JPH0315741A - Near infrared spectrochemical analysis - Google Patents

Abstract

PURPOSE:To make always stable analysis by determining a quadratic curve by absorbance of arbitrary 3 wavelengths among the respective absorbances obtd. from at least >=3 wavelengths of near IR rays past plural pieces of narrow band filters. CONSTITUTION:The light which is emitted from a light source 31 and is made into collimated beams of light is passed through the narrow band filter 33, by which the light is made into the near IR light of the specific wavelength; thereafter, the direction thereof is changed toward a lighting window 36 by a reflecting mirror 32. The near IR light is reflected by the reflecting mirror 32 and is admitted into an integrating sphere 34 via the lighting window 36, and a sample 55 in a sample container 52 is irradiated with the near IR light from right above. While the diffused and reflected light from the sample 55 is reflected in the integrating sphere 34, the light arrives finally at a pair of detectors 35a, 35b disposed in the positions symmetrical with a measuring part 37 as a center. The intensity of the reflected light is measured in this way. A sensor 79 to measure the temp. of the container 52 is provided in a sample container mounting box 5 in the lower part of the container 52 to correct the analysis value by the sample temp.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides a near-infrared spectroscopic analysis method for analyzing the component values or quality of a sample based on the absorbance obtained by a plurality of near-infrared rays transmitted through a plurality of narrow band filters. Regarding.

[Conventional technology]

In conventional near-infrared spectroscopy methods, so-called fixed-filter near-infrared spectrometers that use narrow band filters are compared to scanning-type near-infrared spectrometers that continuously irradiate wavelengths from 100 Qnm to 2600 nrA. Since only a limited wavelength is irradiated using a narrow band filter, the analysis time can be completed in a short time.

In other words, changes in absorbance due to differences in the quality of samples to be analyzed are limited to a specific wavelength range, and in particular, the wavelengths or wavelength ranges in which differences in sample quality are particularly noticeable are often limited to a plurality of wavelengths. Therefore, when constantly analyzing the same type of sample, a near-infrared spectrometer equipped with a narrow band filter that transmits only the wavelengths at which the difference in sample quality becomes noticeable as a difference in absorbance can be used to measure the absorbance in the required wavelength range. The components or quality of the sample can be analyzed in a short time.

In addition, a near-infrared spectrometer only needs to have multiple narrow-band filters and a device to replace these filters as necessary, and can produce fine and continuous irradiation wavelengths like a scanning type. It did not require any adjustment equipment and could be manufactured at low cost.

By the way, the most important issues at present in analysis using near-infrared rays are the pulverized particle size and temperature of the sample. The temperature of the sample directly affects the moisture content in the sample. Furthermore, the environmental temperature at which the measurement is performed increases the generation of optical and electrical noise and drift of the near-infrared spectroscopy analyzer, causing a significant decrease in the accuracy of analytical values.

Furthermore, as the particle size of the sample becomes smaller, the surface reflectance of the sample increases and the absorbance decreases. Conversely, the larger the particle size, the lower the surface reflectance of the sample and the higher the absorbance.

To deal with this problem, in the scanning near-infrared spectroscopy described above, the absorbance is analyzed using almost continuous wavelengths (2 nm intervals), so the results can be expressed as a continuous spectrum. For example, this continuous spectrum is about 2000
It will be recorded as an array of data divided into ~4000 points.

Using this large amount of data, various processes are first performed, such as averaging processing to eliminate the effects of noise in the spectrum, and shrinkage processing to facilitate mathematical calculations, and the values obtained in this way are used to calculate the It is possible to obtain highly accurate analytical values by using it as a variable in a multiple regression equation with minute values or quality.

[Problem to be solved by the invention]

However, in fixed-type near-infrared spectroscopy, the absorbance obtained from the analyzer is not continuous, but only at a plurality of arbitrary specific wavelengths. Therefore, since it is not possible to obtain a continuous spectrum as in the scanning type near-infrared spectroscopy described above, absorbance cannot be handled in the same way as in the scanning type. Therefore, there is a problem in that it is necessary to stabilize the environment and sample temperature and control the pulverized particle size of the sample using peripheral equipment.

