JP6911543B2 - Baseline determination method - Google Patents

Description

The present invention relates to a method for determining a baseline noise level in a chromatogram and for data processing.

Chromatography is a method of separating and quantifying a sample containing a plurality of components with a column. In chromatography, periodic or irregular noise due to various factors is superimposed on the baseline, so a process for eliminating peaks having a certain height and an area value or less is also used. In order to reliably detect the peak that appears, it is preferable to determine the exclusion value according to the degree of baseline noise, and it is necessary to always accurately grasp the degree of baseline noise. As a general noise calculation method, it is performed in accordance with ASTM (ASTM American Society for Testing and Materials E 685-93). This is a method in which the baseline is divided at intervals of 30 seconds, the noise width of each section is measured, and the average of 10 sections is expressed as noise (see FIG. 1).

Since the above method is standardized, it is very effective for checking the difference between devices, checking the performance of devices, and validating, but the baseline is fairly stable in order to calculate the noise value accurately. It is difficult to obtain an accurate noise value unless it is performed in (a state where there is little swell or drift). Further, since the baseline noise is greatly affected by the eluent at the time of use, the detector, the liquid feed pump, the state of the use environment, and the like, there is also a problem that the degree of noise changes depending on the measurement date or measurement.

The present invention provides a method of calculating a noise value from a chromatogram. Further, it provides a method of processing chromatogram data from the noise value under appropriate conditions.

As a result of diligent studies, the present inventor has focused on the difference between the maximum value and the minimum value of the output value in a plot group consisting of a specific plot in the chromatogram and plots of several points before and after the plot. It came to be completed. Hereinafter, the present invention will be described in detail.

The present invention relates to a part or all of the time from the start to the end of chromatogram measurement.
Select the maximum and minimum output values for each plot in a plot group consisting of a specific plot and several plots before and after the plot.
The difference between the maximum value and the minimum value is calculated for each plot.
Create a frequency distribution using the difference as a class,
This is a method of determining the average value of the class indicating the maximum frequency from the lower class of the frequency distribution as the noise value of the chromatogram.

The present invention first plots the maximum and minimum output values in a plot group consisting of a specific plot and plots of several points before and after the plot for a part or all of the time from the start to the end of chromatogram measurement. Select for each. Specifically, it will be described with reference to the chromatograms shown in FIGS. 4a and 2a. FIG. 2a is an enlarged view of FIG. 4a from 0 minutes to 0.15 minutes. Table 1 summarizes the plotted points of the chromatogram shown in FIG. 2a in a table.

When a specific plot is included in the plot group with # 4 (time 0.025) and three points before and after are included in the plot group, # 1 (time 0.000) to # 7 (time 0.050) are the plot groups for # 4. In this plot group, the maximum value is 1.817 (# 1) and the minimum value is 0.300 (# 4). This is repeated after # 5. When three points before and after are included in the plot group, the specific plots are # 4 to # 16. That is, even if the entire time from the start to the end of the chromatogram measurement is the target, this step is not performed for several points immediately after the start of the measurement and several points immediately before the end of the measurement. In the above description, the mode in which the three points before and after are included in the plot group is used, but there is no problem in arbitrarily deciding how many points before and after are included in the plot group.

For the chromatogram shown in FIG. 2a, the maximum and minimum output values in a plot group consisting of a specific plot and plots of three points before and after the plot are plotted to be FIGS. 2b and 2c.

Next, the difference between the maximum value and the minimum value is calculated for each plot. Subsequently, the explanation using the chromatogram shown in FIG. 2a will be as shown in FIG. Table 2 shows the difference between the maximum value and the minimum value of each plot.

Next, a method of creating a frequency distribution using the difference as a class will be described. In the above procedure, the difference is calculated for all the area floors from 0 minutes to 90 minutes, the difference is set as the class, and the frequency is used as the frequency to create the frequency distribution (see FIGS. 4 and 5). It can be seen that the class showing the maximum frequency from the lower class of the created frequency distribution is 1.6 to 1.7, and the average value of 1.65 is the noise of the chromatogram shown in FIG. 4a. It can be determined as a value (Nz).

