JPH07198601A - Near infrared spectroscopy - Google Patents

Abstract

PURPOSE:To create a high-accuracy analytical curve by using absorbances of at least two wavelengthes in a wavelength area of a neutral zone as a descriptive variable of an analytical curve at the time of measuring the component content of food. CONSTITUTION:At least two wavelengthes are selected from a wavelength area (1640nm-1680nm) of a neutral zone in a near infrared area, and the absorbances of the wavelengthes are used as a descriptive variable of an analytical curve at the time of measuring the component content or the like. Thus, any influence of temperature and moisture content of a sample can be removed, and the accuracy of an analytical curve can be improved so as to measure the component content or the like correctly.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to physical characteristics such as the content of ingredients, stickiness or hardness of food by near-infrared spectroscopy, or comprehensive quality characteristics (tasting, umami, processing characteristics, etc.) obtained from these characteristics. ) Relates to a method for creating a calibration curve when measuring. Foods to be measured here include agricultural and marine products such as grains and fruits, processed foods and beverages, and tablet additives.

[0002]

2. Description of the Related Art Near-infrared spectroscopic analysis is an analytical method in which light in the near-infrared region (wavelength range 0.8 to 2.5 μm) is applied to a sample and the content or concentration of the sample is measured from the reflected or transmitted light. is there.

The spectroscopic analysis is an analysis method utilizing the fact that the light incident on a substance shows a characteristic change in the type of atoms and molecules contained in the substance. In the near-infrared spectroscopic analysis, the light absorption peculiar to each substance. The type and amount of the substance is measured from the wavelength and height of the peak of the spectrum. The basic formula for the quantification is as follows.

[0004] There wavelength of the peak of the specific absorption spectrum in substance lambda, attenuation I _t of energy if the energy of the light to the I ₀ is determined from the following equation.

[0005]

[Equation 1] Equation 1 Here, k: constant C: concentration of substance When Equation 1 is rewritten, it becomes as follows.

[0006]

[Equation 2] Above

[Equation 3] Absorbance is the degree of light absorption of a substance.

As described above, there is a certain relationship between the absorbance of a certain component (substance) in a sample and the concentration of that component,
If a calibration curve is created by previously obtaining this relationship, the component concentration of the unknown substance in the sample can be determined from the absorbance at the measurement wavelength based on the calibration curve.

In practice, a sample contains various components, and the absorption spectrum of several components overlap,
It is not possible to simply obtain a calibration curve from the absorbance of one measurement wavelength. In order to remove the influence of other components, the regression equation showing the calibration curve is created by multiple regression analysis. The multiple regression analysis is a multivariate analysis method that obtains an objective variable from a certain result (objective variable) and some causes (explaining variables) that affect it, and the near-infrared spectroscopy. In the case of analysis, the concentration of a component corresponds to the objective variable and the explanatory variable is the absorbance at several different wavelengths.

In the case of grains, the measurement is performed after crushing, but the reflectance of powder samples differs depending on the particle size, and therefore the influence of particle size must be taken into consideration when preparing a calibration curve. In addition, the temperature and water content of the sample are also known to affect accuracy.

Therefore, when measuring the component content of the sample,
In order to eliminate the influence of accuracy due to some of these factors, the absorbance at a different wavelength is measured in addition to the absorbance at the wavelength derived from the component to be analyzed, and these are used as explanatory variables for correction to create a calibration curve. There is a need.

Conventionally, as the explanatory variable for this correction, the absorbance of wavelengths derived from components other than the component to be analyzed is used, or the absorbances of wavelengths that are likely to have a high correlation are sequentially selected by trial and error and used. Or use the absorbance of the wavelength in the neutral band, or differentiate the original spectrum to create a second derivative spectrum that removes the influence of the background, and confirm the absorption band that is difficult to find in the original spectrum, and check that wavelength. Has been proposed (Journal of Japan Society of Food Industry, Vol. 32, No. 9, September 1985, Mutsue Iwamoto et al., "Nondestructive quality measurement of food by near infrared spectroscopy").

[0012]

The accuracy of the calibration curve can be evaluated by the standard error obtained by using other data not used for preparing the calibration curve, for example, data by chemical analysis.
The standard error of the calibration curve by the conventional method is still not sufficiently small even if various corrections are used, and further improvement in accuracy has been a problem.

The inventors of the present invention have conducted extensive studies to solve the problem, and as a result, there is a specific wavelength range showing a certain correlation between the water content of the sample and the second derivative absorbance with the temperature of the sample as a parameter. I found out that. If a calibration curve is created by using such an absorbance in a specific wavelength range as an explanatory variable of at least two wavelengths, it is possible to obtain a calibration curve with a small standard error that is not affected by changes in temperature and water content. I found it.

An object of the present invention is to create a highly accurate calibration curve based on such new knowledge.