In other words, the technical objective of the present invention is to find a method for near-infrared spectroscopic analysis that is always stable without being affected by the pulverized particle size/particle size of paint, sample temperature, environmental temperature, etc.

[Means to solve the problem]

According to the present invention, when a sample is irradiated with near-infrared light, a near-infrared spectrometer is used that is equipped with at least three narrow band filters that transmit only wavelengths at which differences in quality of the sample become noticeable as differences in absorbance. A near-infrared spectroscopic analysis method, in which quadratic curves are obtained from the absorbances of arbitrary three wavelengths among the respective absorbances obtained from near-infrared rays of at least three wavelengths that have passed through the plurality of narrow band filters, and the In order to solve the above-mentioned technical problem by a near-infrared spectroscopic analysis method that analyzes the quality of the sample by performing multiple regression analysis using each value obtained by quadratic differentiation of the quadratic curve as an explanatory variable of the multiple regression equation. It was used as a means.

[For production]

According to the present invention, when a sample is irradiated with near-infrared light, any three of the absorbances obtained from near-infrared light of at least three wavelengths that transmit only those wavelengths in which the difference in quality of the sample is noticeable as a difference in absorbance are transmitted. The value obtained by second-order differentiation of a quadratic curve that is uniquely determined by the value of the absorbance of a wavelength can be determined without being affected by the particle size of the sample or the temperature between the sample and its surroundings.

When the quadratic curve obtained from the absorbance of the three wavelengths is differentiated to the second order, only the square term remains, and the change in absorbance due to the influence of particle size, particle size, and temperature, which is obtained by the linear function, is calculated by the quadratic curve obtained from the three wavelengths. It is not included in the value obtained by second-order differentiation of the following curve.

Therefore, the quadratic differential value is temporarily determined from the quadratic curve of absorbance at any three wavelengths, and is not affected by sample particle size, particle diameter, or temperature, and acts as an explanatory coefficient of the multiple regression equation. Then, specific coefficients such as quality evaluation coefficients or component conversion coefficients are determined by multiple regression analysis with the quality or component values of known samples, and the results are compared with the values obtained by analyzing the unknown sample with a near-infrared spectrometer with a fixed filter. The value obtained by calculation with the coefficient of It has become possible to perform analysis without

〔Example〕

Hereinafter, embodiments of a near-infrared spectroscopic analyzer according to the near-infrared spectroscopic analysis method according to the present invention will be described with reference to FIGS. 1 to 3 of the accompanying drawings.
This will be explained with reference to the figures.

FIG. 1 is a schematic diagram of a near-infrared spectroscopic analyzer 1 according to the present invention viewed from the front. Inside the cabinet 2, a near-infrared spectrometer 3 and a control device 4, the detailed configuration of which will be explained with reference to FIG. 2 below, are disposed. On the front panel of the cabinet 2, there is a sample container mounting box 5 for mounting a sample container (sample placement section) into which a sample to be measured is placed.
A display device 6 in the form of a light emitting diode or CRT for visually displaying the operation procedure of the apparatus, calculation results, etc., push buttons 7 for operation, and a printer 8 for making a hard copy of the calculation results are provided. The control device 4 is connected to the light source, detector, display device 6, operation push button 7, printer 8, etc. of the near-infrared spectrometer 3, and is an input/output signal processing device 4 for processing various signals! and specific coefficients set individually for each component or quality of the sample in order to calculate the component or quality evaluation value, the various conditions entered through the input device (keyboard) 9, and the specific coefficients for each component or quality. a storage device 4b for storing prices, various corrections, various control procedures, etc.;
It consists of a calculation device 4C for calculating the components or quality evaluation values of the sample based on the measurement values obtained by the near-infrared spectrometer 3 and the specific coefficients. Note that specific coefficients and necessary correction values that are individually set for each sample component or quality evaluation value are stored in advance in a read-only memory (hereinafter referred to as ROM) in the storage device 4b. Further, the printer 8 is not limited to a built-in type, and may be an externally connected type.