The noise value determined by the method of the present invention described above is useful for calculating a particularly stable point (base point) in the target chromatogram. Specifically, a plot in which the difference is smaller than the value obtained by multiplying the noise value by a coefficient of 1.0 or less is extracted from the chromatogram (base point extraction), and the extracted plot group (base point group) is extracted. By subtracting the approximate calculation by the mathematical method from the chromatogram, the chromatogram excluding the fluctuation of the baseline can be obtained.

The coefficient for multiplying the noise value may be a value of 1.0 or less, but a value of 0.1 to 0.8 is preferable. As the base point to be used, all the data points that meet the above conditions may be used, but only a part thereof may be used. In particular, it is preferable to select the base points so that the time intervals are as even as possible and perform the calculation, because more accurate approximation is possible. In addition, the mathematical method referred to here can be a simple folding line approximation, a linear approximation by the least squares method, a polynomial approximation by the least squares method, a nonlinear function approximation by the least squares method, an approximation by spline processing, etc. In a general chromatogram, the linear approximation by the least squares method or the quadratic approximation by the least squares method is suitable because the correlation is good and the calculation load is small.

Further, the noise value determined by the method of the present invention described above is also useful in determining the peak height to be excluded during the peak detection process. Specifically, by setting a value obtained by multiplying the noise value by a coefficient of 1.0 or more as the exclusion peak height, unnecessary peaks can be efficiently eliminated. The coefficient for multiplying the noise value may be a value of 1.0 or more, but 3 to 10 is preferable.

In an embodiment of the present invention, there is no problem in determining the noise value for each measurement, and the noise value is used to eliminate the baseline fluctuation and / or determine the peak height to be excluded by the method described above. By doing so, chromatogram data processing may be performed.

According to the present invention, the noise value of the chromatogram can be easily determined, and by using the noise value, data processing for efficiently eliminating baseline fluctuations and unnecessary peaks can be performed.

It is a figure which showed the method of the noise calculation by the general method (ASTM method). FIG. 2a is a chromatogram data point (plot), FIG. 2b is a diagram in which the minimum value for each plot is superimposed, and FIG. 2c is a diagram in which the maximum value for each plot is superimposed. It is a figure which calculated the difference about the chromatogram of FIG. 2a. FIG. 4a is a chromatogram data point (plot), FIG. 4b is a diagram in which the minimum value and the maximum value are superimposed for each plot, and FIG. 4c is a diagram in which the difference is calculated. It is a frequency distribution created based on the difference of FIG. It is a figure which showed the system configuration used in Example 1. FIG. FIG. 7a is a data point (raw data), and FIG. 7b is a diagram in which the data point and the maximum value and the minimum value are superimposed. It is a figure which showed the difference (noise width) which subtracted the minimum value from the maximum value in Example 1. FIG. FIG. 9a is a diagram in which the noise in the first embodiment is frequency-distributed, and the horizontal axis shows the noise and the vertical axis shows the frequency (the number of data points). FIG. 9b is a diagram showing noise calculated by a general method (ASTM method). It is a figure which showed the result of having applied the noise value obtained in Example 1 to an automatic data processor, and performed the peak detection. It is a figure which showed the basic point of the baseline based on the noise value obtained in Example 1, and the result of suppressing the obtained baseline. FIG. 11a is a diagram in which the raw chromatogram and the baseline (reference numeral ○) are superimposed, FIG. 11b is a diagram in which the baseline (linear equation) is subtracted from the raw chromatogram, and FIG. 12c is a diagram in which the baseline (secondary equation) is subtracted from the raw chromatogram. It is the figure which subtracted the formula). It is a figure which showed the system configuration used in Example 2. FIG. 13a is a data point (raw data), and FIG. 13b is a diagram in which the data point and the maximum value and the minimum value are superimposed. FIG. 14a is a diagram in which the noise in the second embodiment is frequency-distributed, and the horizontal axis shows the noise and the vertical axis shows the frequency (the number of data points). FIG. 14b is a diagram showing noise calculated by a general method (ASTM method). It is a figure which showed the result of having applied the noise value obtained in Example 2 to an automatic data processor, and performed the peak detection. It is a figure which showed the basic point of the baseline based on the noise value obtained in Example 2, and the result of suppressing the obtained baseline. FIG. 16a is a diagram in which the raw chromatogram and the base point (reference numeral ◯) are superimposed, and FIG. 16b is a diagram in which the baseline (secondary equation) is subtracted from the raw chromatogram.

