Fractional derivative spectral quantitative analysis method for multi-component system

Abstract

A direct quantitative analysis method for fractional derivative spectra of multi-component systems; following the standard addition method, adding [a specific ingredient] to the multi-component system h to be analyzed. n The analyte f at several gradient concentrations was mixed thoroughly, and the spectra of each sample were acquired using a spectrometer to obtain... F 1- F n; Call the SH algorithm: function S=SHfd(Mraw, v , b , p ),set up b , p , v =v0+kΔ, find the value of v0+kΔ. F 1- F continuous derivatives of n DF 1- DF n; Determine the calibration peak position λ and the derivative order vJ: In DF 1- DF In n, the order vJ is found when the correlation between the derivative value at a certain peak position and the amount added is greater than 0.99. The data point corresponding to this peak position is the calibration peak position λ. Under the determined vJ and calibration peak position λ, the content of the analyte f in the multi-component system H can be calculated by plotting the relationship between the amount added C and the derivative value DF = aC + b.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fractional derivative spectrum quantitative analysis, in particular to a fractional derivative spectrum quantitative analysis method for multi-component system. BACKGROUND

[0002] Derivative spectrum method is a multi-component quantitative method other than multivariate statistics, which can eliminate the interference of low-frequency baseline, narrow the peak width and improve the selectivity of the measured component. It does not need secondary modeling and complicated statistical model, and is simple and clear. Existing solvers include wavelet, Fourier transform and gradient noise reduction finite difference, etc., among which the SG method has the highest acceptance. However, the accuracy of these algorithms is not high in the presence of noise, and there is a lack of a tool suitable for actual measurement data and easy to use. SUMMARY

[0003] In order to solve the problems existing in the prior art, the purpose of the present application is to provide a fractional derivative spectrum direct quantitative analysis method for multi-component system.

[0004] In order to achieve the above purpose, the technical scheme of the present application is as follows:

[0005] A fractional derivative spectrum direct quantitative analysis method for multi-component system.

[0006] Composed of the following steps:

[0007] 1) Using a spectrometer to collect the sample spectrum of the multi-component system to be measured h H ;

[0008] 2) According to the standard addition method, adding n a gradient concentration of the component to be analyzed f in the multi-component system to be measured h, mixing uniformly and then using a spectrometer to collect the spectrum of each sample to obtain F 1- F n;

[0009] 3) Calling the SH algorithm: function S = SHfd (Mraw, v , b , p ), wherein Mraw is F 1- F n, setting b as a positive real number to adjust according to the signal noise intensity; p as a positive real number, v = v0+kΔ, wherein k = 0, 2…m; 0.01 < Δ < 0.1, 0 < v < 4; the continuous order derivative F 1- F n of DF 1- DF n is calculated.

[0010] 4) Determine the calibration peak position lambda and the derivative order vJ: In DF 1- DF n, search for the derivative value at a certain peak position and the correlation between the added amount is greater than 0.99 when the order vJ, the data point corresponding to the peak position is the calibration peak position lambda;

[0011] 5) Under the determined vJ and calibration peak position lambda, the relationship between the added amount C and the derivative value DF is plotted DF=aC+b, and the content of the component f to be measured in the multi-component system H can be calculated.

[0012] Further, in the step 2) n >5.

[0013] Compared with the prior art, the beneficial effects of the present application are:

[0014] The present application proposes that the noise-containing frequency spectrum measured can be enveloped by a function of the sum of an exponential function and a constant, and a more accurate noise separation response function, the hyperbolic secant (Sech) function, is constructed on the envelope. The simulated hyperbolic function (SH) after Fourier inverse transform is a usable convolution kernel in the time domain. Around this kernel, through numerical calculation (Grünwald-Letnikov) and corresponding adjustment, any order calculation on the original frequency spectrum can be realized. On this basis, an executable MATLAB script is proposed, and interactive mathematical analysis verification is carried out in the presence and absence of noise. BRIEF DESCRIPTION OF DRAWINGS

[0015] Fig. 1 is the relationship diagram of different added amounts and 1.5 order derivative values in H1 sample;

[0016] Fig. 2 is the sample derivative spectrum diagram in the range of 3000-3120 data points under 1.5 order;

[0017] Fig. 3 is the relationship diagram of different added amounts and 1.5 order derivative values in H2 sample;

[0018] Fig. 4 Linear relationship between derivative value of cane juice spectrum and cane sugar added amount. DETAILED DESCRIPTION

[0019] The technical solutions of the present application will be further described in detail below in combination with the drawings and specific embodiments:

[0020] As Figs. 1-4 shown, a fractional order derivative spectrum quantitative analysis method of a multi-component system;

[0021] The present invention uses the standard addition method to collect the spectra of the sample to be measured and the component to be measured, and uses the SH algorithm to establish a direct quantitative analysis method for the fractional derivative spectra of a multi-component system.