By the way, near-infrared rays irradiated onto a sample are absorbed by the sample because the chain of atoms that make up the molecule vibrates due to thermal energy, and the natural frequency differs depending on the type of atoms and the chain state. In addition, the magnitude of vibration changes in the near-infrared wavelength region, causing heat absorption. In addition, if the sample initially has little thermal energy (low m degrees), the amount of absorption due to differences in molecular structure cannot be measured accurately due to small vibrations, so it is necessary to correct the temperature. . Normally, no correction is required when the temperature is 20°C or higher.

The temperature setting device 77 adjusts the near-infrared spectrometer 1 to a constant temperature, and when the temperature is low, it operates the heating device 78 and normally
Set to ℃. This has the purpose of preventing changes in the sample temperature and eliminating errors due to the temperature of the electric circuit.

FIG. 2 is a sectional view of a main part of an example of the near-infrared spectrometer 3 disposed inside the cabinet 2. As shown in FIG. The illustrated near-infrared spectrometer 3 is of a reflection type, and has a light source 31, a reflecting mirror 32, a narrow band pass filter 33, an integrating sphere 34, and detectors 35g and 35b as main components. The light emitted from the light source 31 passes through an appropriate optical system (not shown) and becomes parallel light. After passing through the narrow band pass filter 33, the light becomes near-infrared light of a specific wavelength. A reflecting mirror 32 configured to freely change the direction of the integrating sphere 34 can be used to change the direction toward a lighting window 36 provided by opening the top of the integrating sphere 34. The near-infrared light reflected by the reflecting mirror 32 and entering the interior of the integrating sphere 34 through the lighting window 36 of the integrating sphere 34 is transmitted to the measuring section 37 provided by opening the bottom of the integrating sphere 34, and thus to the sample container. A sample 55 in a sample container 52 located at a predetermined rear position of the mounting box 5 is irradiated from directly above. The diffusely reflected light from the sample 55 is reflected by the integrating sphere 3.
4, a pair of detectors 35m, 35 are finally arranged at symmetrical positions centering on the measuring section 37.
b, from which the intensity of the reflected light is measured. The sample container mounting box 5 below the sample container 52 is provided with a sensor 79 for measuring the temperature of the sample container, and as described in detail regarding the temperature setting device 77, the analysis value is corrected based on the sample temperature. Further, in the illustrated embodiment, a pair of detectors, namely two detectors designated by reference numbers 35a and 35b, are provided in order to correct the optical symmetry and efficiently receive the reflected light from the sample 55. The number of detectors is not limited to two, and may be one or three or more.

Here, the configuration of the narrow band pass filter 33, which is provided between the light source 31 and the reflecting mirror 32, and through which the light emitted from the light source 31 becomes near-infrared light of a specific wavelength, and the requirements for this filter are explained. Explain the physical characteristics etc. The narrow band pass filter 33 is composed of a plurality of arbitrary filters (for example, six filters 331 to 33I) each having a different dominant wavelength pass characteristic, and these are attached to a rotating disk and rotated by an appropriate angle. Accordingly, the light source 31
The structure is such that a desired filter can be selected and replaced in sequence so that it is located on the line connecting the filter and the reflecting mirror 32. Note that in the transmission characteristics of a filter, the dominant wavelength refers to the maximum transmission wavelength of near-infrared rays that are transmitted when the incident optical axis is perpendicular to the filter surface.

Next, the physical characteristics required of the narrow band pass filter 33 will be explained based on FIG. 3. FIG. 3 is a graph (absorbance curve) showing the relationship between irradiation wavelength and absorbance when different samples are irradiated with near-infrared light whose wavelength changes continuously. The absorbance logIo/I is the common logarithm of the reciprocal of the ratio of the amount of reflected light from the sample (■) to the reference amount of irradiated light (total irradiated light amount) 10. It can be easily understood that the amount of content of each of the above components is clearly expressed as a difference in absorbance. Since the present invention utilizes this phenomenon to measure the absorbance of a sample at an arbitrary wavelength, the wavelength of the near-infrared light irradiated onto the sample for measurement is in the wavelength range 110.
Depending on the component or quality of the sample, specific peaks can be seen on the absorbance curve between 0 and 2500 nm (in this example, P.
The absorbance is measured in the wavelength range (am, P2nm-P6nm). Therefore, each of the filters 331 to 33t included in the narrow band pass filter 33 has a specific pass characteristic for each of the wavelengths in order to produce near-infrared light of each of the wavelengths suitable for calculating the amount or quality evaluation value contained in the sample. , that is, it is required to have it as the dominant wavelength.

