with reduced stray light influence can be measured. Furthermore, the applicable range of the measurement using the new model formula "AT = V" becomes wider.

[0024] <Estimation of Parameter A> In addition, data of the transmittance t by wavelength of a plurality of standard samples (transmittance spectrum matrix TREF ) and the data of the transmitted light amount spectrum (transmitted light amount spectrum matrix V) obtained by measuring these standard samples with a spectrometer REF ) If there is, for example, a calibration parameter (light amount distribution parameter A) for reducing the influence of stray light can be estimated by using a solution method based on the quadratic programming method. Therefore, it is not necessary to measure monochromatic light deliberately, the high cost of the spectrometer can be avoided, and the efficiency of the calibration work can be improved.

[0025] <Update of parameter A> Also, for example, when the conditions of the blank measurement change (due to changes in the solvent or measurement environment, etc.) and the light amount distribution parameter A fluctuates, the parameter A can be updated only by performing the blank measurement of the spectrometer.

[0026] <More practical measurement method of transmittance or absorbance> Also, when the light amount distribution parameter A fluctuates, the transmittance or absorbance of the sample can be efficiently measured based on the result of the blank measurement of the spectrometer without updating the existing light amount distribution parameter A.

Brief Description of Drawings

[0027]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Mode for Carrying Out the Invention

[0028] Hereinafter, a first embodiment of a method for measuring transmittance or absorbance of the present invention will be described with reference to the drawings.

[0029] <Configuration of the Spectrometer> Fig. 1 shows the optical configuration of a typical spectrometer. The spectrometer shown in Fig. 1 is a device that can measure optical property values (e.g., absorbance, transmittance) of a sample for each measurement wavelength and output them as spectral information. Fig. 1 is merely a representative example, and the measurement method of the present invention can be applied to any spectrometer using an optical dispersion element 30 (diffraction element G or prism). The wavelength range to be spectrally analyzed is not limited, but for example, it is suitable for spectrometers in the ultraviolet, visible, near-infrared to far-infrared wavelength range of "190 nm to 28.5 μm".

[0030] The spectrometer can be said to extract the component of the desired wavelength from the light passing through the sample chamber 20 (e.g., the flow cell in Fig. 1) and individually measure the light quantity with the detector 40, or measure the light quantities of multiple wavelengths together using a PDA. However, in a real spectrometer, the light quantity detected as the light of a certain measurement wavelength contains not only the light quantity of the same wavelength as that measurement wavelength (also called the true light quantity), but also a small amount of the light quantity of wavelengths different from that measurement wavelength (called stray light).

[0031] The purpose of this document is to show a method for measuring the transmittance t or absorbance abs that contains as little stray light as possible (or eliminates the influence of stray light as much as possible) in a spectrometer.

[0032] <Spectrometer model> Fig. 2 shows the basic matters of the spectrometer model. In this document, the wavelength of the true light quantity spectrum L before spectroscopy is called the "distribution wavelength" and is represented by j (j = 1 to n), and the wavelength of the transmitted light quantity spectrum V actually measured by the spectrometer is called the "measurement wavelength" and is represented by i (i = 1 to m) to distinguish between the two. For the sake of briefly explaining the method of linear algebra, the number of points n of the distribution wavelength and the number of points m of the measurement wavelength are set to n = m. The expressions "true" and "close to the true value" are equivalent to the meaning of "containing no stray light".

[0033] The true light quantity spectrum L is the light quantity spectrum of the light passing through the sample chamber 20 (a state without a sample, for example, only a solvent with a sample concentration of zero), and as shown in Fig. 2, it has the light quantity for each true wavelength (l1 to l n ) as components. On the other hand, the transmitted light quantity spectrum V is a light quantity spectrum created based on the detection signal from the detector 40 and includes stray light. The main causes of stray light in an actual spectrometer are reflections and scatterings of the components of the spectrometer (such as the walls of the housing, slits, detectors, window plates, etc.). In the spectrometer model, regardless of the cause of the stray light, the relationship between the true light quantity spectrum L and the transmitted light quantity spectrum V is defined as follows.

[0034] For the true wavelength-specific light quantity l1 of the first distribution wavelength (j = 1), most of it is detected as the light quantity of the first measurement wavelength (i = 1), and although a small amount is also detected as the light quantity of other measurement wavelengths (i = 2 to m). Thus, the extent to which the true wavelength-specific light quantity l1 is distributed in the measurement wavelength range (i = 1 to m) of the detector is represented by the light quantity a i1 as shown. Similarly, for the true wavelength-specific light quantity l2 of the second distribution wavelength (j = 2), most of it is detected as the light quantity of the second measurement wavelength (i = 2), and although a small amount is also detected as the light quantity of other measurement wavelengths (i = 1, 3 to m), it is distributed in the measurement wavelength range (i = 1 to m) of the detector by the light quantity a i2 as shown. That is, the light quantity a ij represents the distributed light quantity of the distribution wavelength j included in the light quantity v1 of the measurement wavelength i. In this way, the distribution of the true wavelength-specific light quantity l j can be represented by the light quantity distribution parameter A with the light quantity a ij as the matrix component.

[0035] For example, the light quantity of the first measurement wavelength (i = 1) of the transmitted light quantity spectrum V is equal to the sum of the distributed light quantities a 1j for all distribution wavelengths j (j = 1 to n), and is represented by "a 11 + a 12 +... + a 1n ". When monochromatic light is spectroscopically measured with an actual spectrometer, a bandwidth occurs, so the diagonal components and the components in the vicinity of the matrix a ij indicate light quantities of magnitudes corresponding to that bandwidth.

[0036] Next, the "AT = V" model showing the spectrometer with a sample in the sample chamber will be described. A: Light quantity distribution parameter of the spectrometer (aij (a matrix having as components) T: The transmittance spectrum of a certain sample (t j (a vector having as components) V: The transmitted light quantity spectrum obtained by actually measuring the above sample (v i (a vector having as components) The transmittance spectrum T is the wavelength - specific transmittance of the sample measured by a more accurate spectrometer rather than the spectrometer that is the object of calibration here. Alternatively, literature values for the transmittance spectrum of the sample may be used.

[0037] In an ideal spectrometer, for example, for the light quantity l1 of the first distribution wavelength of the incident light to the sample chamber, the value obtained by multiplying it by the transmittance t1 of the sample at that distribution wavelength is measured as the light quantity v1 of the first measurement wavelength, and no stray light is generated. Therefore, the light quantity distribution parameter A of the ideal spectrometer has the total amount of the respective true wavelength - specific light quantities l j in the diagonal components, and is zero for components other than the diagonal components, and the "AT = V" model becomes Equation (1). For the component v1 of the first measurement wavelength of the transmitted light quantity spectrum, it is the value (l1t1) obtained by multiplying the true wavelength - specific light quantity l1 of the first distribution wavelength by the transmittance t1 of that distribution wavelength.

[0038]

Equation

[0039] However, in a real spectrometer, as shown in Equation (2), most of the light of the first distribution wavelength (j = 1) is measured as the light quantity of the first measurement wavelength (i = 1), but a very small part of the light of the first distribution wavelength (j = 1) becomes stray light and is measured as the light quantity of other measurement wavelengths (i = 2, 3...). Therefore, the "AT = V" model of a real spectrometer is represented by Equation (2). For the light quantity v1 of the first measurement wavelength of the transmitted light quantity spectrum, for each light quantity component (a 11 a 12 … a 1n ) distributed as the light quantity of the first measurement wavelength (i = 1), the sum of the products of each corresponding transmittance of the sample (t1t2… t n ) (v1 = a 11 t1 + a12 t2 + … + a 1n t n ) becomes

[0040] [Number]

[0041] The “AT = V” model obtained by expanding Equation (2) to p samples s is shown in Equation (3). Here, T: Transmittance spectrum of sample s (s = 1~p) (t js as a component matrix) V: Transmitted light quantity spectrum obtained by actually measuring each sample s (v is as a component matrix) Here, as described above, assuming the number of measurement wavelengths i is m and the number of distribution wavelengths j is n, the size of the matrix of parameter A is m rows × n columns, the size of the matrix of transmittance spectrum T is n rows × p columns, and the size of the matrix of transmitted light quantity spectrum V is m rows × p columns.

[0042] [Number]

[0043] The light quantity distribution parameter A represents the characteristics inherent to the spectrometer. If the light quantity distribution parameter A can be obtained in advance, when measuring the transmitted light quantity spectra V of various samples with that spectrometer, from the “AT = V” model, the transmittance spectrum T (= A -1 V) of the sample not affected by stray light can be calculated, and the transmittance spectrum T close to the true value of the sample, that is, the calibration spectrum, can be obtained. Furthermore, by taking the common logarithm (-log js (t 10 (t js )) of the transmittance t, the absorbance can be obtained, and the absorbance spectrum S of the sample can also be obtained. In this embodiment, since the diagonal elements (or the diagonal elements and the elements in their vicinity) of the matrix A are much larger than the other elements, the inverse matrix A -1 is easy to calculate. Generally, the inverse matrix A -1The calculability is determined by calculating the condition number of matrix A. However, since the condition number of matrix A in this embodiment is at most several hundreds even when large, the inverse matrix A with sufficient accuracy for the present application can be obtained. -1 Therefore, as long as matrix A, which is the light quantity distribution parameter, can be obtained, a calibration spectrum close to the true value can be acquired from the transmitted light quantity spectra of various samples by equation (2). Here, for the sake of simplicity in explaining the model, the inverse matrix A -1 is used. However, according to the knowledge of numerical calculations in linear algebra, when obtaining T from "AT = V", it should be noted that a method using LR decomposition or QR decomposition, which is excellent in calculation accuracy and speed, should be adopted.

[0044] The concept underlying the "AT = V" model is also known from Patent Document 1, and it is possible to obtain the light quantity distribution parameter A using monochromatic light. However, the calibration method using monochromatic light requires the provision of a device for performing spectroscopic measurement using monochromatic light, which is costly, and the calibration work is cumbersome and inefficient, so it is not a practical method.

[0045] There is no specific description of a method for obtaining the light quantity distribution parameter A without using monochromatic light. The reason is that there are difficult problems in each of the generally conceivable methods as described below.

[0046] A generally conceivable first method is shown. According to the knowledge of linear algebra, A can be obtained from VT -1 However, in order to uniquely determine the inverse matrix T -1 the p sample spectra must be independent of each other. What this means is that, for example, if the wavelength step is 1 nm in the measurement wavelength range of 200 to 700 nm, about 501 independent sample spectra are required. Selecting samples having a large number of independent spectra, performing unit operations such as pretreatment of individual samples, and measuring the spectra of these samples are very time-consuming.