[0022] Steps:

[0023] 1) Use a spectrometer to collect the sample spectra of the multi-component system h to be measured H ;

[0024] 2) According to the standard addition method, in the multi-component system h to be measured, add n ( n >5) gradient concentrations of the component f to be analyzed, mix evenly, and then use a spectrometer to collect the spectra of each sample, obtaining F 1- F n;

[0025] 3) Call the SH algorithm: function S = SHfd (Mraw, v , b , p ), where Mraw is F 1- F n, set b (a positive real number, adjusted according to the signal noise intensity), p (a positive real number), v = v0 + kΔ (k = 1, 2…m; 0.01 < Δ < 0.1, 0 < v < 4), and obtain F 1- F n consecutive order derivatives DF 1- DF n;

[0026] 4) Determine the calibration peak position λ and the order of differentiation vJ: In DF 1- DF n, search for the order vJ when the correlation between the derivative value at a certain peak position and the added amount is greater than 0.99, and the data point corresponding to this peak position is the calibration peak position λ;

[0027] 5) Under the determined vJ and calibration peak position λ, plot the relationship between the added amount C and the derivative value DF as DF = aC + b, and the content of the component f to be measured in the multi-component system H can be obtained.

[0028] Example 1

[0029] Analysis of DL-aspartic acid

[0030] 1. Take analytical pure DL-aspartic acid, D-aspartic acid, and L-aspartic acid, and prepare two mixed samples containing 45.2% and 68.5% of DL-aspartic acid respectively;

[0031] 2. Raman spectra (DL, D, L, H1, and H2) of DL-aspartic acid, D-aspartic acid, L-aspartic acid, and two mixed samples were acquired using a wide-field transmission Raman spectrometer (PERT1000, CEPT (Beijing) Co., Ltd.). The acquisition parameters were: integration time 500 ms, smoothing point count 0, center wavelength 532 nm, power level 5, and 200 cm⁻¹. -1 ~3442cm -1 );

[0032] 3. Analysis of DL-aspartic acid in H1:

[0033] 1) Weigh 1.0g into the multi-component system H1 to be tested, and add 0.1g, 0.2g, 0.3g, 0.4g, 0.5g, and 0.6g of the analyte DL asparagic acid according to the standard addition method, and grind and mix to obtain B1-B6;

[0034] 2) On the Matlab computing platform, call the function S = SHfd(Mraw, v, b, p), where Mraw is the spectrum to be processed; set b=100, p=5, and obtain the continuous derivative spectra of DL, D, and L aspartic acid; where v = 0.5 + kΔ (k=0, 1,2, 3, 4; Δ=0.5, 0.5) <v<2);

[0035] 3) Searching for a derivative order vd of 1.5 at the peak data point position 3080, the linear relationship between the DL derivative value and the amount added is obtained as y = 0.1356x - 6.2119, R 2 =0.9966, such as Fig. 1 . Fig. 2 The derivative spectra of each sample in the range of 3000-3120 data points under order 1.5.

[0036] 4) According to the formula, the content of DL-aspartic acid in H1 can be calculated to be 43.67%, with an absolute error of 1.53% and a relative error of 3.38% compared with the prepared content.

[0037] 4. Analysis of DL-aspartic acid in H2:

[0038] 1) Weigh 5 portions of 1.0g each into the multi-component system H2 to be tested, and add 0.05 g, 0.1g, 0.15g, 0.2g, and 0.25g of the analyte DL asparagic acid according to the standard addition method, grind and mix to obtain C1-C5;

[0039] 2) Call function S = SHfd(Mraw, 1.5, 100, 5), where Mraw is the spectrum to be processed, C1-C5. Run the search to obtain the derivative values ​​of C1-C5 at the 3070 data point position in order 1.5, and obtain the relationship between the amount added and the derivative value: y = -0.1138x - 8.113, R. 2 =0.9917. Based on the relationship, the content of DL-aspartic acid in H2 can be calculated to be 71.29%, with an absolute error of 2.79% and a relative error of 4.07% compared to the prepared content.