Next, the specific operation of the infrared spectrometer according to the present invention having the above configuration will be explained. First, the light source 31 is turned on by operating the operating push button 7, and the near-infrared spectrometer 3 The entire area is preheated using a heating device 78 or the like. After a predetermined period of time for preheating has elapsed, the sample container mounting box 5 is pulled out from the cabinet 2 of the apparatus, and the sample container 52 filled with the crushed sample 55 is placed in a predetermined position.
Complete the measurement preparation by inserting the At this time, the temperature sensor 79 of the sample container device box 5 measures the temperature of the sample container 52. Note that, in order to avoid errors in the measured values, it is desirable that the sample 55 be pulverized to a particle size of about 50 microns or less, but it is not necessarily necessary to be pulverized. In addition, in order to reduce the loss of light due to diffuse reflection, the sample 55
The sample container 52 is placed in such a state that its surface becomes a flat surface.
It is preferable that the surface be filled with a transparent glass plate and that the surface be covered with a transparent glass plate while applying some pressure.

When the measurement preparation work is completed, first P. nm
The filter 33A having a main wavelength of Pln is selected to be on the line connecting the light source 31 and the reflecting mirror 32, and
Then, the work begins to measure the reflected absorbance when the sample 55 is irradiated with near-infrared light of m. The measurement work of reflection absorbance involves measuring the total amount of irradiation irradiated onto the sample 55, that is, the reference amount of light, and measuring the amount of reflected light actually reflected by the sample 55 when the sample 55 is irradiated with the reference amount of irradiation light. It consists of two measurements. Although it is possible to perform either of the two measurements on one filter first, in the following explanation, it is assumed that the measurement of the reference irradiation light amount is performed first. The reference irradiation light amount is measured by setting the inclination angle of the reflecting mirror 32 whose inclination angle is variable so that the reflected light directly hits the inner wall of the integrating sphere 34.
It is carried out in different states using a rotating means (not shown) using an electric motor or the like. By doing this, the light from the reflecting mirror 32 that is directly applied to the inner wall of the integrating sphere 34 is diffusely reflected on the inner wall in multiple directions, and finally reaches the detectors 351, 35b.
reaches and is detected as the reference irradiation light amount. On the other hand, sample 5
Measurement of the amount of reflected light from reflector 5 is performed according to the above-described principle after the inclination angle of reflector 32 is returned to the original position shown in FIG. In addition, each execution from the selection of the first filter after the measurement preparation is completed to the measurement of the reference irradiation light amount and the measurement of the reflected light amount is performed automatically according to the program by storing a procedure program in the ROM in the storage device 4b of the control device 4. Needless to say, there are many things you can do to make it more practical. Furthermore, it is also useful to measure the reference irradiation light amount and reflected light amount for one filter a plurality of times, and to take the average of the measurements as the measured value. Each measurement value based on the reference irradiation light amount and the reflected light amount from the sample 55 detected by the detectors 35I and 35b is communicated to the control device 4 and stored in a writable memory (hereinafter referred to as RAM) in the storage device 4b. Once it is memorized.

After completing the measurement of absorbance at the irradiation wavelength P, nm,
The process moves on to measuring the absorbance at the next irradiation wavelength, ie, P2 nm in this example. Here again, the measurement of the reference irradiation light amount and the measurement of the reflected light amount are performed using the same method and procedure as in the case of P+flffl described above. As in the previous case, each measured value is communicated to the control device 4 as actual measurement data, and is temporarily stored in the RAM in the storage device 4b. Similarly, each remaining absorbance measurement, i.e., wavelength P3 nm, P4 nm
, p, nm, and P6 nm are sequentially measured, and each measured value is communicated to the control device 4 as actual measurement data and stored in the RAM. Note that each of the filters 331 to 33 of the narrow band pass filter 33 occurs when the absorbance measurement at a certain specific wavelength is completed and the transition is made to the absorbance measurement at the next specific wavelength.
Normally, the exchange and selection operation of 331 is performed automatically according to a procedure program written in advance in the ROM in the storage device 4b of the control device 4. However, even in the case of this embodiment, it is not necessary to change the absorbance for all of the above six wavelengths. It is not necessary to perform measurement, and the wavelength to be measured can be arbitrarily selected by considering the accuracy required for the desired quality evaluation value, the time required for measurement, etc. This can be done using the measurement wavelength selection button in the operation push button 7.