[0047] A second generally conceivable approach is shown. As a method for efficiently measuring the transmitted light quantity spectra V of a plurality of samples, for example, a liquid feeding system capable of mixing a large number of reference samples is connected to a flow cell of a spectrometer or the like, and while sequentially changing the concentration of the reference sample in the solvent, the sample mixed solvent is introduced into the flow cell of the spectrometer, whereby spectra of a large number of sample mixed solvents with different concentrations can be acquired in a short time. However, since the spectra measured by this method have no or low (as vectors) independence between the spectra, the condition number of T becomes an extremely large number, and it is difficult to calculate T. Therefore, it is difficult to directly obtain A from A = VT. -1 In this book, the calibration of the spectrometer means solving the light quantity distribution parameter A in the above-mentioned "AT = V" model of the spectrometer. Here, a method for efficiently estimating such a light quantity distribution parameter A with high precision will be described. -1 The above-mentioned second approach has a great merit in that a plurality of spectra can be measured in a short time. Therefore, the inventors examined a method for obtaining the light quantity distribution parameter A without calculating T while using this measurement method.

[0048] As a result, regarding the optical characteristics represented by the light quantity distribution parameter A, that is, since the light quantity distribution parameter A is the distribution of the already spectro-dispersed light, when comparing the diagonal components (or the diagonal components and the components in the vicinity thereof) of the matrix of the light quantity distribution parameter A with the amounts of the other components, it was noted that "the amount of the diagonal components >> the amounts of the other components", and the inventors considered that the solution method of the so-called "quadratic programming method", which appropriately sets the initial value and range (constraint conditions) of the solution to be obtained for each element of the light quantity distribution parameter A and calculates the solution that satisfies a predetermined relational expression, can be applied to the method for estimating the parameter A of the present embodiment.

[0049] The above-mentioned second approach has a great merit in that a plurality of spectra can be measured in a short time. Therefore, the inventors examined a method for obtaining the light quantity distribution parameter A without calculating T while using this measurement method. -1 As a result, regarding the optical characteristics represented by the light quantity distribution parameter A, that is, since the light quantity distribution parameter A is the distribution of the already spectro-dispersed light, when comparing the diagonal components (or the diagonal components and the components in the vicinity thereof) of the matrix of the light quantity distribution parameter A with the amounts of the other components, it was noted that "the amount of the diagonal components >> the amounts of the other components", and the inventors considered that the solution method of the so-called "quadratic programming method", which appropriately sets the initial value and range (constraint conditions) of the solution to be obtained for each element of the light quantity distribution parameter A and calculates the solution that satisfies a predetermined relational expression, can be applied to the method for estimating the parameter A of the present embodiment. As a result, regarding the optical characteristics represented by the light quantity distribution parameter A, that is, since the light quantity distribution parameter A is the distribution of the already spectro-dispersed light, when comparing the diagonal components (or the diagonal components and the components in the vicinity thereof) of the matrix of the light quantity distribution parameter A with the amounts of the other components, it was noted that "the amount of the diagonal components >> the amounts of the other components", and the inventors considered that the solution method of the so-called "quadratic programming method", which appropriately sets the initial value and range (constraint conditions) of the solution to be obtained for each element of the light quantity distribution parameter A and calculates the solution that satisfies a predetermined relational expression, can be applied to the method for estimating the parameter A of the present embodiment. "the amount of the diagonal components >> the amounts of the other components" By appropriately setting the initial value and range (constraint conditions) of the solution to be obtained for each element of the light quantity distribution parameter A and calculating the solution that satisfies a predetermined relational expression, that is, by applying the solution method of the so-called "quadratic programming method", the inventors considered that it can be applied to the method for estimating the parameter A of the present embodiment. From the results of measuring a reference sample with an actual spectrometer and evaluating the vertical axis accuracy characteristics of the spectrometer, the initial value and range can be easily estimated. For example, in the spectrometer to be calibrated shown in this embodiment, when the magnitude of the diagonal component is "1", the magnitudes of the other components can be estimated to be "about 0.02, or 0.02 or less". This is set as the range of the constraint conditions. Also, the initial value can use the light quantity spectrum at the time of blank. Note that since the value of 0.02 varies depending on the spectrometer to be calibrated, this value should be determined from the evaluation of the vertical axis accuracy characteristics of the spectrometer.

[0050] In summary, the solution method using the inverse matrix T -1 is a method that directly tries to solve the matrix of parameter A so that all components a ij of parameter A become the expected values. Since the (vector-based) independence between transmittance spectra is low, there is a high possibility that the solution cannot be determined. On the other hand, in this embodiment, since the possible range of the component a ij to be obtained is determined, if a method such as the quadratic programming method is used to obtain an approximate solution within the limited solution range, it is easy to obtain the solution.

[0051] For example, when the measurement range is 200 to 700 nm, it is advisable to use a plurality of types of reference samples and measure a plurality of concentration spectra for each reference sample. Preferably, by using reference samples having transmittance spectra with a high degree of independence of about 5 types or more, the light quantity distribution parameter A can be obtained with higher accuracy.

[0052] When estimating parameter A using reference samples with multiple concentrations, the transmittance spectrum matrix T REF consisting of the known transmittance spectra of the reference samples for each concentration, and the transmittance light quantity spectrum matrix V REF consisting of the transmittance light quantity spectra obtained by actually measuring these reference samples for each concentration with a spectrometer are used as input values respectively. Then, for these input values, the light quantity distribution component a REF that satisfies the model formula "AT REF =V ijwill be estimated by a computer. Here, in the estimation of the light quantity distribution component a using reference samples of multiple concentrations, it is advisable to determine the reference concentration of the reference sample for each concentration as follows. That is, select a wavelength range with relatively low absorbance from the wavelength range of the transmitted light quantity spectrum of the reference sample for each measured concentration, and determine the concentration of the reference sample based on the absorbance in the selected wavelength range as the reference concentration. Then, based on the determined reference concentration, by determining the correspondence between the transmitted light quantity spectrum of the reference sample for each measured concentration and the transmittance spectrum of the reference sample for each known concentration, the model formula "AT ij In the estimation of, the reference concentration of the reference sample for each concentration may be determined as follows. That is, select a wavelength range with relatively low absorbance from the wavelength range of the transmitted light quantity spectrum of the reference sample for each measured concentration, and determine the concentration of the reference sample based on the absorbance in the selected wavelength range as the reference concentration. Then, based on the determined reference concentration, by determining the correspondence between the transmitted light quantity spectrum of the reference sample for each measured concentration and the transmittance spectrum of the reference sample for each known concentration, the model formula "AT REF =V REF " can be used to more efficiently estimate the light quantity distribution component a ij .

[0053] Furthermore, in the estimation of the above parameter A, by using multiple types of reference samples and samples with multiple concentrations for each reference sample, it becomes more efficient and practical. That is, the transmittance spectrum matrix T REF composed of the known transmittance spectra of the reference samples for each type and each concentration, and the transmitted light quantity spectrum matrix V REF composed of the transmitted light quantity spectra of the reference samples for each type and each concentration measured by the spectrometer, for each input value, estimate the light quantity distribution component a REF =V REF " that satisfies the model formula "AT ij using a computer in advance. Also in this case, as described above, the reference concentration of the reference sample for each concentration may be determined, and based on the reference concentration, the correspondence between the measured transmitted light quantity spectrum and the known transmittance spectrum may be determined.

[0054] As an example, when estimating the light quantity distribution parameter A using spectra in the measurement range of 190 to 700 nm with a wavelength step of 4 nm (or 2 nm), six types of reference samples are selected, and samples of each reference sample with 21 to 22 types (or 42 to 43 types) of concentrations are used. That is, 128 (or 256) spectra, the same number as the number of wavelength points to be measured, may be obtained, and the parameter A may be estimated based on these spectra. Details will be described in the examples below.

[0055] The specific solution method of parameter A by the quadratic programming method (a representative example of the non-linear programming method) is within a limited solution range (constraint condition: the range that the component (a ij ) can take is shown by an inequality or the like), and a method of obtaining the optimal solution of the component (a ij ) such that a predetermined objective function (quadratic function) becomes the maximum value or the minimum value. It can estimate the components of parameter A with high accuracy more reliably and efficiently than the solution method using the inverse matrix T -1 .

[0056] In the calibration method of this embodiment, the quadratic programming method is applied to estimate the light quantity distribution parameter A from the "AT = V" model of the spectrometer.

[0057] In advance, the transmittance spectra T REF of p reference samples are obtained from the measurement results using a more accurate spectrometer or from the survey results of documents or the like, and stored in the memory 60 of the spectrometer. p reference samples are actually measured with the spectrometer to be calibrated, and the obtained transmitted light quantity spectra V REF are stored in the memory 60. Next, the arithmetic unit 50 of the spectrometer executes the quadratic programming estimation program of the "AT REF = V REF " model with the stored transmittance spectra T REF and the transmitted light quantity spectra V REF as input values to estimate the light quantity distribution parameter A. Finally, the estimated light quantity distribution parameter A is stored in the memory 60. In addition, as a spectrometer with higher precision, it is advisable to use one that can exhibit linearity up to a range where at least absorbance (AU) 1 is added to the calibration range of the desired absorbance (AU) for the spectrometer to be calibrated.

[0058] <Estimation of Light Quantity Distribution Parameters by the Quadratic Programming Method> In the execution of the quadratic programming calibration program, the arithmetic unit 50 calculates the distributed light quantity a for each row of the light quantity distribution parameter A, and finally synthesizes these to obtain a matrix of the light quantity distribution parameter A. ij and finally combines these to obtain a matrix of the light quantity distribution parameter A.

[0059] Equation (4) is obtained by extracting the light quantity of the first row of the light quantity distribution parameter A, that is, the light quantity distributed for each wavelength of the first measurement wavelength (a 11 a 12 … a 1n ) from Equation (3) of the "AT = V" model. As explained in Equation (3), the first column of the transmittance spectrum T is the transmittance for each wavelength of the first sample (t 11 t 21 … t m1 ). Therefore, the product of the distributed light quantity (a 11 a 12 … a 1n ) and the transmittance for each wavelength (t 11 t 21 … t m1 ) is the transmitted light quantity at the first measurement wavelength (v 11 = a 11 t 11 + a 12 t 21 + … + a 1n t m1 ).