[0040] The calculation results show that the method of the present invention can quickly and conveniently quantify the content of analytes in a multi-component system.

[0041] Example 2

[0042] Determination of sucrose content in sugarcane juice

[0043] 1. Take 10mL of freshly squeezed sugarcane juice in 5 portions, and add 0.5001g, 1.0030g, 1.5008g, 2.0022g, and 2.5001g of sucrose (AR) in sequence. After stirring and dissolving, ZZT1-ZZT5 are obtained.

[0044] 2. Near-infrared spectra of sugarcane juice ZZT1-ZZT5 and sucrose ZT were collected using a near-infrared spectrometer (PE). The acquisition parameters were: integration time 60 s, resolution 2 cm⁻¹. -1 The data interval is 1nm, and the wavelength range is 1000-2500nm.

[0045] 3. In the Matlab computing platform, call the function S = SHfd(Mraw, v, b, p), where Mraw is the spectrum to be processed, ZZT1-ZZT5; set b=100, p=4, and calculate the continuous order derivative spectra of the spectrum ZZT1-ZZT5, where v = 1.8 + kΔ (k=0, 1, 2…10; Δ=0.01, 1.8). <v<2.5);

[0046] The calibration peak position λ of sucrose was determined to be at data point 3349. The derivative order vJ was 2.11, yielding the linear relationship between the derivative value of the cane juice spectrum and the amount of sucrose added: y = -3.1016x - 4.6085, r = 0.9902. Based on this relationship, the sucrose content in the cane juice was calculated to be 148.58 mg / mL.

[0047] 4. Freshly squeezed sugarcane juice was diluted 20 times with pure water. 0.5 mL of the sample was placed in a 10 mL volumetric flask, brought to volume with pure water, and filtered through a 0.22 μm microporous membrane. HPLC analysis was performed (amino column, flow rate 1 mL / min, mobile phase acetonitrile:water = 80:20 (V / V), differential refractive index detector, column temperature 35℃, injection volume: 20 μL). The calculated sucrose content in the sugarcane juice was 145.60 mg / mL.

[0048] 5. Compared with HPLC determination, the absolute error of the method of this invention is 2.98 mg / mL, and the relative error is 2.05%. This method only requires the acquisition of sample spectra, and direct quantification can be achieved through the fractional derivative of the spectra.

[0049] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions conceived without inventive effort should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A direct quantitative analysis method for fractional derivative spectra of multi-component systems; It consists of the following steps: 1) The sample spectrum of the multi-component system h to be tested was acquired using a spectrometer. H ; 2) Following the standard addition method, add [the following to] the multi-component system h to be tested: n The analyte f at several gradient concentrations was mixed thoroughly, and the spectra of each sample were acquired using a spectrometer to obtain... F 1- F n; 3) Call the SH algorithm: function S = SHfd (Mraw, v , b , p ), where Mraw is F 1 - F n, set b as a positive real number, adjusted according to the signal noise intensity; p is a positive real number, with a value > 2, v = v0 + kΔ, where k = 0, 2…m; 0.01 < Δ < 0.1, 0 < v < 4; obtain F 1 - F the continuous order derivative of n DF 1 - DF n; according to 0 < v < 4, 0.01 < Δ < 0.1, m is 40, and the m sequence is 0, 2, 4, 6 until 40; 4) Determine the calibration peak position λ and the derivative order vJ: In DF 1- DF In n, the order vJ when the correlation between the derivative value at a certain peak position and the amount added is greater than 0.99 is searched, and the data point corresponding to that peak position is the calibration peak position λ. 5) Under a defined vJ and calibrated peak position λ, by plotting the relationship between the added amount C and the derivative value DF, DF=aC+b, the content of the analyte f in the multi-component system H can be calculated. C: n gradient concentrations C1, C2, ..., Cn.

2. The method according to claim 1, characterized in that: In step 2), n >5.

Images