Next, the arithmetic unit 4C of the control device 4 executes R of the storage device 4b.
A quadratic curve uniquely obtained from the measured values of any three wavelengths among the large number of actual measured data obtained by absorbance measurements stored in the AM, that is, the measured values of the reference irradiation light amount and reflected light amount at each measurement wavelength. Second derivative value and ROM of storage device 4b
The component or quality evaluation value of the sample is calculated based on the pre-stored fraction or the specific coefficient value for calculating the quality evaluation value. Note that this specific coefficient value, which is written in advance in the ROM of the storage device 4b, is a component or quality evaluation value obtained, for example, in a sensory test for a large number of samples, and a second derivative value of the absorbance measurement value from the detector. This is a constant determined by multiple regression analysis method.

Here, an example will be shown of second-order differential values used as explanatory variables in multiple regression analysis for determining specific coefficients according to the present invention. For example, six filters P + nm, P2nm, P3nm,
FIG. 4 shows absorbance measurements using a detector for samples using P4 nm, P5 nm, and P6 nm.

In Figure 4, the horizontal axis represents near-infrared wavelength and the vertical axis represents absorbance, and the absorbance measured by each filter is plotted.
The absorbance for each wavelength of B was determined.

In this way, even for the same sample, the absorbance required for each wavelength varies depending on the pulverized particle size/particle size, sample temperature, or environmental temperature. Moreover, the amount of change increases as the wavelength becomes longer, and is generally expressed by the following equation.

Letting the amount of change be Δr and the wavelength be λ, then Δr=mλ+n...(1) (m, n are arbitrary constants) Here, as mentioned above, unlike the scanning type near-infrared spectroscopy process, in the present invention is processed as follows.

Select three arbitrary points of the absorbance for each wavelength determined in FIG. 4, and obtain a quadratic curve uniquely determined by the three arbitrary points. The quadratic curve obtained from A in Fig. 4 is shown in Fig. 5. Although this figure shows an example in which the quadratic curve was obtained from three adjacent points, it is not limited to this example, and any three points can be used. , and ensure that each plotted point is included in one of the quadratic curves.

Pay attention to A of the quadratic curves A1 to A4 formed here. Assume that the quadratic equation S (Pi) of the quadratic curve A is as follows.

S +pn=aP1+bp,+c---(2)S
+p2+=aP2+bp2+c−-(3)S tpv,
=a Pi +b P3+ c・------(4) Here, solving (2), (3), and (4) for a gives S
(p21 = 2a. Furthermore, regarding B in FIG. 4, where the sample has a different pulverized particle size, consider the same way as A. Assuming that B is a quadratic curve just like A, equation 81 becomes as follows.

S (Pll = S (p1》10m P + + n
--(e)S (P21 = S (P2+ +m
P 2 10n−”{7)S tp3+ = S (P3
1 +m P 3 + n −”{8}Here, mP+
n is the amount of change influenced by the particle size/grain size determined by equation (1), and the amount of change Δr is the difference between A and B in Figure 4.
Adding to A and assuming the required quadratic curve for B? be.

Here, when (6), (7), and (8) of B are solved for a, only the numerator of equation (5) becomes as follows.