[0060] Hereafter, to explain the solution method for the light quantity (a 11 a 12 … a 1n ) of the first row of the light quantity distribution parameter A in the form of Ax = b, which is familiar in linear algebra problems, both sides of "AT REF = V REF " are transposed to obtain "T REF T A T = V REF TUse the formula of 」. This transformation is only for convenience and has no essential meaning. The transposed form of both sides of Equation (4) is shown in Equation (5). Equation (5) may be abbreviated as 「T REF T a j =v p 」. a j and v p are vectors. Note that the 「 T 」 on the upper right of the matrix represents transposition. However, in the following description, the transposition symbol 「 T 」 may be omitted.

[0061]

Number

[0062]

Number

[0063] ·Constraint conditions When solving the light amounts (a 11 a 12 … a 1n ) in the first row of parameter A, set the characteristics of these distributed light amounts as constraint conditions in the quadratic programming method. First, according to the definition of each light amount a 11 , a 12 , …, a 1n in the first row, all light amounts a 11 , a 12 , …, a 1n are non - negative (non - negative condition).

[0064] The light amount distribution parameter A represents the distribution of light that has already been spectro - analyzed by the spectrometer. When the step interval of the measurement wavelength i is much wider than the bandwidth of the monochromatic light spectro - analyzed by the spectrometer, that is, when two or more measurement wavelengths i are not included in the bandwidth of the monochromatic light, the light amount a 11 of the main wavelength, which is the diagonal component of the matrix A, becomes a sufficiently large value compared to the light amounts a 12 ~a 1n (corresponding to the stray light component). That is, a 11 >>Σa1j (Here, j = 2 to n). Therefore, as a constraint condition in the quadratic programming method, for example, "when the amount of light a of the main wavelength is 1, the amounts of light a of other wavelengths 11 ~a 12 ~a 1n are each 0.02 or less". Refer to Equation (6-1) and Equation (6-2). Also, for example, add "the sum of stray light components Σa 1j (here, j = 2 to n) is less than 0.05". Refer to Equation (6-3).

[0065]

Number

[0066]

Number

[0067]

Number

[0068] Here, Equation (6-1) is a 11 +a 12 +…+a 1n =v 11 ···(7-1) and is equivalent. Also, the condition of "v 11 :δ j =1:0.02 or less (j = 2 to n)" in Equation (6-2) is expressed by the following equation together with the non-negativity condition. 0≦a 12 ≦0.02v 11 0≦a 13 ≦0.02v 11 · · ···(7-2) · 0≦a 1n ≦0.02v 11 Also, Equation (6-3) is a 12+a 13 +…+a 1n <0.05 ···(7-3) is equivalent to

[0069] ·Objective function Next, an objective function for solving the light quantity (a 11 a 12 … a 1n ) of the first row of parameter A is defined as in Equation (8). The objective function takes the difference "T REF T (a 11 a 12 … a 1n )-(v 11 v 12 … v 1p )" of the vectors on both sides of Equation (5), and is set to calculate the optimal solution of the light quantity that minimizes its magnitude from the range of solutions set as the constraint conditions.

[0070] ObjFunc =Σ(T REF T a 1j -v 1s ) 2 =Σ(t 1s a 11 +t 2s a 12 +…+t ns a 1n -v 1s ) 2 ···(8) Here, s represents the sample, s = 1 to p, j represents the distribution wavelength, j = 1 to n, and the summation symbol Σ represents the sum of the values of the expressions within Σ when the variable s is from 1 to p.

[0071] The intention of the objective function is that the vector of the result of multiplying the transposed matrix of the transmittance spectrum matrix T REF by the light quantity distribution spectrum of the first measurement wavelength (represented by T REF T a 1j ) and the vector of the actually measured transmitted light quantity spectrum V REF (represented by v 1sThe amount of light (a 1j ) is determined such that the magnitude of the difference vector from (represented by) is minimized. Whether the minimum value has reached the target value can be used to determine whether the calculated amount of light (a 1j ) is appropriate. If it is not appropriate, the calculation is repeated until an appropriate solution is obtained, such as by changing the standard sample.

[0072] The method for estimating the light quantity distribution parameter A by such a quadratic programming method is applied to each row of the parameter A (the amount of light for each distribution wavelength j at the measurement wavelength i) in the same manner as in Equation (5). If the amount of light (a ij ) for each distribution wavelength j at all measurement wavelengths i can be estimated, they are combined to obtain the light quantity distribution parameter A.

[0073] The above estimation method is realized by an information processing apparatus (arithmetic unit 50) configured by a computer executing an estimation program. The estimation program will be described using the flow of FIG. 3.

[0074] Step S1: Input the actually measured data (transmission light quantity spectrum V REF ) for p reference samples from the outside. Step S2: Read the data of the transmittance spectrum T REF (transmittance for each distribution wavelength j) of p reference samples from the memory. Step S3: Read the set values of the constraint conditions (such as the solution range of the light quantity distribution parameter A) from the memory. Step S4: Using the input values of the transmission light quantity spectrum V REF and the transmittance spectrum T REF , under the constraint conditions of the light quantity distribution parameter A, solve the row components of the light quantity distribution parameter A that minimize the objective function "Σ(T REF T a 1j -v 1s ) 2 ". Here, the quadratic programming method is applied individually to each row of the light quantity distribution parameter A. Step S5: The components of all rows (a ijAfter solving the equation (0)), the solutions are combined to obtain the light quantity distribution parameter A, and the program ends.

[0075] In the estimation program of the present invention, for the sake of easy understanding in this document, the "AT = V" model is transformed as in Equation (4) or Equation (5), but such transformation processing is not necessary in the estimation program.

[0076] The main effects of this estimation method are as follows. · It is not necessary to introduce monochromatic light into the spectrometer. · Since calibration is performed using the sample chamber used for actual sample measurement, an appropriate light quantity distribution parameter A close to actual measurement can be obtained. · A light quantity distribution parameter A that can remove stray light inherent in the spectrometer from the measurement spectrum of the sample can be obtained. · Since it is a solution of the quadratic programming method independent for each measurement wavelength, parallel processing is possible. For example, the acquisition time of the light quantity distribution parameter A can be shortened in proportion to the number of computers (the number of calculators, or the number of cores of a CPU composed of multiple cores) that execute the estimation program.

[0077] The estimation method of the present invention can also be realized by an external server (estimation unit 70, memory 80) as shown in FIG. 4, for example, together with the calculation unit 50 of the spectrometer or instead of the calculation unit 50, by executing the estimation program.

[0078] The estimated parameter A is registered in the memory 60 of the spectrometer and used for normal spectroscopic measurement. FIGS. 5(A) and (B) show the spectroscopic measurement flow of the sample using the light quantity distribution parameter A.

[0079] The spectroscopic measurement program is stored in advance in the memory 60 of the spectrometer together with the estimated light quantity distribution parameter A, and is executed by the calculation unit 50 composed of a computer or the like. Step S11: Actually measure the sample to obtain the transmitted light quantity spectrum V. Step S12: Using the spectrometer model "AT = V", obtain the transmittance spectrum T of the sample. Step S13: Calculate the absorbance spectrum based on the transmittance spectrum T, and end the spectroscopic measurement program. Note that the spectroscopic measurement program may be configured such that the user can select whether to execute the measurement method using the model "AT = V" of the present embodiment during the spectroscopic measurement with the spectrometer.

[0080] Instead of the objective function of Equation (8), Equation (9) based on the transmittance can also be adopted. Similar to Equation (8), Equation (9) is the objective function when focusing on the first measurement wavelength (i = 1).

[0081] ObjFunc = Σ{(T REF T a 1j / a sum ) - (v 1s / v B1 )} 2 = Σ{t 1s (a 11 / a sum ) + t 2s (a 12 / a sum ) + … + t ns (a 1n / a sum ) - (v 1s / v B1 )} 2 ···(9) Here, a sum is the sum of a 1j at the first measurement wavelength (i = 1), that is, Σa 1j . v B1 is the measured light amount when there is no sample absorption at the first measurement wavelength (i = 1) (during blank measurement). s represents the sample, s = 1 to p, j represents the distribution wavelength, j = 1 to n, and the summation symbol Σ represents the sum of the values of the expressions within Σ when the variable s is from 1 to p.

[0082] Note that the estimation of the light amount distribution component a ij by the objective function of Equation (9) is executed for each measurement wavelength i. For the measurement wavelength i, "(A / a sum )TREF -(V REF / v Bi )」(where a sum is the sum of the components at the measurement wavelength i (a ij ), j = 1 to n).) The magnitude of the vector represented by is minimized, and the component (a ij ) is obtained by optimizing within the solution range of the component. This is based on the concept.

[0083] Alternatively, instead of the objective function of Equation (8), Equation (10) based on absorbance can also be adopted. Equation (10) is the objective function when focusing on the first measurement wavelength (i = 1), similar to Equation (8).

[0084] ObjFunc = Σ{ -log 10 (T REF T a 1j / a sum ) + log 10 (v 1s / v B1 )} 2 = Σ[ -log 10 {t 1s (a 11 / a sum )} - log 10 {t 2s (a 12 / a sum )} - … - log 10 {t ns (a 1n / a sum )} + log 10 (v 1s / v B1 )] 2 ···(10) Here, a sum is the sum of a 1j , that is, Σa 1j . v B1is the measured light quantity when there is no sample absorption at the first measurement wavelength (i = 1) (during blank measurement). s represents the sample, where s = 1 to p, j represents the distribution wavelength, where j = 1 to n, and the summation symbol Σ represents the sum of the values of the expressions within Σ when the variable s ranges from 1 to p.

[0085] In addition, the estimation of the light quantity distribution component a ij by the objective function of Equation (10) is performed for each measurement wavelength i. For the measurement wavelength i, "-log 10 {(A / a sum )T REF}+ log 10 (V REF / v Bi )" (where a sum is the sum (j = 1 to n) of the components (a ij ) at the measurement wavelength i.) is obtained by optimizing the component (a ij ) within the solution range of the component such that the magnitude of the vector represented by it is minimized. This is based on the idea.

[0086] Furthermore, depending on the magnitude of the transmittance value (v 1s / v B1 ) of the reference sample at the first measurement wavelength (i = 1), the expression within the objective function is switched to the following Equation (11) or Equation (12) each time, and the sum of their squares can also be used as the objective function. For example, with a transmittance boundary value of 0.1, when v 1s / v B1 is 0.1 or more, Equation (11) is used, and when it is less than 0.1, Equation (12) is used, and the sum of their squares is used as the objective function.