[0015]

The specific wavelength range found by the present inventors is a neutral wavelength range not derived from any component.

That is, the present invention is characterized in that the absorbance of at least two wavelengths in the wavelength range of the neutral zone is used as an explanatory variable of a calibration curve when measuring the content of ingredients in food.

FIGS. 1 and 2 are graphs showing the relationship between the protein content of grains and the second-order differential absorbance with temperature as a parameter. FIG. 1 shows the case where the wavelength is 1668 nm, and FIG. 2 shows the case where the wavelength is 1660 nm. Each case is shown. From these tables, it can be seen that there is a clear correlation between the water content and the second derivative absorbance at a constant temperature.

[0018]

EXAMPLES Next, examples in which the absorbance at such a specific wavelength is used as an explanatory variable will be described.

Here, the protein content of the grain is C, the absorbance at a certain wavelength is represented by Q (numerical value of wavelength in nm unit), and its second derivative is represented by Q ₂ . For example, Q ₂ (21
80) indicates the second derivative of the absorbance at a wavelength of 2180 nm. Example 1

[0020]

[Equation 4] Equation 3 Here, K (n) is a regression coefficient (n = 0, 1, 2, 3),
It is a constant as follows.

K (0) = 82.958 K (1) = − 596.156 K (2) = − 56.411 K (3) = 594.164 K (4) = 139.801 Standard of Example 1 The deviation SEC is SEC = 0.266
Then, the multiple correlation coefficient MR is MR = 0.9342. Example 2

[0022]

[Equation 5] Formula 4 Here, K (n) is as follows.

K (0) = 7.312 K (1) = − 614.174 K (2) = − 2250.330 K (3) = 2998.757 K (4) = 541.652 Multiple correlation coefficient MR with SEC = 0.284
Is MR = 0.9245. Comparative example

[0024]

[Equation 6] Equation 5 Here, K (n) is as follows.

K (0) = 14.386 K (1) = − 1392.998 K (2) = 1347.335 K (3) = 432.078 K (4) = 903.382 This comparative example is SEC. = 0.314, the multiple correlation coefficient M
R has MR = 0.066.

From these, it can be seen that Examples 1 and 2 have a smaller standard error and higher accuracy than the comparative example of the conventional method. This is because the comparative example sets the neutral wavelength to 1640 nm.
While only the wavelength is used, in Examples 1 and 2, it is 1
It can be inferred that the use of two wavelengths of 666 nm and 1658 nm eliminated the influences of temperature and water content, which are the factors of variation in the neutral zone, more appropriately.

In Examples 1 and 2, the standard error of Example 1 in which the ratio of absorbance at two wavelengths is taken is smaller than that of Example 2 in which the difference between the two is taken, but this is generally due to some cause. When explaining an outcome, when there are several causes that affect the outcome, taking a ratio rather than taking the difference between those two causes eliminates the background effects common to those causes. It is due to being effective in.

Therefore, in order to remove the influence of the background even more effectively, the second derivative spectrum of the neutral zone of FIGS. Equation 6 is obtained. Example 3

[0029]

[Equation 7] ... Equation 6

[0030]

As described above, according to the present invention, in the near-infrared analysis, the absorbance of at least two wavelengths in the neutral wavelength band that does not originate from any component is used to measure the component content of food. By using it as an explanatory variable of the calibration curve at that time, the effect of temperature and water content can be removed, and the accuracy of the calibration curve is improved.

[Brief description of drawings]

FIG. 1 is a table showing the relationship between the protein content of grains and the second-order differential absorbance with temperature as a parameter, showing the case where the wavelength is 1668 nm.

FIG. 2 is a chart showing the relationship between the protein content of grains and the second-order differential absorbance with temperature as a parameter, showing the case where the wavelength is 1660 nm.

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] April 4, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0017

[Correction method] Change

[Correction content]

FIGS. 1 and 2 are graphs showing the relationship between the water content of grains and the secondary differential absorbance with temperature as a parameter. FIG. 1 shows the case where the wavelength is 1668 nm, and FIG. 2 shows the case where the wavelength is 1660 nm. Each case is shown. From these tables, it can be seen that there is a clear correlation between the water content and the second derivative absorbance at a constant temperature.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Name of item to be corrected] Brief description of the drawing

[Correction method] Change

[Correction content]

[Brief description of drawings]

FIG. 1 is a table showing the relationship between the water content of grains and the second-order differential absorbance with temperature as a parameter, showing the case where the wavelength is 1668 nm.

FIG. 2 is a chart showing the relationship between the water content of grains and the second-order differential absorbance with temperature as a parameter, showing the case where the wavelength is 1660 nm.

Claims (1)

[Claims] 1. Near red, characterized in that at least two wavelengths are selected from the neutral bands in the near-infrared region, and the absorbances at those wavelengths are used as explanatory variables of a calibration curve when measuring component contents and the like. Spectroscopic analysis.