The molecule of formula (5) is (Pi −P+ ) (SLpo−Lp2+) +m(
P+ -P2)-(PI-P2)(Lpi+-
St. ,) ++e(P3-p, ), and =(P3-P+) (Stp+BiScp2l)-(1
-P2 ) (Srps+-3pu)+m(P+
-F'2) (P3-PI) -m(P, -I1
2) (P3-PI) (P3-PI) (Lp
n-Lp2,)-(P+-P2) (Lpi+-
St ■→. This is the same as equation <5>, and since there is no term for the amount of change Δr=mP+1 due to the influence of particle size, S (P)
It can be proven that the value obtained by the second derivative of =2a is not affected by particle size, particle size, temperature, etc. In other words, the quadratic differential value of the quadratic curve, which is uniquely determined from the absorbance at three arbitrary points, is not affected by particle size, particle diameter, temperature, etc., and can be used as an explanatory variable in the multiple regression analysis described in detail later. This is possible and effectively works to improve accuracy when determining the composition or quality of a sample.

Next, multiple regression analysis using the above-mentioned explanatory variable, that is, the second-order differential value a, will be explained.

Let the second-order differential values obtained by A1 to A4 (or B), respectively, be a, to a4. At this time, it is assumed that the following linear relationship holds.

X=FO +F, a 1+ F2 a2+=・Fa
a4+E r '' (9) At this time, X: Known sample component or quality evaluation value F: Specific coefficient value a by multiple regression analysis: Secondary differential value ε obtained from A1 to A4: Error.

In the same way, absorbance analysis was performed on n samples (9)
Substituting into the multiple regression equation, X1=Fot+Flx' aI1+-F2% a2, 10-+F4w" 841 + εH
x+”'+ F 4W ' a 42 ten 5 2X+
t =Fax 10F+% a1n+-F2xs a2n+
-...+F4% a. ．． n+ε. Then, multiple regression analysis is performed using the above n multiple regression equations and F
. 8~F When calculating the specific coefficient value of 4w, X=Fo, +F
+za a, +F2xm a, +-F4t#4...
・・・・・・Old・・・・・・・・・・・・・・・([1)
Therefore, a relational expression between the component or quality evaluation value of the sample and the second-order differential value obtained from the measured absorbance value is established.

By the way, the filters P1 to P6 are shown as examples, and to ensure accuracy, the optimal filters are selected by performing the regression analysis at 110θnm to 250flnm.
The wavelength range is subdivided into wavelength ranges, for example, absorbances obtained at intervals of 2+++n are used to find out by combining all of them in a matrix manner.

In addition, using six filters, the absorbance of n samples is measured with a detector for other components or quality evaluation values, and the following multiple regression equation is established to obtain the same components or quality evaluation values as described above.

Y=Fo, +F,,a a,+F2,・a2+-tenF4
F-a4 As described above, multiple regression analysis is similarly performed for each component to determine the specific coefficient of the sample, and the fraction or quality evaluation value is determined by absorbance measurement with a detector. In other words, when determining the specific coefficient for each component or quality evaluation value of a sample, the second derivative value determined from the absorbance value of any sample to be analyzed can be calculated using the formula a1
~ a4 and calculates the fraction or quality evaluation value by calculating with the specific coefficient value obtained previously.

By the way, it is clear that the moisture contained in a sample affects the pulverized particle size during pulverization or the absorbance during near-infrared spectroscopic analysis, so when calculating the second derivative value and specific coefficient based on the absorbance of the sample, The moisture value is added to the calculation as a correction value.

This moisture value can also be determined by the absorbance measurement described above. In other words, the natural frequency of water is clear from the molecular formula, and the wavelength is 1940 nm, so based on the absorbance measurement value obtained in this wavelength range, the moisture value is calculated by calculation with the moisture conversion coefficient in the storage device. It is something. This moisture conversion coefficient can also be determined more simply than the above specific coefficient from the relational expression between the absorbance measurement value and the moisture measurement value obtained by chemical analysis or the like.

The component or quality evaluation value calculated by the formula of the second derivative value based on the quality evaluation specific coefficient and absorbance obtained above is visually displayed on the display device 6 and automatically The hard copy is delivered from the printer 8 automatically or based on a command to the operation push button 7. Further, the moisture and the like determined during the process of determining the fraction or quality evaluation value may be visually displayed on the display device 6 at the same time based on the components or the quality evaluation value.

A comprehensive evaluation value that comprehensively evaluates the components or quality evaluation values determined above can be determined.