[0087] (t j1 a 1j / a sum ) - (v 1s / v B1 ) ···(11) Note that Equation (11) is equivalent to the following equation. t 1s (a 11 / a sum ) + t 2s (a 12 / a sum) + … + t ns (a 1n / a sum ) - (v 1s / v B1 )

[0088] -log 10 (t j1 a 1j / a sum ) + log 10 (v 1s / v B1 ) ···(12) Note that Equation (12) is equivalent to the following equation. -log 10 {t 1s (a 11 / a sum )} - log 10 {t 2s (a 12 / a sum )} - … - log 10 {t ns (a 1n / a sum )} + log 10 (v 1s / v B1 )

[0089] The reason for switching in this way is that since absorbance calculation is logarithmic calculation, in the objective function calculation of Equation (8), the difference in the high absorbance region is relatively small compared to the difference in the low absorbance region, so the relative weight (degree of contribution to the objective function value) in the objective function tends to be small. In this application, since the concentration and the measured value are linear, the absorbance, which is frequently used in practice, can also be an important calibration target value, so this should be fully considered.

[0090] Equations (8) and (9) can obtain excellent calibration results in the low absorbance region (high transmittance), but it is difficult to obtain sufficient calibration results in the high absorbance region (low transmittance). On the other hand, Equation (10) is the opposite. Therefore, in the objective function, it is more suitable for this application to appropriately switch between the transmittance difference that is easier to compare with the absorbance directly in the low absorbance region (high transmittance) and the absorbance difference in the high absorbance region (low transmittance), and use the sum of their squares as the objective function.

[0091] Note that for the switching method according to the above absorbance, for a reference sample with a relatively low absorbance at measurement wavelength i, the difference represented by "(T REF T a ij / a sum )-(v is / v Bi )" is calculated for that measurement wavelength i. For a reference sample with a relatively high absorbance at measurement wavelength i, "-log 10 (T REF T a ij / a sum )+log 10 (v is / v Bi )" (where a sum is the sum (j = 1 to n) of the components (a ij ) at measurement wavelength i.) is calculated for that measurement wavelength i. By optimizing the component (a ij ) such that the sum of the squares of these differences is minimized within the setting range of the solution of the component, the light quantity distribution component a ij is estimated based on this concept.

[0092] However, depending on the switching boundary value, a step may occur between the two difference values, affecting the function of the objective function. In terms of implementation, it is advisable to take measures such that the sensitivity of the evaluation function at the switching boundary value is continuous.

[0093] The method for making the transmittance difference in Equation (11) and the absorbance difference in Equation (12) continuous is easily understood by comparing the differential values of both as shown in FIG. 6. FIG. 6 is a graph with absorbance on the horizontal axis and the differential value thereof (Δ absorbance) and the transmittance differential value (Δ transmittance) on the vertical axis.

[0094] The line of ΔAbs in the graph (parallel to the x-axis and with a value of 1) is the absorbance differential value. The curve of ΔTRate is the differential value at the transmittance corresponding to the absorbance on the horizontal axis. The magnitude of the differential value becomes the magnitude of the sensitivity in the objective function. From this graph, it can also be seen that, as described above, depending on the magnitude of the transmittance value, the transmittance difference in Equation (11) in the objective function has high sensitivity in the low absorbance region, while the absorbance difference in Equation (12) has high sensitivity in the high absorbance region.

[0095] Here, if the switching boundary value is assumed to be the transmittance differential value and the transmittance corresponding to an absorbance of 0.3622 where the absorbance differential value coincides, the differential values of both become 1 and coincide, so the sensitivity becomes continuous. On the other hand, when the switching boundary value is the transmittance corresponding to an absorbance of 0.2, the differential values of both do not coincide, so it is necessary to multiply the transmittance difference value in Equation (11) by approximately 0.7 so that they coincide. (Refer to the curve of ΔTRate×0.7)

[0096] Thus, when switching the transmittance difference and the absorbance difference, it is important to ensure sensitivity continuity by multiplying Equation (11) or Equation (12) by an appropriate coefficient so that the transmittance differential value and the absorbance differential value coincide at the switching boundary value, and using the sum of their squares as the objective function.

[0097] In the above, a method for switching the transmittance difference and the absorbance difference that is easy to assume from an optical meaning within the objective function was exemplified. However, a more general method is to prepare a weight function for obtaining a larger absorbance value in the low absorbance region, and substitute the value obtained by substituting the value of " 10 (v 1s / v B1 )" in Equation (12) into the weight function to calculate the obtained value (weight coefficient), and use this value as the difference in Equation (12) "-log 10 (tj1 a 1j / a sum ) + log 10 (v 1s / v B1 )」 is multiplied by, which is the method.

[0098] The above weighting method is for the measurement wavelength i as "-log 10 (T REF T a ij / a sum ) + log 10 (v is / v Bi )」 (where a sum is the sum (j = 1~n) of the components (a ij ) at the measurement wavelength i.) Calculate the difference represented by, and according to the absorbance of the reference sample at the measurement wavelength i, perform weighting on the difference, and optimize the component (a ij ) such that the sum of the squares of the weighted differences is minimized within the solution range of the component, thereby estimating the light quantity distribution component a ij . This is based on the idea.

[0099] Note that in the low absorbance range, the method of switching from the absorbance difference in Equation (12) to the transmittance difference in Equation (11) is an example of this weighting method. The process of weighting the difference calculation value in the objective function is more suitable for this application.

[0100] <Response when the light quantity distribution fluctuates> The light quantity distribution parameter A is a parameter representing the light quantity distribution of the spectrometer itself. If the light quantity distribution of the spectrometer itself fluctuates, the accuracy of spectrum calibration will decrease, so it is necessary to correct (re - estimate) the light quantity distribution parameter A.

[0101] For example, even in the case of a calibrated spectrometer, the shape of the spectrum due to blank measurement may vary. As a result, the user can recognize that the light quantity distribution parameter A, particularly the true light quantity lj for each wavelength before spectroscopy, has changed. Not only for the user, but the computer of the spectrometer may be configured to determine whether there is a change in the shape of the spectrum due to blank measurement, and when a change is recognized, the computer may automatically update the light quantity distribution parameter A.

[0102] The variation in the light quantity distribution of the spectrometer is, for example, that the light quantity distribution of the incident light itself to the sample varies. As its cause, (1) for short-term variations, for example, the "solvent" for sample dissolution changes, and (2) for long-term variations, for example, "optical elements" and "light sources" deteriorate.

[0103] As an example, when the spectrometer is configured as a liquid chromatography detector, the sample dissolution solvent (also called the mobile phase) is selected according to each sample so that the separation degree between the samples to be analyzed is high and the elution is completed in a short time. Therefore, often the transmittance of the mobile phase itself varies greatly, causing variation in the light quantity distribution of the spectrometer.

[0104] In the case of short-term variations, in the above-described method for estimating the light quantity distribution parameter A (Equations (2) to (8)), measurement of the transmitted light quantity spectra of a plurality of reference samples is required. Therefore, performing these measurements frequently is very laborious and not practical.

[0105] Therefore, in the second embodiment, a method for calibrating the spectrometer based on a model constructed on the premise that the light quantity distribution parameter A of the spectrometer changes from the beginning so as to save such labor will be described, that is, an update of the light quantity distribution parameter A will be described.

[0106] · "DL = V B " Model derivation In the "measurement in the state without a sample" (generally called blank measurement), all the light passing through the sample chamber is detected by the detector. However, the distribution wavelength j component of the light passing through the sample chamber is not necessarily detected as the corresponding measurement wavelength i component, and stray light of the distribution wavelength j (j ≠ i) is also included in the measurement wavelength i component.

[0107] Therefore, in the present embodiment, the light amount for each distribution wavelength j of the light passing through the sample chamber is represented by " j " (j = 1 to n), and this is called the "true light amount by wavelength before spectroscopy". For example, the light source spectrum may be applied to " j ". Further, when light passes through the sample chamber or the diffraction grating G without a sample and is detected by the detector, as a parameter indicating the ratio of the light amount of the true light amount by wavelength l j of the light passing through the sample chamber distributed in the measurement wavelength range (i = 1 to m) of the detector, a light amount ratio distribution parameter D is introduced, and its component is represented by " ij ". Then, the light amount (v Bi ) of the measurement wavelength i measured in the state without a sample is the light amount (d j ) represented by "true light amount by wavelength (l ij ) × the ratio (d ij ) at which the light amount is distributed to the measurement wavelength i", and the sum of this light amount (d j ) for all distribution wavelengths (j = 1 to n).

[0108] For example, the light amount v B1 measured as a blank for the first measurement wavelength (i = 1) is the light amount (d j ) represented by "light amount of the distribution wavelength (l 1j ) × the ratio (d 1j ) at which it is distributed as the light amount of the first measurement wavelength", and is the sum of the light amount (d j ) of the light amount (d B1 ) of the light amount (d 1j ) of the light amount (d j ) of the light amount (d Bi ) of the measurement wavelength "v

[0109]

Equation

[0110] D: Light quantity ratio distribution parameter of the spectrometer (d ij ≥ 0) L: True light quantity spectrum before spectroscopy (l j ≥ 0) V B : Light quantity spectrum measured in the blank measurement (Also called the blank light quantity spectrum. Component v Bi )

[0111] When expanding the "DL = V B " model introduced as above, the component (v Bi ) of the blank light quantity spectrum is v B1 = d 11 l1 + d 12 l2 + … + d 1n l n v B2 = d 21 l1 + d 22 l2 + … + d 2n l n ··· v Bm = d m1 l1 + d m2 l2 + … + d mn l n It becomes like this. On the other hand, based on Equation (2) of the above-mentioned "AT = V" model, if we similarly represent the "state without a sample", the magnetic permeability (t1t2 … ) for each distribution wavelength j of the sample becomes (1 1 … ), and the measured light quantity (v i ) measured in the state without a sample is v B1 = a 11 + a 12 + … + a 1n v B2 = a 21 + a 22 + … + a 2n ··· v Bm = a m1 + a m2 + … + amn This results in. Comparing each term in the expansion formulas of both models, the relationship among the light quantity distribution parameter A, the light quantity ratio distribution parameter D, and the true light quantity spectrum L is expressed by Equation (14).

[0112]

Number

[0113] Extracting the corresponding column elements from both sides is shown in Equation (15).

