JPS59171837A - Data correction for transmissivity measurement with infrared spectroscope - Google Patents

Abstract

PURPOSE:To measure the transmissivity spectrum of a sample accurately by dividing data of a transmissivity curve by data of a background curve to correct the transmissivity curve. CONSTITUTION:A transmissivity curve is memorized with respect to the wavelength or the number of waves of a sample measured with an infrared spectroscope. Then, after the differentiation of the transmissivity curve, the differentiation value is squared and the results are divided in terms of a fixed wavelength or a fixed number of waves and the sum of the squared differentiation values in each divided section is calculated. Background curves are prepared respectively by using data corresponding to the sections involved of the transmissivity curve when the value of the sum is below the predetermined threshold and data of the transmissivity curve corresponding to the sections on both sides of the sections involved when the value of the sum exceeds the threshold. The data of the transmissivity curve is divided by the data of the background curves to correct the transmissivity curve. Thus, a correct transmissivity spectrum of the sample can be measured.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for correcting transmittance measurement data in a spectroscopic device that measures the transmittance of a sample in the infrared region, such as a Fourier interference infrared spectrometer.

As is well known, Fourier interference infrared spectrometers are widely used, which use an interferometer such as a Michelson interferometer to obtain a spectrum of a sample, that is, a change curve of transmittance with respect to wavelength, by Fourier transformation.

Incidentally, in this type of apparatus, KBr (potassium bromide) is often used as an optical element such as a sample window or a prism. KBr can be used with other optical crystal materials such as Na
This is because it has higher transparency to infrared rays than C1 and the like. However, even when KBr is used as a sample window, the baseline, that is, the transmittance curve corresponding to 100% transmittance (corresponding to an empty sample), may be tilted or curved due to absorption and scattering effects due to its wavelength characteristics. Sometimes I sip.

If the baseline is tilted or curved in this way, the transmittance of the sample cannot be accurately measured at the tilted portion or at the curved wavelength. For example, as shown in Figure 1 (A), with respect to the true transmittance line A of 100%, the transmittance that actually appears due to absorption and scattering due to the wavelength characteristics of optical elements such as KBr in the optical path of the spectrometer is minus 100% base. If line B was tilted, the true spectrum of the sample would be the first
In the case shown by the broken line C in figure (B), that is, the wavelength λ1
In the case of a sample where absorption occurs only around 1λ2, the actual data output after Fourier transformation is shown in Figure 1 (B).
As shown by the solid line, it may become impossible to accurately identify the sample.

However, in the conventional Fourier interference infrared spectrometer,
The reality is that no particular consideration is given to the slope or curvature of the baseline as described above.

This invention has been made in view of the above circumstances, and is a correction method that corrects the slope and curvature of the baseline corresponding to 100% transmittance, thereby making it possible to accurately measure the transmittance spectrum of a sample. The purpose is to provide the following.

That is, the correction method of the present invention stores a transmittance curve for the wavelength or wavenumber of a sample measured by an infrared spectrometer, differentiates the transmittance curve, squares the differential value, and calculates the differential value by 2. Divide the multiplication value by a certain wavelength or a certain wave number and calculate the differential 2 within each divided section.
The sum of the multiplication values is calculated, and in the section where the value of the sum is less than a predetermined threshold value, the data corresponding to that section of the transmittance curve is calculated, and in the section where the value exceeds the same value, the data corresponding to the section on both sides thereof is calculated. A background curve is created using transmittance curve data, and the transmittance curve is corrected by dividing the transmittance curve data by the background curve data.

The method of the present invention will be explained in detail below with reference to FIGS. 2 and 3.

FIG. 2 shows a flowchart for implementing the correction method of the present invention, and FIG. 3 is a diagram showing the contents of spectral data at each stage of the correction method of the invention.

Note that in actual spectrum data, the horizontal axis is often set to the wave number ν, but in FIG. 3, the horizontal axis is set to the wavelength λ for convenience, and the explanation will be based on this. Of course, the case where the wave number ν is plotted on the horizontal axis is substantially the same as the case where the wavelength λ is plotted on the horizontal axis, and it is only necessary to replace the wavelength in the following explanation with the wave number.

Prior to implementing the correction method of the present invention, the transmittance, that is, the spectrum, corresponding to each wavelength of the sample is measured using a Fourier interference infrared spectrometer, and this is stored in a storage device.

To carry out the correction method of the present invention, first, the above-mentioned measurement data is read out and squared.

That is, for example, raw spectrum data as shown in FIG. 3 (A>, that is, raw data B (λ
) is read out, and this data is treated as a curve and differentiated to obtain data B'(λ) as shown in FIG. 3(B). Here, the absorption by the sample is usually at a specific wavelength, for example λl in Figure 3.
, λ2, and hardly occurs at wavelengths other than λ1 and λ2, so the steep dip parts Dl and D2 near the wavelengths λl and λ2 due to absorption of the sample are removed from the curve in Figure 3 (A). The overall slope is due to the baseline slope. Since the change in the slope of this baseline is small compared to the rise and fall of the dip portions Dl and D2, the differential value B'
As shown in FIG. 3(B), (λ) is almost constant except around the wavelengths λ1 and λ2.