For example, you can evaluate the evaluation method for each quality evaluation value out of 10 points and use the total score of each quality evaluation value as the overall evaluation value, or you can change the degree of involvement of each quality evaluation value in the overall evaluation to more precisely match your preferences. Various forms can be considered, such as those that can be evaluated in a certain field, and ultimately, as a comprehensive evaluation, this sample will be able to calculate an evaluation value such as belonging to this rank in the field, and it will be possible to calculate the evaluation value that belongs to this rank in the field. According to the present invention, the above components or quality evaluation values can be determined fairly in a short time and not influenced by human factors.

The components or quality evaluation values of each sample calculated by this quality evaluation device can be stored as data in an external storage device using a magnetic medium such as a floppy disk. When calculating a ratio, etc., it is also possible to read data from an external storage device into the RAM of the storage device 4b in this device and perform necessary calculations based on this data.

Now, in order to analyze a sample using the above-mentioned quality evaluation device, the sample is pulverized and analyzed, and a so-called centrifugal type sample pulverizer, etc., which consists of pulverizing blades that rotate at high speed within a wire mesh, is used for pulverizing the sample. .

The quality evaluation device of the above embodiment is a reflective near-infrared spectrometer that measures the absorbance when a sample is irradiated with near-infrared light of a specific wavelength by measuring the intensity of reflected light from the sample. However, it is also possible to use a transmission-type near-infrared spectrometer that measures the intensity of transmitted light that has passed through the sample, and it is also possible to measure absorbance based on both reflected light and transmitted light. It is also possible to create a more precise near-infrared spectrometer that performs this.

〔Effect of the invention〕

As described above, according to the present invention, in stationary near-infrared spectroscopy, it is possible to eliminate the influence of the pulverized particle size and particle size of the sample and temperature or environmental temperature, which could only be obtained with near-infrared spectroscopy using continuous scanning spectra. This eliminates the need for conventional peripheral equipment to stabilize the environment and sample temperature, and to control sample crush particle size and particle size, allowing for highly accurate near-infrared spectroscopic analysis that is unaffected by human or external factors. It can also be realized using a formula, and because the structure is simple, it has become possible to obtain the same accuracy as a scanning type even with an inexpensive measuring device.

[Brief explanation of the drawing]

FIG. 1 is a schematic front view of a quality evaluation device according to an embodiment of the present invention, FIG. 2 is a sectional side view of the main part of a near-infrared spectrometer used in the quality evaluation device of FIG. 1, and FIG. 3 is a diagram showing different samples. FIG. 4 is a graph (absorbance curve) showing the relationship between the near-infrared irradiation wavelength and absorbance for a given sample, FIG. 4 is a graph showing the absorbance of this example, and FIG. 5 is a quadratic curve obtained from the absorbance in FIG. 4. In the figure, 1... coffee bean quality evaluation device, 2...
...Cabinet, 3...Near-infrared spectrometer, 4.
...Control device, 4a...Input/output signal processing device, 4b.
...Storage device (ROM, RAM), 4 c...Arithmetic device, 5...Sample container mounting box, 6...Display device, 7
...Push button for operation, 8...Printer, 9
... Input device (keyboard), 31 ... Light source, 32
...Reflector, 33...Narrow band pass filter, 33
a~33f...filter, 34...integrating sphere, 35
a, 35b...detector, 36...lighting window, 37...
・Measurement part, 52... Sample container, 55... Sample (coffee beans), 77... Temperature setting device, 78... Warming device, 79... Temperature sensor

Claims (1)

[Claims] A near-infrared spectroscopic analysis method using a near-infrared spectrometer equipped with at least three or more narrow band filters that transmit only wavelengths at which the difference in quality of the sample is noticeable as a difference in absorbance when a sample is irradiated with near-infrared light. Among the respective absorbances obtained from near-infrared rays of at least three wavelengths that have passed through the plurality of narrowband filters, a quadratic curve is obtained from the absorbance at any three wavelengths, and the quadratic curve is A near-infrared analysis method, characterized in that the quality of the sample is analyzed by performing multiple regression analysis using each value obtained by second-order differentiation as an explanatory variable in a multiple regression equation.