[0114]

Number

[0115] That is, the component a of the light quantity distribution parameter ij and the component d of the light quantity ratio distribution parameter ij are in a proportional relationship, and the true light quantity l by wavelength j is the proportionality constant. The characteristic point of the "DL = V B " model is that the distributed light quantity a i of the distributed wavelength j contained in the light quantity v ij (stray light component) at the measurement wavelength is proportional to the total light quantity (true distributed wavelength light quantity l j ) at that distributed wavelength j. Thus, the "DL = V B " model is a model that conforms to the characteristics of stray light generated by reflection and scattering of the components of the spectrometer (such as the housing wall, slit, detector, window plate, etc.) (the reflectivity at the reflecting surface and the scattering rate at the scattering surface for a certain wavelength are always constant). Therefore, as shown in Equation (16), the component d of the light quantity ratio distribution parameter ij is obtained by dividing each element in the j column of the light quantity distribution component a ij by the true light quantity l by wavelength before spectroscopy j . Equation (16) shows the formula for calculating the light quantity ratio distribution parameter D from the light quantity distribution parameter A.

[0116]

Number

[0117] Now, the column elements of the light quantity distribution parameter A on the left side of Equation (15) represent how much monochromatic light of distribution wavelength j is distributed in the measurement wavelength range (i = 1 to m) of the detector when the monochromatic light of distribution wavelength j is spectroscopically measured without a sample (a 1j … a ij … a mj ). These light quantities (a 1j … a ij … a mj ) can be obtained by repeating spectroscopic measurement n times using monochromatic light of distribution wavelengths (j = 1 to n) as in the prior art.

[0118] When monochromatic light of distribution wavelength j is introduced into the spectrometer, the detected light quantity is the sum of the light quantity measured as the light quantity of the monochromatic light at the measurement wavelength (the same measurement wavelength i as the distribution wavelength j) and the light quantity measured as the light quantity of each measurement wavelength other than the monochromatic light (measurement wavelength i different from the distribution wavelength j, that is, stray light). Therefore, the ratio of the total light quantity of distribution wavelength j at measurement wavelength i to the total light quantity of distribution wavelength j can be obtained by "light quantity of distribution wavelength j at measurement wavelength i / sum of distribution wavelength j". However, it is desired that these ratios can be obtained without introducing monochromatic light into the spectrometer.

[0119] A renewal program for renewing the light quantity distribution parameter A according to light quantity fluctuations by using the light quantity ratio distribution parameter D (d ij =a ij / l j ) without using monochromatic light will be described using the flow of FIG. 7. Here, the existing light quantity distribution parameter before the fluctuation is distinguished as A, and the renewed light quantity distribution parameter is distinguished as A NEW . Similarly, the blank light quantity spectra before and after the fluctuation are distinguished as V B , V B NEW , and the true light quantity spectra before and after the fluctuation are distinguished as L, L NEW .

[0120] Step S21: The blank light quantity spectrum V obtained by blank measurement of the spectrometer to be calibratedB Determine whether there is a change in Step S22: V B If there is a change in j (=Σa ij , where i = 1~m) to obtain the true light quantity l by wavelength from the light quantity distribution parameter A before the change ij =a ij / l j ) to calculate the component (d Step S23: Blank light quantity spectrum V after the change B NEW Based on NEW (=D -1 V B NEW ) to calculate the true light quantity spectrum L after the change Step S24: True light quantity spectrum L after the change NEW Based on the true light quantity spectrum L after the change NEW (a ij =d ij l j ) to calculate the light quantity distribution parameter A after the change Step S25: Light quantity distribution parameter A after the change NEW Register (update) the light quantity distribution parameter A after the change in the memory. Or, use the light quantity distribution parameter A after the change NEW to calculate the measurement spectrum of the sample

[0121] Here, the light quantity ratio distribution parameter D is a matrix with very large diagonal elements, and the condition number of the matrix indicating the calculability of the inverse matrix D -1 is at most several hundred even at its largest, and in the application of this embodiment, the inverse matrix D -1 can be calculated with sufficient accuracy without problems. Also, the light quantity ratio distribution parameter D hardly changes and can be used for a long time. Therefore, once an appropriate light quantity ratio distribution parameter D is set, even if there is a change in the light quantity distribution, in step S23, "L NEW =D -1 V BNEW Based on the formula of "", the true light quantity spectrum L after the change can be calculated. Further, in step S24, "a NEW =d ij =d ij l j Based on the formula of "", the light quantity distribution parameter A after the change can be obtained. The light quantity distribution parameter A updated in this way NEW By directly calculating the measurement spectrum of the sample using, a transmittance spectrum T close to the true value can be obtained. NEW

[0122] In the update program of FIG. 7, the true light quantity spectrum L before spectroscopy is calculated. However, the true light quantity spectrum L is divided into the light quantities of each measurement wavelength, and since the light quantity distribution parameter A can directly correct the measurement spectrum including stray light, the true light quantity spectrum L itself cannot be directly used for correction. The true light quantity spectrum L is positioned as an intermediate variable necessary for optical understanding.

[0123] The characteristic point of the "DL = V B " model of this embodiment is that, as shown in formula (14), the distributed light quantity a of the distributed wavelength j included in the light quantity of a certain measurement wavelength i ij is proportional to the true wavelength-specific light quantity l at that distributed wavelength j (d j = a ij / l ij ). This is also consistent with the fact that stray light is generated due to reflection and scattering of the components of the spectrometer (such as the housing wall, slit, detector, window plate, etc.). That is, this model is constructed in accordance with the characteristics of stray light. The reflectivity and scattering rate of the components of the spectrometer are considered difficult to change even after a long time due to their characteristics, and the light quantity ratio distribution parameter D of the spectrometer can be used for a long time. Therefore, by introducing the light quantity ratio distribution parameter D into the model of this embodiment, even if fluctuations in the light quantity distribution occur due to deterioration of the light source and optical elements or changes in the sample dissolution solvent, the light quantity distribution parameter A after these changes can be easily calculated. j

[0124] ​​If the accuracy of the light quantity ratio distribution parameter D is suspected, re-investigation of the light quantity ratio distribution parameter D is necessary. However, by re-executing the estimation program described in the first embodiment, that is, the measurement of a plurality of reference samples and the derivation of the light quantity distribution parameter A based on the quadratic programming method, the light quantity ratio distribution parameter D can be accurately obtained. Therefore, as long as the reference sample can be prepared, it is possible to obtain the light quantity ratio distribution parameter D even at the spectrometer installation location (onsite). Such operation is not possible with the conventional method of deriving the light quantity distribution parameter A using monochromatic light.

[0125] <Application of the "DL = V B " model to liquid chromatography> The baseline correction method for elution peaks will be described. In liquid chromatography, in order to improve the analysis efficiency, it is frequently performed (also called the gradient method) to change the elution conditions by changing the composition of the mobile phase (referring to the sample dissolution solvent) over time. At this time, the transmittance of the mobile phase fluctuates in a short time, and the light quantity distribution (or the true light quantity spectrum L) fluctuates. In addition, liquid chromatography also has a function of measuring the transmittance spectrum at the elution peak and performing a library search on the transmittance spectrum. That is, the spectrum measurement may be executed while the transmittance of the mobile phase changes over time.

[0126] Here, an example of applying the "DL = V B " model to the calibration of various detectors for liquid chromatography (spectrometers constituting ultraviolet-visible spectrometric detectors, PDA detectors, circular dichroism detectors, etc.) will be described.

[0127] As described above, since the spectrum measurement at the elution peak is often executed while the composition ratio of the mobile phase changes over time, the light quantity distribution is different at the timing of measuring each spectrum. Therefore, in order to execute the spectrum measurement using the light quantity distribution parameter A, the same light quantity distribution parameter A cannot be used, and it is necessary to recalculate the light quantity distribution parameter A corresponding to the light quantity distribution for each timing of the spectrum measurement, that is, to correct the baseline of the elution peak.

[0128] Using a three-dimensional (3D) chromatogram composed of rows and columns, the direction to rows 1 to m is represented as the retention time, and the direction to columns 1 to n is represented as the absorbance spectrum. For example, the first row indicates the absorbance spectrum at the start of measurement. Let the matrix of the 3D baseline absorbance chromatogram when only gradient elution is performed without injecting a certain sample be Bti (t: retention time, i: measurement wavelength). Let the matrix of the 3D absorbance chromatogram obtained when a certain sample is injected and gradient elution is performed be Gti (the subscripts t and i are the same as above). The 3D chromatogram obtained by subtracting Bti from Gti represents the 3D absorbance chromatogram Cti of the sample itself (the subscripts are the same as above) and is expressed by the following formula. Cti = Gti - Bti ···(17)

[0129] The chromatogram in Fig. 8 (the part from t1 to t1 is shown as a solid line, t1 to tn is shown as a dotted line, and the part after tn is shown as a solid line) is obtained by extracting the absorbances at measurement wavelengths 1, 2, ···, m from the 3D absorbance chromatogram Gti. The dotted line in Fig. 8 exemplifies the 3D absorbance chromatogram Cti of the sample itself, and the solid line exemplifies the 3D baseline absorbance chromatogram Bti. At the retention time t0, perform baseline light quantity spectrum measurement and absorbance auto-zero. For single-wavelength detection, perform measurement wavelength scanning at times t1 to tn, measure the absorbance spectrum, and save it in the memory. For a PDA detector, extract the absorbance spectra at times t1 to tn.

[0130] In the "AT = V" model, the terms required to obtain the calibrated absorbance spectrum T are the light quantity distribution parameter A and the transmitted light quantity spectrum V. First, A is the light quantity distribution parameter at time ti and needs to be calculated each time. Specifically, (1) Extract the baseline absorbance spectrum at time ti from the 3D baseline absorbance chromatogram Bti. (2) Calculate the baseline light quantity spectrum at time ti from the baseline light quantity spectrum before baseline variation (at time t0) and the absorbance spectrum at time ti extracted from the 3D baseline absorbance chromatogram Bti. This corresponds to V B in the "DL = V B " model. (3) Calculate L from L = DV B -1 , and obtain the light quantity distribution parameter A NEW at that time ti using a ij = d ij l j . In this document, A NEW is also referred to as the updated value of the light quantity distribution parameter. NEW

[0131] Next, V is the transmitted light quantity spectrum after sample absorption at time ti and needs to be calculated each time. Specifically, (4) Extract the sample absorbance spectrum at time ti from the 3D absorbance chromatogram Cti of the sample itself. This sample absorbance spectrum is the value before calibration. (5) Obtain the transmitted light quantity spectrum V at time ti from the "baseline light quantity spectrum (V B ) at time ti" calculated in (2) above and the "sample absorbance spectrum at time ti" extracted in (4) above.

