Method for rapidly detecting raffinate concentration of raffinate in hydrogen peroxide production process

Abstract

The invention relates to the field of spectrum detection, and discloses a method for rapidly detecting the raffinate concentration of raffinate in the hydrogen peroxide production process. The method comprises the following steps: (1) obtaining a plurality of standard raffinate; (2) obtaining a characteristic spectrum wave band of the standard raffinate; (3) respectively carrying out outer product and outer sum on the characteristic spectrum wave bands to obtain three-dimensional spectrums Xwj and Xwh; (4) respectively carrying out regression analysis on the three-dimensional spectrum and the raffinate concentration of a standard raffinate to obtain regression models Ywj and Ywh; fusing the Ywj and the Ywh to obtain a fusion model M; and (5) obtaining raffinate with unknown raffinate concentration, obtaining corresponding Xwj'and Xwh 'according to the steps (2)-(3), and substituting Xwj' and Xwh 'into the fusion model M to obtain the raffinate concentration to be measured. The method can quickly, simply and accurately obtain the raffinate concentration in the raffinate.

Description

Technical Field

[0001] The invention relates to the field of spectrum detection, and in particular to a method for rapidly detecting the raffinate concentration of raffinate in a hydrogen peroxide production process. Background Art

[0002] The production of hydrogen peroxide (hydrogen peroxide) mainly adopts the anthraquinone method. However, the raw materials, intermediates and products in the anthraquinone method are almost all flammable and explosive or combustion-supporting substances. Therefore, it is particularly necessary to accurately and timely monitor the quality indicators of the hydrogen peroxide production process and make real-time adjustments according to the situation.

[0003] The raffinate is the liquid remaining after hydrogen peroxide is extracted from the oxidation reaction during the production of hydrogen peroxide by the anthraquinone method. The main components are water, some catalysts and organic impurities such as (organic acid), and a small amount of hydrogen peroxide. The concentration of hydrogen peroxide is the raffinate concentration. The raffinate concentration is an important analytical item in the production of hydrogen peroxide by the anthraquinone method. The raffinate concentration can reflect the quality of hydrogen peroxide production. The determination method of the raffinate concentration is generally: take part of the sample, add potassium permanganate titration solution, and calculate the corresponding raffinate concentration (that is, the actual hydrogen peroxide content in the sample) according to the content of the titration solution consumed at the titration endpoint, in g / L. In order to ensure the quality indicators of hydrogen peroxide production, the analysis interval of the production process is very short, and it needs to be analyzed once every 2 hours on average. The current mainstream manual titration method is cumbersome, has high requirements for operators, and the analysis speed of this method is slow, which is obviously not suitable for the production process of hydrogen peroxide.

[0004] Therefore, it is necessary to develop a method for accurately, simply and quickly obtaining the raffinate concentration of the raffinate. Summary of the invention

[0005] The purpose of the present invention is to overcome the above problems existing in the prior art and to provide a method for quickly detecting the raffinate concentration of the raffinate in the hydrogen peroxide production process, which method can quickly, simply and accurately obtain the raffinate concentration in the raffinate.

[0006] In order to achieve the above object, the present invention provides a method for quickly detecting the raffinate concentration of the raffinate in the hydrogen peroxide production process, the method comprising:

[0007] (1) obtaining a plurality of standard raffinates, wherein the raffinate concentrations of the plurality of standard raffinates are known and different;

[0008] (2) obtaining a spectrum of a standard raffinate and selecting a characteristic spectral band;

[0009] (3) Perform outer product and outer sum on the characteristic spectral bands to obtain the three-dimensional spectrum X wj and three-dimensional spectrum X wh ;

[0010] (4) The above three-dimensional spectrum is regressed with the raffinate concentration of the standard raffinate to obtain the regression model Y wj and the regression model Y wh ;

[0011] The regression model Y wj and the regression model Y wh Perform fusion to obtain a fusion model M;

[0012] (5) Obtain a raffinate with an unknown raffinate concentration, and obtain the corresponding three-dimensional spectrum X according to steps (2)-(3) wj ' and three-dimensional spectrum X wh ', and X wj ' and X wh 'Substitute into the fusion model M to obtain the raffinate concentration to be measured.