Next, the differential value data B'(λ) as described above is squared,
Data (B'(λ))2 as shown in FIG. 3(C) is obtained. This data is the dip part D+ of raw data B(λ)
, has a peak at the falling and rising edges of D2.

On the other hand, a predetermined wavelength interval (
(or at predetermined wave number intervals) and store them. Then, the differential value data squared as described above, that is, the differential square value data (B'(λ))2
is divided by the dividing points, and the sum of differential square values at an arbitrary number of sampling wavelengths, such as three or five points, within each divided section is calculated. Differential 2 of each interval
The sum S of the multiplication values is shown in FIG. 3(D). Next, the sum S of the differential square values of each section is compared with a predetermined threshold value, and a background curve BS(λ) is created based on the comparison result. That is,
In the interval where the sum S is @([1- or less (all of these intervals are represented by P), the raw data B(λ) (see Fig. 4 (A>)) in that interval before differentiation is used as the interval background. The curves Or, Ca, and Cs are assumed, and on the other hand, in the sections where the sum S exceeds the threshold (these sections are represented by Q), the raw data B (λ) of that section is discarded, and the sections on both sides (i.e., the raw data B ( A linear approximation is made from the data of the section) in which λ) is adopted, and these straight lines are used as the section background curves C2 and C4.In other words, in the section Q, the raw data B(
Connect the values of λ) with a straight line. The background curve BS(λ) created in this way is shown in FIG. 3(E). However, in practice, in order to make the background curve BS(λ) smooth, it is preferable to perform smoothing by sequentially taking three-point average data after linear approximation as described above.

As mentioned above, the sections where the sum S of the differential square values is less than or equal to the threshold value correspond to the portions of the transmittance B mutualistic data B(λ) where the sample does not absorb, and in those sections, the raw data B(λ) is It is used as the background curve.

On the other hand, the sum S of the differential square values is the threshold it! ! The section higher than L corresponds to the dip part Di, D2 of the transmittance curve raw data B(λ) near the wavelength λ11λ2 where the sample absorbs, and in that part the raw data B(λ) is discarded and its A background curve B S,(λ) is created by linear approximation using the raw data B(λ) of the non-absorption sections on both sides. Therefore, the background curve is approximated as much as possible to the baseline of the raw data B(λ) excluding the influence of absorption by the sample. Background curve 88 (λ
) is divided by the transmittance curve raw data B(λ). That is, B(λ)/BS(λ) is determined for each wavelength.

In this way, as shown in FIG. 3(F), the transmittance curve BC( λ) is obtained.

That is, the bending and inclination of the baseline due to absorption and scattering of optical elements such as KBr are corrected, and data of the transmittance curve BC(λ) that more accurately represents the transmittance due to absorption of the sample is obtained.

Note that the dividing point roughness for the differential square value (B'(λ)) is
Usually, the wave number is about 4 force iser, but depending on the sample, it may be set within the range of about 2 to 16 force iser.

As described above, according to the method of the present invention, the transmittance is 1 due to absorption and scattering by optical elements such as windows in the optical path of an infrared spectrometer.
A remarkable effect is obtained in that the curve and slope of the baseline corresponding to 00% can be easily and easily measured, and the accurate transmittance spectrum of the sample can be measured.

[Brief explanation of drawings]

Figures 1 (A) and (B) are diagrams showing an example of the baseline of the transmittance curve and an example of the transmittance curve, respectively, in a Fourier infrared interference spectrometer, and Figure 2 is a diagram showing the correction method of the present invention. Figures 3(A) to 3(F) are diagrams showing data at each stage of the method of the present invention. Applicant: Japan Bunko Industries Co., Ltd. Representative Patent Attorney Yutaka 1) Hisashi Take (1 idiot) Foreign book in Figure 11, 1 (no change in content) Figure 1 Wavelength λ1 λλ Re-class leader Figure 2 Figure 3 λl 82 Loss (B) CD) Include 1 ;Aλ Wavelength F ) λI Mu te Lawful Procedural Amendment (Method) July 15, 1980 Director-General of the Patent Office Kazuo Wakasugi 1, Indication of Case 1982 Patent Application No. 46168 2, Title of Invention Infrared Spectroscopy Data correction method for transmittance measurement using a device 3, relationship with the case of the person making the correction Patent applicant address: 5th name, 2967 Ishikawa-cho, Hachioji-shi, Tokyo, Japan Bunkko Kogyo Co., Ltd. 4, agent address: Port of Tokyo 3-4-18-5, Mita-ku, Date of amendment order: June 28, 1982 (shipment date) 6. Subject of amendment

Claims (1)

[Claims] The transmittance curve for the wavelength or wave number of the sample measured by an infrared spectrometer is stored, and after differentiating the transmittance curve, the differential value is squared, and the differential square value is set at a fixed wavelength interval or at a fixed value. Divide into each wave number interval and calculate the sum of the differential square values in each divided section, and in the section where the sum value is less than a predetermined threshold, data corresponding to that section of the transmittance curve is calculated. In the section where the sum value is equal to or greater than the threshold value, a background curve is created by linear approximation from the transmittance curve data of the sections on both sides, and the transmittance curve data is divided by the background curve data. 1. A data correction method in transmittance measurement using an infrared spectrometer, the data being corrected by correcting transmittance curve data.