[0132] Using A NEW and V at each time ti obtained as described above, calculate the calibrated transmittance spectrum T at each time ti from "T = A NEW -1 V", and obtain the calibrated absorbance spectrum S by converting this. This step can be carried out after the peak elution is completed, or it can also be carried out after all measurements are completed. The calibration method may be selected according to the intended use. If the 3D baseline absorbance chromatogram Bti is not measured, the 3D absorbance chromatogram Cti of the sample itself may be obtained by interpolating linearly or curvilinearly, for each wavelength, the absorbance spectrum at the sample elution start time t1 and the absorbance spectrum at the sample elution end time tn of the 3D absorbance chromatogram Gti obtained by gradient elution measurement, and using the resulting spectrum as the baseline absorbance spectrum Bti.

[0133] In this way, by the measurement method using the light quantity ratio distribution matrix D, even when the mobile phase solvent composition changes and the transmitted light quantity spectrum of the baseline fluctuates, efficient calibration of the absorbance spectrum becomes possible, and a more robust absorbance spectrum can be measured.

[0134] When the computer determines whether there is a variation in the light quantity distribution parameter A, and if a variation is recognized, it automatically updates the light quantity distribution parameter A, and uses the updated A NEW to obtain the absorbance spectrum, as shown in the flowchart of FIG. 9.

[0135] Step S31: Read the light quantity ratio distribution parameter D from the memory of the spectrometer. Step S32: Measure the light quantity spectrum of the sample dissolution solvent (blank light quantity spectrum V B ) with the spectrometer at time ti. Step S33: Determine whether the blank light quantity spectrum V B of the sample dissolution solvent is the same (no variation) or not the same (with variation). Step S34: If "no variation" is determined in S33, use the same light quantity distribution parameter A as the previous time to calculate the absorbance spectrum. Step S35: If "with variation" is determined in S33, newly calculate the light quantity distribution parameter A B from the blank light quantity spectrum V NEW at time ti. Step S36: Use the updated light quantity distribution parameter A NEW to calculate the absorbance spectrum.

[0136] A third embodiment of the method for measuring the transmittance or absorbance of the present invention will be described.

[0137] In the above application example (corresponding to the case where the baseline light source spectrum fluctuates), the usefulness of the light quantity ratio distribution parameter D was explained. Here, it will be explained that the above method can be simplified when obtaining the corrected value of the transmittance / absorbance.

[0138] The basic formula "AT = V" can be rewritten as Equation (18) using the light quantity ratio distribution parameter D and the true light quantity spectrum L. Here, the true light quantity spectrum L is represented by a diagonal matrix having the component lj as the diagonal component, and this is denoted as L DIAGONAL for simplicity.

[0139]

Equation

[0140] Equation (19) is the equation obtained by multiplying both sides of Equation (18) from the front by the inverse matrix of DL DIAGONAL in order to obtain the corrected transmittance spectrum T.

[0141]

Equation

[0142] Here, using the property of the inverse matrix "(AB) -1 = B -1 A -1 " and the matrix combination rule, Equation (20) is obtained from Equation (19). The meaning of the spectral correction by Equation (20) in the photometer is that by multiplying the light quantity spectrum V (vector) measured by a spectrometer with stray light by the inverse matrix D -1 of the light quantity ratio distribution parameter and then multiplying it by a matrix having the reciprocal of the true light quantity spectrum L (vector) as the diagonal elements, the corrected transmittance spectrum T can be obtained. Here, the important point is that, for example, at the transmittance of the corrected first measurement wavelength, the reciprocal (1 / l j ) of the first component of the true light quantity spectrum L serves as a scalar coefficient.

[0143]

Number

[0144] The measured light quantity spectrum V of the sample dissolution solvent (blank light quantity spectrum V B ) is defined as "TA" for the transmittance spectrum calibrated by formula (20), and the measured light quantity spectrum V of a certain sample at a certain concentration dissolved in the sample dissolution solvent is defined as "TB" for the transmittance spectrum calibrated by formula (20).

[0145] To simplify the explanation, focus on the first measurement wavelength. In this book, matrix and vector division may be denoted by the symbol " / " (for example, the following "TB1 / TA1", etc.), but the symbol " / " indicates the division of the elements of the matrix and vector. The transmittance at the first measurement wavelength can be calculated by TB1 / TA1 (the subscript indicates the element position of each transmittance spectrum). What this means is that since "1 / l1" in formula (20) appears in both the numerator (TB1) and the denominator (TA1), they cancel each other out.

[0146] Here, we focused on the first measurement wavelength, but the same can be said for any measurement wavelength from the structure of the formula. From this, when calculating the ratio of the transmittance obtained by formula (20), that is, the transmittance of a certain sample at a certain concentration, it can be considered that the multiplication of the matrix (the first term) with the reciprocal of the true light quantity spectrum (1 / l j ) as diagonal elements is omitted. Or, looking at it another way, even if l j is set to any value greater than 0, it can also be considered that the calibrated transmittance spectrum T can be obtained.

[0147] Based on the above relationship, a method for obtaining the calibrated transmittance spectrum T of the sample in a certain sample dissolution solvent based on formula (21) is shown.

[0148] T={(DL DIAGONAL )-1 V} / {(DL DIAGONAL ) -1 V B} ···(21) Here, the " / " in the formula means division between elements. T: Calibrated transmittance spectrum of the sample V: Measured transmitted light quantity spectrum of the sample V B : Measured blank light quantity spectrum D: Light quantity ratio distribution parameter L DIAGONAL : Diagonal matrix with arbitrary values greater than 0 as diagonal elements

[0149] The term DL in formula (21) DIAGONAL can be the light quantity distribution parameter A with l j taken as the true light quantity spectrum of the spectrometer to be calibrated, or if l j is assumed to be 1, the term of L DIAGONAL becomes the identity matrix and can be omitted, so it can also be the light quantity ratio distribution parameter D.

[0150] The light quantity ratio distribution parameter D can be calculated from the light quantity distribution parameter A before the change. As described above, if the transmittance spectrum T after the change can be calibrated using the light quantity distribution parameter A before the change, the process of calculating the light quantity distribution parameter A after the change can be omitted, resulting in simplification and no unnecessary calculations, so an improvement in calculation accuracy can also be expected.

[0151] The above-mentioned "DL = V for liquid chromatography" BIn "Model Application", the calibration of the transmittance spectrum during gradient elution where the sample dissolution solvent changes over time was described. In these calibration steps, it was necessary to update the light intensity distribution parameter A for each measurement time from the light intensity ratio distribution parameter D, which was cumbersome. However, if the knowledge that "the transmittance spectrum T after the change can be calibrated using the light intensity distribution parameter A before the change" is used, it becomes possible to calibrate the transmittance spectrum T by using the light intensity ratio distribution parameter D or the same light intensity distribution parameter A regardless of the measurement time, and simplification can be achieved.

[0152] The absorbance spectrum S can be obtained by taking the common logarithm in Equation (21) and taking the difference as shown in Equation (22).

[0153] S: -log 10 (T) = -log 10 ((DL DIAGONAL ) -1 V) + log 10 ((DL DIAGONAL ) -1 V B ) ···(22) Here, log 10 (x) means the common logarithm of each element of the matrix x. S: Calibrated absorbance spectrum of the sample T: Calibrated transmittance spectrum of the sample V: Measured transmittance light intensity spectrum of the sample V B : Measured blank light intensity spectrum D: Light intensity ratio distribution parameter L DIAGONAL : Diagonal matrix with arbitrary values greater than 0 as diagonal elements

[0154] The measurement flow of the transmittance spectrum T of this embodiment is shown in Fig. 10(A), and the measurement flow of the absorbance spectrum S is shown in Fig. 10(B).

[0155] The measurement flow of the transmittance spectrum T is as follows. Step S41: Obtain the true light quantity l for each wavelength from the existing light quantity distribution parameter A before the change, and calculate the light quantity ratio distribution parameter D (d j =a ij =a ij / l j ). (D calculated in advance may be read.) Step S42: Using the light quantity ratio distribution parameter D, the blank light quantity spectrum V measured after the change of A B , the transmitted light quantity spectrum V of the sample measured after the change of A, and the diagonal matrix L with arbitrary values greater than 0 as diagonal elements DIAGONAL , calculate T={(DL DIAGONAL ) -1 V} / {(DL DIAGONAL ) -1 V B} to obtain the calibrated transmittance spectrum T of the sample after the change of A.

[0156] Also, the measurement flow of the absorbance spectrum S is as follows. Step S51: Obtain the true light quantity l for each wavelength from the existing light quantity distribution parameter A before the change, and calculate the light quantity ratio distribution parameter D (d j =a ij =a ij / l j ). (D calculated in advance may be read.) Step S52: Using the light quantity ratio distribution parameter D, the blank light quantity spectrum V measured after the change of A B , the transmitted light quantity spectrum V of the sample measured after the change of A, and the diagonal matrix L with arbitrary values greater than 0 as diagonal elements DIAGONAL , calculate S=-log 10 ((DL DIAGONAL ) -1 V)+log 10 ((DL DIAGONAL ) -1 V B ) to obtain the calibrated absorbance spectrum S of the sample after the change of A.

[0157] <Example> As a specific application example of the "AT = V" model of the present invention, an example of a method for estimating the light quantity distribution parameter A using a spectrum in a measurement range of 190 to 700 nm with a wavelength step of 4 nm using an actual spectrometer, and an example of a method for measuring an absorbance spectrum using the light quantity distribution parameter A with an actual spectrometer are shown below.

[0158] 1. Estimation of light quantity distribution parameter A 1.1. Acquisition of transmitted light quantity spectrum V in AT = V First, select a plurality of types of reference samples having transmittance spectra with high mutual independence. In this example, six types of reference samples were used. From the conditions of the measurement range (190 to 700 nm) and the wavelength step (4 nm), since the number of wavelength points is 128, the same 128 transmittance spectra are required. Therefore, it was decided to acquire transmittance spectra of 21 to 22 types of concentrations for each reference sample. In this example, using a gradient liquid delivery system capable of delivering only the dissolution solvent (blank) or a mixed solvent between the reference sample and the dissolution solvent, with an actual spectrometer, the change in the reference sample concentration of the mixed solvent was measured, and the transmittance spectrum (and transmitted light quantity spectrum V) for each concentration was acquired, and based on this, the absorbance spectrum for each concentration was also acquired.