The effect of the present invention was verified using an actual chromatogram. However, the present invention is not limited to the description of Examples.

(Example 1)
Verification was carried out in an ion chromatography system (non-suppressed anion). The actual measurement was performed using the liquid chromatogram system shown in FIG. The system consists of a solvent degassing device (SD-8020) 2, a liquid feed pump (DP-8020) 3, a sample injection device (AS-8020) 4, a column oven (CO-8020) 6, and an electrical conductivity meter (CM-). It was composed of 8020) 7 and a data processing device (LC-8020II) 9 (both manufactured by Toso Co., Ltd.).
As the analysis column 5, TSKgel IC-Anion-PWXL (4.6 mm ID × 7.5 cm) manufactured by Tosoh Corporation was used, and Br ⁻ , NO ₃ ⁻ , PO ₄ ^2- , and SO ₄ ^2- ion were used. Was separated.

Other conditions are as follows.
Data collection sampling: 200 msec
Injection amount: 30uL
Column temperature: 40 ° C
Flow velocity: 1.200 mL / min,
Eluent: Mix the following and use pure water to make up to 1 L. Boric acid: 360 mg, sodium tetraborate tetrahydrate: 575 mg, glycerin: 5.0 g, potassium gluconate: 350 mg, acetonitrile: 40 mL, n-butanol: 30 mL

FIG. 7 is a chromatogram obtained under the above conditions. FIG. A shows a raw chromatogram, and FIG. B shows the maximum and minimum values calculated at three points before and after.

FIG. 8 is a diagram showing (maximum value-minimum value), that is, the noise width. As can be seen from FIG. 10, the noise value thus obtained ^{is slightly higher near the peaks of the target Br −} , NO ₃ ⁻ , PO ₄ ^2- , and SO ₄ ^2- ions, but on the chromatogram. It can be seen that the values are almost constant even in the part where the swell is generated and the part where the baseline is drifting.

Next, the frequency distribution of the noise value obtained above is created.
FIG. 9a is a diagram showing the frequency by setting the noise value from 0 mV to 0.1 mV in increments of 0.005 mV. In this case, since the frequency of the section of 0.015 mV to 0.020 mV is the highest at 894, the average value of 0.0175 mV is set as the noise value (Nz).

For comparison, when the noise is calculated by the ASTM method, which is often used in the field of liquid chromatography, the result is as shown in FIG. 9b. The noise width was calculated by dividing the interval by 30 seconds by the method shown in FIG. 1, and the average of 10 sections was taken. Calculated by this method, it is 6.404 mV in the first section, 1.214 mV in the second section, and 1.855 mV in the third section, which are compared with the noise value (0.0175 mV) obtained by this method. It is calculated as a value 10 times or more larger.

FIG. 10 is a diagram showing the results of actual data processing under the following peak detection conditions. The basic peak detection conditions are as follows. The height of the exclusion peak was set to 0.088 mV as 5 times (S / N = 5) of the noise value (0.0175 mV).
Peak start detection sensitivity: 0.500 [mV / min]
Peak-end detection sensitivity: 0.500 [mV / min]
Exclusion peak area: 0.000 [mV x seconds]
Exclusion peak half width: 0.000 [seconds]
Exclusion peak height: 0.088 (0.0175 × 5) [mV]
In this case, it is determined that the noise on the baseline is not a peak, and the total number of peaks is 7, which is almost only the peak of the target ionic component. As described above, by using the method of the present invention, it is possible to automatically calculate the excluded peak height at the time of peak detection, and it can be seen that an appropriate peak detection result can always be obtained.

The effect of baseline suppression using the noise value obtained by the above method was also confirmed. First, for the target chromatogram, data having a noise value (maximum value-minimum value) obtained by multiplying the noise value (Nz) by a coefficient (Nf) of 0.80 was extracted.
As shown in FIG. 11a, it can be seen that the region without peaks and waviness is extracted as the base point (reference numeral ◯). The base point group was then approximated mathematically and subtracted from the original chromatogram. FIG. 11b is obtained by subtracting the baseline obtained by approximating with the linear equation, and FIG. 11c is obtained by subtracting the baseline obtained by approximating with the quadratic equation. In either case, it can be seen that the chromatogram drift can be canceled.