[0013] Through the above technical solution, the raffinate concentration of the raffinate can be obtained quickly, simply and accurately. Compared with the prior art, the method provided by the present invention can avoid complex manual processing processes, is more efficient, greatly improves the analysis operation environment, and has obvious advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a correlation diagram of the measured values ​​and calculated values ​​of the verification set in Example 1. DETAILED DESCRIPTION

[0015] The endpoints and any values ​​of the ranges disclosed in this article are not limited to the precise ranges or values, and these ranges or values ​​should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of each range, the endpoint values ​​of each range and the individual point values, and the individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed in this article.

[0016] The present invention provides a method for quickly detecting the raffinate concentration of a raffinate in a hydrogen peroxide production process, the method comprising:

[0017] (1) obtaining a plurality of standard raffinates, wherein the raffinate concentrations of the plurality of standard raffinates are known and different;

[0018] (2) obtaining a spectrum of a standard raffinate and selecting a characteristic spectral band;

[0019] (3) Perform outer product and outer sum on the characteristic spectral bands to obtain the three-dimensional spectrum X wj and three-dimensional spectrum X wh ;

[0020] (4) The above three-dimensional spectrum is regressed with the raffinate concentration of the standard raffinate to obtain the regression model Y wj and the regression model Y wh ;

[0021] The regression model Y wj and the regression model Y wh Perform fusion to obtain a fusion model M;

[0022] (5) Obtain a raffinate with an unknown raffinate concentration, and obtain the corresponding three-dimensional spectrum X according to steps (2)-(3) wj ' and three-dimensional spectrum X wh ', and X wj ' and X wh 'Substitute into the fusion model M to obtain the raffinate concentration to be measured.

[0023] It is understandable that hydrogen peroxide is mainly produced by the anthraquinone method, and the main reaction liquids involved in the production process generally include hydrogenation liquid, oxidation liquid and raffinate. The main process of the anthraquinone method includes: preparing anthraquinone (generally 2-ethyl anthraquinone) and an organic solvent (such as a mixed solvent of C9-C10 heavy aromatics, trioctyl phosphate and tetrabutyl urea) into a working liquid, and controlling the total effective anthraquinone content to about 160g / L. Hydrogenation stage: Under the conditions of a pressure of 0.3MPa or above, a temperature of 40-80°C and a catalyst (such as a Pd catalyst), the anthraquinone in the working liquid is hydrogenated. 2 Hydrogenation reduction to obtain anthrahydroquinone or tetrahydroanthrahydroquinone; in the oxidation stage, the material after the hydrogenation stage is reacted with O at 30-60°C and slightly compressed. 2 Oxidation reaction is carried out to oxidize anthrahydroquinone and tetrahydroanthrahydroquinone to generate H 2 O 2 and anthraquinone; then the material after the oxidation stage is extracted (the raffinate is the extracted H 2 O 2 The remaining solution is then regenerated, refined and concentrated to produce 20wt%-50wt% H 2 O 2 Aqueous solution.

[0024] The inventors of the present invention have found in their research that for the raffinate in the anthraquinone process, processing according to the above method, especially performing outer product and outer sum processing to obtain a three-dimensional spectrum, can effectively avoid the influence of overlapping peaks and can fully extract the information related to the raffinate concentration in the spectrum, so that the result has high accuracy; by fusing the regression model to obtain a fusion model, when it is used to calculate the unknown raffinate concentration, it can also ensure that a high-accuracy result is obtained.