[0159] The concentration of the reference sample indicated by the acquired transmittance spectrum is determined as follows. First, select a wavelength range where the absorbance is relatively low in the absorbance spectrum of the reference sample. If possible, it is desirable that the absorbance is 0.5 or less even at the highest concentration. Then, based on the proportional relationship between the measured absorbance spectrum of the reference sample and the known absorbance spectrum of the reference sample in the selected wavelength range, the concentration of the reference sample indicated by the acquired transmittance spectrum can be easily determined.

[0160] Figure 11 shows a concentration gradient absorbance curve (also called a concentration gradient chromatogram) obtained by extracting the change in absorbance over time at 306 nm, which has a relatively low absorbance, from the 3D chromatogram of the coloring agent "Food Blue No. 1" (Brilliant Blue FCF) of the reference sample 1 measured with an actual spectrometer after connecting a gradient liquid delivery system. In this example, in order to exclude the influence of baseline fluctuations, the point where the retention time Rt on the horizontal axis is 1.5 min was set as the starting point, and the point where it is 28.5 min was set as the ending point, and the interval between them was defined as the baseline interval. Then, the transmittance spectra (and the transmitted light amount spectrum V) at a total of 22 points, including the starting point (1.5 min) of the baseline interval and 21 points evenly divided in the starting - ending interval from 3.0 min to 9.0 min, were extracted from the above 3D chromatogram. The reason for adding the starting point (1.5 min) of the baseline interval is to obtain the light amount spectrum during blank measurement.

[0161] 1.2. Acquisition of the transmittance spectrum T at AT = V According to the concentration series for each reference sample obtained above, a known transmittance spectrum concentration series was acquired, and these were used as the transmittance spectra T.

[0162] 1.3. Estimation of the light amount distribution parameter A Figure 12 shows, in a bird's - eye view, the light amount distribution parameter A obtained by the quadratic programming method shown in "Estimation of Parameter A by Quadratic Programming Method of the Present Invention" using the transmitted light amount spectrum V and the transmittance spectrum T acquired above. The vertical axis in Figure 12 is the measurement wavelength i, and the horizontal axis is the distribution wavelength j. The intensity at the coordinates (i, j) of the bird's - eye view corresponds to the light amount distribution component a of the estimated light amount distribution parameter A ij and is displayed on a common logarithm (Log 10 ) scale. Also, Figure 13 shows, in a bird's - eye view, the light amount ratio distribution parameter D used in "Update of Parameter A of the Present Invention" using the same vertical axis and horizontal axis as in Figure 12. The intensity at the coordinates (i, j) of the bird's - eye view corresponds to the light amount ratio distribution component d of the light amount ratio distribution parameter D ij and is displayed on a common logarithm scale.

[0163] 2. Calibration of Absorbance Spectrum FIG. 14 is a graph for comparing the known absorbance spectrum of reference sample 1 (diagram indicated by “known”), the absorbance spectrum measured by an actual spectrometer (diagram indicated by “actual spectrometer measurement”), and the absorbance spectrum calibrated using existing parameter A according to the present invention (points indicated by “after calibration”). As shown in FIG. 14, the absorbance spectrum calibrated by the application of the present invention has substantially the same spectral shape as the known absorbance spectrum obtained from a high-precision spectrometer, except at 190 nm and 194 nm.

[0164] FIG. 15 is an enlarged graph with the wavelength axis of FIG. 14 set to 600 - 700 nm. In the absorbance spectrum measured by the actual spectrometer, distortion of the absorbance spectrum (the portion enclosed by the ellipse in FIG. 15) is observed in the wavenumber range of 620 - 660 nm due to the influence of the 656 nm emission line derived from a deuterium lamp (D2 light source), but the calibrated absorbance spectrum is substantially in agreement with the known absorbance spectrum.

[0165] From these facts, it can be seen that the calibration of the absorbance spectrum according to the present invention is performed with high accuracy.

[0166] Finally, FIG. 16 is a graph showing the linearity of concentration - absorbance at 630 nm, which is the spectral peak top of reference sample 1. Based on the absorbance spectrum measured by the actual spectrometer, the linearity gradually deteriorates from around where the absorbance exceeds 0.4, whereas it can be seen that the linearity is ensured based on the calibrated absorbance spectrum according to the present invention until the absorbance reaches about 2.0.

Explanation of Signs

[0167] 10 Light source 20 Sample chamber 30 Dispersion element 40 Detector 50 Arithmetic unit 60 Memory 70 Estimation unit 80 Memory

Claims

1. Let \(T\) be the transmittance spectrum composed of the transmittance \(t\) of the sample for each wavelength, and let \(V\) be the transmitted light quantity spectrum composed of the transmitted light quantity \(v\) of the sample for each wavelength measured by a spectrometer. Let \(A\) be the light quantity distribution parameter which is a matrix composed of the light quantity distribution component \(a\) ｉｊ peculiar to the spectrometer. A method for measuring transmittance or absorbance using a model formula expressed as \(AT = V\). The light quantity distribution component a ｉｊ is the amount of light per wavelength before splitting in the blank measurement of the spectrometer. ｊ (j indicates the distribution wavelength before splitting) is a component indicating the amount of light distributed in each band of the measurement wavelength i of the detected light after splitting, and is obtained in advance. A method for measuring transmittance or absorbance, characterized by the following.

2. The method for measuring transmittance or absorbance according to Claim 1, further A transmittance spectrum matrix T composed of known transmittance spectra of a plurality of types of reference samples ＲＥＦ , and a transmitted light quantity spectrum matrix V composed of transmitted light quantity spectra of the plurality of types of reference samples measured by the spectrometer ＲＥＦ For each input value of ＲＥＦ , the light quantity distribution component a ＲＥＦ that satisfies the model formula "AT ｉｊ = V ＲＥＦ " is estimated in advance using a computer A method for measuring transmittance or absorbance, characterized by the following.

3. The method for measuring transmittance or absorbance according to Claim 2, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, the component (a ＲＥＦ -V ＲＥＦ ) for which the magnitude of the vector represented by "AT ｉｊ " is minimized is optimized within the setting range of the solution of the component, and the measurement method of transmittance or absorbance is characterized by a quadratic programming method of obtaining it.

4. The method for measuring transmittance or absorbance according to Claim 2, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, For the measurement wavelength i, the component (a ｓｕｍ ), such that the magnitude of the vector represented by "(A / a ＲＥＦ )T ＲＥＦ -(V Ｂｉ / v ｉｊ )" is minimized, is obtained by optimizing within the setting range of the solution of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, and v Ｂｉ is the measurement light quantity of the blank measurement at the measurement wavelength i.

5. The method for measuring transmittance or absorbance according to Claim 2, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, and For the measurement wavelength i, the magnitude of the vector represented by "-log １０ {(A / a ｓｕｍ ) T ＲＥＦ}+ log １０ (V ＲＥＦ / v Ｂｉ )" is minimized, and the component (a ｉｊ ) is obtained by optimizing within the solution setting range of the component, by means of the quadratic programming method as described above. A method for measuring transmittance or absorbance, characterized by the following. Here, a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, and v Ｂｉ is the amount of measurement light in the blank measurement at the measurement wavelength i.

6. The method for measuring transmittance or absorbance according to Claim 2, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For a reference sample with a relatively low absorbance at the measurement wavelength i, the difference represented by “(T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) − (v ｉｓ / v Ｂｉ )” is calculated for the measurement wavelength i. For a reference sample with a relatively high absorbance at the measurement wavelength i, the difference represented by “−log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )” is calculated for the measurement wavelength i, The component (a) that minimizes the sum of the squares of these differences is obtained by optimizing within the solution range of the component, according to the quadratic programming method ｉｊ ). A method for measuring transmittance or absorbance, characterized by the following. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, s is the number of reference samples (s = 1 to p), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

7. The method for measuring transmittance or absorbance according to Claim 2, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, calculate the difference represented by "-log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )", According to the absorbance of the reference sample at the measurement wavelength i, perform weighting on the difference, The component (a ｉｊ ) that minimizes the weighted sum of squared differences is obtained by optimizing within the solution range of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum of the components (a ｉｊ ) at the measurement wavelength i (j = 1 to n), s is the number of reference samples (s = 1 to p), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

8. The method for measuring transmittance or absorbance according to Claim 1, further Using a plurality of concentrations of one type of reference sample, a transmittance spectrum matrix T composed of known transmittance spectra of the reference sample for each concentration ＲＥＦ , and for each input value of a transmitted light quantity spectrum matrix V composed of transmitted light quantity spectra of the reference sample for each concentration measured by the spectrometer ＲＥＦ , for the model formula "AT ＲＥＦ = V ＲＥＦ ", estimate the light quantity distribution component a ｉｊ that satisfies it in advance using a computer A method for measuring transmittance or absorbance, characterized by the following.

9. The method for measuring transmittance or absorbance according to Claim 8, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, the component (a ＲＥＦ -V ＲＥＦ ) for which the magnitude of the vector represented by "AT ｉｊ " is minimized is optimized within the setting range of the solution of the component, and the transmittance or absorbance measurement method is characterized by the quadratic programming method of obtaining by doing so.

10. The method for measuring transmittance or absorbance according to Claim 8, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, the component (a ｓｕｍ ), which minimizes the magnitude of the vector represented by "(A / a ＲＥＦ )T ＲＥＦ -(V Ｂｉ / v ｉｊ ))", is obtained by optimizing within the solution setting range of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

11. The method for measuring transmittance or absorbance according to Claim 8, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, For the measurement wavelength i, "−log １０ {(A / a ｓｕｍ ) T ＲＥＦ}+ log １０ (V ＲＥＦ / v Ｂｉ )" is obtained by optimizing the component (a ｉｊ ) such that the magnitude of the vector represented is minimized within the solution setting range of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, a ｓｕｍ is the sum of the components (a ｉｊ ) at the measurement wavelength i (j = 1 to n), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

12. The method for measuring transmittance or absorbance according to Claim 8, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For a reference sample with a relatively low absorbance at the measurement wavelength i, a difference represented by “(T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) − (v ｉｓ / v Ｂｉ )” is calculated for the measurement wavelength i. For a reference sample with a relatively high absorbance at the measurement wavelength i, a difference represented by “−log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )” is calculated for the measurement wavelength i, The component (a ｉｊ ) that minimizes the sum of the squares of these differences is obtained by optimizing within the solution range of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum of the components (a ｉｊ at the measurement wavelength i (j = 1 to n), s is the number of reference samples (s = 1 to p), and v Ｂｉ is the amount of measurement light in the blank measurement at the measurement wavelength i.