(Example 2)
Verification was performed on a molecular exclusion chromatography system (GPC). The actual measurement was performed using the liquid chromatogram system shown in FIG. The system consisted of a high-speed GPC device HLC-8320 GPC, an ultraviolet-visible detector (254 nm) (UV-8020) 12, and a data processing device (GPC-8020II) 9, all of which were manufactured by Tosoh Corporation.
As the analysis column 5, four TSKgel Mutipore HXL-M manufactured by Tosoh Corporation were used to separate the epoxy resin.

Other conditions are as follows.
Data collection sampling: 500 msec
Injection amount: 100uL
Column temperature: 40 ° C
Flow velocity: 1.000 mL / min
Eluent: tetrahydrofuran (THF).
In this example, verification was performed using the chromatogram obtained by the ultraviolet-visible detector 12.

FIG. 13 is a chromatogram obtained under the above conditions. FIG. 13a shows a raw chromatogram, and FIG. 13b shows the maximum and minimum values calculated at five points before and after.

FIG. 14 is a diagram showing the difference between the maximum value and the minimum value obtained above, that is, the frequency distribution of the noise value, in which the noise value is set in intervals of 0.002 mV from 0 mV to 0.03 mV and the frequency is shown. be. In this case, since the frequency in the section of 0.006 mV to 0.008 mV is the highest at 1706, the average value of 0.007 mV is set as the noise value (Nz).

When the noise is calculated by the ASTM method for comparison, the result is as shown in FIG. 14b. The noise width was calculated by dividing the interval by 30 seconds by the method shown in FIG. 1, and the average of 10 sections was taken. When calculated by this method, it can be seen that there are considerable variations of 0.0127 mV in the first section, 115.3458 mV in the fourth section, and 3.3262 mV in the sixth section.

FIG. 15 is a diagram showing the results of actual data processing under the following peak detection conditions. The basic peak detection conditions are as follows. The height of the exclusion peak was set to 0.035 mV as 5 times (S / N = 5) of the noise value (0.007 mV).
Peak start detection sensitivity: 0.050 [mV / min]
Peak-end detection sensitivity: 0.050 [mV / min]
Exclusion peak area: 0.000 [mV x seconds]
Exclusion peak half width: 0.000 [seconds]
Exclusion peak height: 0.035 (0.0175 × 5) [mV]
In this case, it is determined that the noise on the baseline is not a peak, and the total number of peaks is 11, which is almost only the target oligomer peak.

The effect of baseline suppression using the noise value obtained by the above method was also confirmed. First, for the target chromatogram, data having a noise value (maximum value-minimum value) obtained by multiplying the noise value (Nz) by a coefficient (Nf) of 0.80 was extracted.
As shown in FIG. 16a, it can be seen that the region without peaks and waviness is extracted as the base point (reference numeral ◯). Next, the base point group was approximated by a quadratic equation and subtracted from the original chromatogram, and it can be seen that the drift of the chromatogram could be canceled as shown in FIG. 16b.

1. 1. Eluent 2. Degassing device 3. Liquid feed pump (sample side)
4. Sample injection valve 5. Analytical column 6. Column constant temperature bath 7. Electrical conductivity meter 8. Waste liquid 9. System control and data processing equipment 10. Differential refractometer 11. Resistance tube 12 / UV-visible detector 13. Liquid feed pump (reference side)
14. Pump oven 15. GPC system

Claims (3)

For some or all of the time from the start to the end of chromatogram measurement
Select the maximum and minimum output values for each plot in a plot group consisting of a specific plot and several plots before and after the plot.
The difference between the maximum value and the minimum value is calculated for each plot.
Create a frequency distribution using the difference as a class,
A method for determining the average value of the classes indicating the maximum frequency from the lower class of the frequency distribution as the noise value of the chromatogram. A plot in which the difference is smaller than the value obtained by multiplying the noise value determined in claim 1 by a coefficient of 1.0 or less is extracted from the chromatogram.
A method for processing chromatogram data, which comprises subtracting an approximate calculation of an extracted plot group by a mathematical method from the chromatogram. A chromatogram data processing method, characterized in that a value obtained by multiplying the noise value determined in claim 1 by a coefficient of 1.0 or more is set as the exclusion peak height.

Images