[0025] Generally, in the three-dimensional processing of the spectrum, the three dimensions of the obtained three-dimensional spectrum are wave number (which can be the X-axis and can range from 1000-10000cm -1 ), absorbance (which can be the Y axis, and the range can be 0-100%) and band type (which can be the Z axis). Compared with other materials, the method of the present invention can obtain more accurate results for the raffinate in the hydrogen peroxide production process.

[0026] According to the present invention, preferably, in step (2), the spectrum is a near infrared spectrum or a mid infrared spectrum, more preferably a near infrared spectrum. The inventors of the present invention further found that when a near infrared spectrum is used, the accuracy of the result can be further ensured.

[0027] When collecting near infrared spectrum, the sample can be injected into about two-thirds of the volume of the cuvette, and the cuvette containing the sample can be sealed with a sealing film, and the sealed cuvette can be placed in a temperature-controlled sample cell rack for spectrum collection. The optical path of the cuvette can be 1-5 mm. The instrument when collecting the near infrared spectrum is not particularly limited and can be a conventional choice in the art, for example, Antaris II near infrared spectrometer (Thermo Fisher Company) can be used.

[0028] According to the present invention, preferably, before performing outer product and outer sum processing, the method further comprises: preprocessing and segmenting the characteristic spectral band.

[0029] According to the present invention, preferably, the characteristic spectrum band is preprocessed and divided into two sections, and the wave number interval of the first section includes 3880-4204cm -1 and 4458-5164cm -1 The second wave number range includes 5168-6194cm -1 It can be understood that the outer product and outer sum refer to operations between matrices (or vectors), generally between two matrices (or vectors). After the characteristic spectrum band is divided into two segments as above (the first segment includes two parts, the two parts are spliced ​​to obtain the first segment, and the two segments respectively correspond to a matrix, and the two matrices are outer-producted to obtain the three-dimensional spectrum X wj ; Perform the outer sum of the two matrices to obtain the three-dimensional spectrum X wh The characteristic spectrum band is divided into two segments as above. During calculation, the number of rows and columns of the matrices corresponding to the two characteristic spectrum bands are the same.

[0030] According to the present invention, preferably, the pretreatment method is selected from the second order differential or the first order differential, and more preferably the second order differential. The inventors of the present invention further found that by adopting the above pretreatment method, even if the raffinate concentration in the raffinate is small, sufficient information can be obtained, which can further improve the accuracy of the results.

[0031] The window width of the second-order differential may be 15-28 (for example, 15, 18, 20, 21, 23, 25, 28).

[0032] According to the present invention, preferably, in step (1), the raffinate concentration of the standard raffinate is 0.01-1 g / L, more preferably 0.02-0.8 g / L (for example, 0.02 g / L, 0.04 g / L, 0.05 g / L, 0.1 g / L, 0.2 g / L, 0.3 g / L, 0.4 g / L, 0.5 g / L, 0.6 g / L, 0.7 g / L, 0.8 g / L). It can be understood that the raffinate concentration of the raffinate refers to the concentration of hydrogen peroxide in the remaining liquid after the hydrogen peroxide is extracted, and the concentration is generally small. By using the method provided by the present invention, the information in the spectrum of the raffinate with a small raffinate concentration can also be fully extracted.

[0033] When collecting samples, it is generally required that the sample conditions are evenly distributed within a possible range, and the raffinate concentration in the standard raffinate is also evenly distributed within the above range.

[0034] According to the present invention, preferably, in step (1), the number of standard raffinates is not less than 28, and more preferably is 50-210 (for example, it can be 50, 80, 100, 130, 150, 180, 200, 210, and a range formed by any two of the above values, and a value within the range).

[0035] The present invention has no particular limitation on the conditions for obtaining the spectrum. However, preferably, the conditions for obtaining the near-infrared spectrum include: a temperature of 12-52°C, more preferably 20-30°C (for example, 20°C, 21°C, 22°C, 23°C, 24°C, 25°C, 26°C, 27°C, 28°C, 29°C, 30°C); a wave number interval of 3200-10800cm -1 ; Resolution is 2-18cm -1 , more preferably 6-10cm -1 . Can scan 100-140 times.