13. The method for measuring transmittance or absorbance according to Claim 8, The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, calculate the difference represented by "-log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )", According to the absorbance of the reference sample at the measurement wavelength i, perform weighting on the difference, The component (a ｉｊ ) that minimizes the weighted sum of squared differences is obtained by optimizing within the solution setting range of the component, according to the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum of the components (a ｉｊ ) at the measurement wavelength i (j = 1 to n), s is the number of reference samples (s = 1 to p), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

14. The method for measuring transmittance or absorbance according to Claim 8, The light quantity distribution component a using the reference samples of the plurality of concentrations ｉｊ In the estimation of, a wavelength range with relatively low absorbance is selected from the wavelength range of the transmitted light quantity spectrum of the reference sample for each measured concentration, and the concentration of the reference sample based on the absorbance in the selected wavelength range is determined as the reference concentration. Based on the determined reference concentration, determine the correspondence between the transmitted light amount spectrum of the reference sample for each measured concentration and the transmittance spectrum of the reference sample for each known concentration. A method for measuring transmittance or absorbance, characterized by this.

15. The method for measuring transmittance or absorbance according to Claim 1, further Using a plurality of types of reference samples, and using those with multiple concentrations for each of the reference samples, a transmittance spectrum matrix T composed of known transmittance spectra of the reference samples for each type and each concentration ＲＥＦ , and a transmitted light quantity spectrum matrix V composed of transmitted light quantity spectra of the reference samples for each type and each concentration measured by the spectrometer ＲＥＦ For each input value of ＲＥＦ “AT ＲＥＦ = V ｉｊ ”, estimating in advance, using a computer, the light quantity distribution component a that satisfies the model formula, a method for measuring transmittance or absorbance.

16. The method for measuring transmittance or absorbance according to Claim 15, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, the component (a ＲＥＦ -V ＲＥＦ ) for which the magnitude of the vector represented by "AT ｉｊ -V" becomes minimum is optimized within the setting range of the solution of the component, and the transmittance or absorbance measurement method is characterized by the quadratic programming method of obtaining by doing so.

17. The method for measuring transmittance or absorbance according to Claim 15, the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, For the measurement wavelength i, the component (a ｓｕｍ ), for which the magnitude of the vector represented by "(A / a ＲＥＦ )T ＲＥＦ -(V Ｂｉ / v ｉｊ )" is minimized, is obtained by optimizing within the solution setting range of the component, by means of the quadratic programming method A method for measuring transmittance or absorbance, characterized by the following. Here, a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

18. The method for measuring transmittance or absorbance according to claim 15, wherein the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, For the measurement wavelength i, the magnitude of the vector represented by "-log" １０ {(A / a ｓｕｍ )T ＲＥＦ}+log １０ (V ＲＥＦ / v Ｂｉ )" is minimized, and the component (a ｉｊ ) is obtained by optimizing within the setting range of the solution of the component, according to the quadratic programming method. a method for measuring transmittance or absorbance, characterized by the above. Here, a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, and v Ｂｉ is the amount of measurement light in the blank measurement at the measurement wavelength i.

19. The method for measuring transmittance or absorbance according to claim 15, wherein the light quantity distribution component a ｉｊ is estimated for each measurement wavelength i, For a reference sample with a relatively low absorbance at the measurement wavelength i, calculate the difference represented by “(T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) − (v ｉｓ / v Ｂｉ )” for the measurement wavelength i. For a reference sample with a relatively high absorbance at the measurement wavelength i, calculate the difference represented by “−log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )” for the measurement wavelength i, The component (a) that minimizes the sum of the squares of these differences ｉｊ is obtained by optimizing within the solution range of the component, according to the quadratic programming method a method for measuring transmittance or absorbance, characterized by the above. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum (j = 1 to n) of the components (a ｉｊ ) at the measurement wavelength i, s is the number of reference samples (s = 1 to p), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

20. The method for measuring transmittance or absorbance according to claim 15, wherein The light quantity distribution component a ｉｊ is estimated for each measurement wavelength i and For the measurement wavelength i, calculate the difference represented by "-log １０ (T ＲＥＦ Ｔ a ｉｊ / a ｓｕｍ ) + log １０ (v ｉｓ / v Ｂｉ )". weighting is performed on the difference according to the absorbance of the reference sample at the measurement wavelength i, The component (a ｉｊ ) that minimizes the weighted sum of squared differences is obtained by optimizing within the solution range of the component, according to the quadratic programming method a method for measuring transmittance or absorbance, characterized by the above. Here, T ＲＥＦ Ｔ is the transposed matrix of T ＲＥＦ , a ｓｕｍ is the sum of the components (a ｉｊ ) at the measurement wavelength i (j = 1 to n), s is the number of reference samples (s = 1 to p), and v Ｂｉ is the measured light quantity of the blank measurement at the measurement wavelength i.

21. The method for measuring transmittance or absorbance according to claim 15, wherein The light quantity distribution component a using the reference samples of the plurality of concentrations ｉｊ In the estimation of, a wavelength region with relatively low absorbance is selected from the wavelength range of the transmitted light quantity spectrum of the reference sample for each measured concentration, and the concentration of the reference sample based on the absorbance in the selected wavelength region is determined as the reference concentration, a method for measuring transmittance or absorbance, characterized by determining the correspondence between the transmitted light amount spectrum of the reference sample for each measured concentration and the transmittance spectrum of the reference sample for each known concentration based on the determined reference concentration.

22. The method for measuring transmittance or absorbance according to claim 1, further including a method for updating the light amount distribution parameter A, the updating method being In the blank measurement of the spectrometer, the light quantity l by wavelength before spectroscopy ｊ (l ｊ = a １ｊ + a ２ｊ + ··· + a ｍｊ ), the light quantity spectrum before spectroscopy is defined as L. The light quantity ratio distribution component d ｊ indicating to what extent the light quantity l by wavelength before spectroscopy is distributed in each band of the measurement wavelength i of the detected light after spectroscopy ｉｊ is used to form the light quantity ratio distribution parameter matrix D. From the existing light quantity distribution component a ｉｊ to the light quantity ratio distribution component d ｉｊ (d ｉｊ = a ｉｊ / l ｊ Here, "a ｉｊ / l ｊ " means dividing each element in the j-th column of the light quantity distribution component a ｉｊ by the light quantity l by wavelength ｊ ).) is calculated, Performing a blank measurement of the spectrometer to obtain a blank light quantity spectrum V Ｂ and The light quantity ratio distribution parameter D and the blank light quantity spectrum V Ｂ For each input value of Ｂ a new light quantity spectrum L before spectroscopy that satisfies the model formula "DL = V" ＮＥＷ is calculated by the computer, The light quantity ratio distribution parameter D and the light quantity spectrum L before spectroscopy that is new ＮＥＷ are used to calculate an updated value of the light quantity distribution parameter A ＮＥＷ where (a ｉｊ = d ｉｊ l ｊ ), and a method for measuring transmittance or absorbance, characterized by including this

23. The method for measuring transmittance or absorbance according to claim 22, further The blank light quantity spectrum V obtained by performing a blank measurement using the spectrometer Ｂ and the past blank light quantity spectrum V stored in the memory Ｂ are compared to determine the variation of the blank light quantity spectrum V Ｂ ​ automatically performing the update of the light amount distribution parameter A when there is a variation, a method for measuring transmittance or absorbance, characterized by including the above.

24. The method for measuring transmittance or absorbance according to claim 1, wherein In the blank measurement of the spectrometer, the light quantity l by wavelength before spectroscopy ｊ (l ｊ = a １ｊ + a ２ｊ + ··· + a ｍｊ ) is defined as the light quantity spectrum L before spectroscopy. The light quantity ratio distribution component d ｊ indicating to what extent the light quantity l by wavelength before spectroscopy is distributed in each band of the measurement wavelength i of the detected light after spectroscopy ｉｊ is used to form the light quantity ratio distribution parameter matrix D. From the existing light quantity distribution component a ｉｊ to the light quantity ratio distribution component d ｉｊ (d ｉｊ = a ｉｊ / l ｊ Here, "a ｉｊ / l ｊ " means dividing each element in the j-th column of the light quantity distribution component a ｉｊ by the light quantity l by wavelength ｊ ). Calculate (), Execute the blank measurement of the spectrometer to obtain the blank light quantity spectrum V Ｂ and measuring the sample with the spectrometer to obtain the transmitted light amount spectrum V, Let \(L\) be a diagonal matrix having diagonal elements with any value greater than 0 ＤＩＡＧＯＮＡＬ and input each value of the light quantity ratio distribution parameter \(D\), the blank light quantity spectrum \(V\) ＤＩＡＧＯＮＡＬ and －１ the transmitted light quantity spectrum \(V\) of the sample into the formula of “\(T = \{(DL ＤＩＡＧＯＮＡＬ ) －１ V\} / \{(DL Ｂ ) Ｂ V\}\)” to calculate the transmittance spectrum \(T\) of the sample by the computer, or Let \(L\) be a diagonal matrix having diagonal elements of any value greater than \(0\). ＤＩＡＧＯＮＡＬ “\(S = -\log\)” １０ “\((DL\)” ＤＩＡＧＯＮＡＬ ) －１ “\(V)+\log\)” １０ “\((DL\)” ＤＩＡＧＯＮＡＬ ) －１ “\(V\)” Ｂ “ In the formula, each value of the light quantity ratio distribution parameter \(D\), the blank light quantity spectrum \(V\), and the transmitted light quantity spectrum \(V\) of the sample is input, and an absorbance spectrum \(S\) of the sample is calculated by the computer.” Ｂ “A method for measuring transmittance or absorbance, characterized in that.”

25. The method for measuring transmittance or absorbance according to claim 1, further including selecting whether to execute or not execute a measurement method using the model formula expressed as "AT = V" during the measurement with the spectrometer. A method for measuring transmittance or absorbance, characterized by including the above.

26. A transmittance or absorbance measurement program for causing a computer to execute the procedure of the transmittance or absorbance measurement method according to any one of claims 1 to 25.

27. A spectrometer, characterized by comprising a computer capable of executing the transmittance or absorbance measurement program according to claim 26.

Images