[0036] According to the present invention, preferably, the characteristic spectral band is in the wave number interval of 3880-4204cm -1 and 4458-6194cm -1 The inventors of the present invention further discovered that the selection of the above characteristic spectral bands can further fully reflect the information related to the raffinate concentration.

[0037] The present invention has no particular limitation on the specific method of regression analysis. However, preferably, the regression analysis method is multidimensional partial least squares analysis. The above regression analysis method can further ensure the accuracy of the results.

[0038] According to the present invention, preferably, the fusion method includes: wj Substitute into the regression model Y wj The predicted value R is obtained wj ;

[0039] The three-dimensional spectrum X wh Substitute into the regression model Y wh The predicted value R is obtained wh ;

[0040] The predicted value R wj and the predicted value R wh , assign values ​​a and b respectively, where a and b are respectively greater than 0 and less than 1, and a+b=1; according to aR wj +bR wh , adjust the values ​​of a and b to, aR wj +bR wh The calculated result is close to the actual raffinate concentration of the raffinate, and the fusion model M = aY wj +bY wh .

[0041] It can be understood that the actual raffinate concentration refers to the raffinate concentration obtained by the currently available manual titration method after obtaining the standard raffinate. In addition, in the process of adjusting a and b, there are many combinations that can be used, and the values ​​of a and b that can make the calculation results closer to the actual raffinate concentration can be taken from these combinations.

[0042] Generally, the main components of the raffinate are water, some organic impurities (such as organic acids), catalysts, etc., and also contain a small amount of hydrogen peroxide.

[0043] The inventors of the present invention further discovered in the study that for the raffinate in the process of preparing hydrogen peroxide by the anthraquinone method, by selecting the specific characteristic spectral bands defined by the present invention, and coordinating the spectral processing methods of second-order differential, segmentation, outer product and outer sum as described above, especially coordinating the multidimensional partial least squares analysis applicable to the three-dimensional spectrum, and fusing the obtained regression models in the manner defined by the present invention, the obtained fusion model has a higher accuracy. Compared with the use of other spectra, or with other processing methods for the spectrum, or not obtaining a three-dimensional spectrum and using a method (such as partial least squares analysis) applicable to constructing a regression model corresponding to a two-dimensional spectrum, the preferred near-infrared spectrum of the present invention, and the processing method of the present invention, the obtained fusion model has a higher accuracy and the accuracy of the calculation result is higher. The specific method provided by the present invention is particularly suitable for determining the raffinate efficiency of the raffinate in the process of preparing hydrogen peroxide by the anthraquinone method.

[0044] The present invention will be described in detail below through examples.

[0045] In the following examples, the raffinate is obtained during the process of preparing hydrogen peroxide by the anthraquinone process.

[0046] In step (1) of the following examples, the raffinate concentration of the standard raffinate is determined by potassium permanganate titration, and the raffinate concentration value obtained by potassium permanganate titration is recorded as the measured value.

[0047] The instrument used to collect near-infrared spectra was Antaris II near-infrared spectrometer (Thermo Fisher).

[0048] The operations after spectrum acquisition were performed in MATLAB.

[0049] Example 1

[0050] Used to illustrate the method provided by the present invention for quickly detecting the raffinate concentration of the raffinate in the process of hydrogen peroxide production

[0051] (1) 54 raffinates were obtained and the raffinate concentrations thereof were measured (the raffinate concentration values ​​were distributed in the range of 0.041-0.755 g / L, and the difference between every two adjacent raffinate concentration values ​​was in the range of 0.01-0.1 g / L).

[0052] (2) Inject the sample into a cuvette (optical path length of 2 mm) until it reaches two-thirds of the cuvette volume, seal the cuvette containing the sample with a sealing film, and place the sealed cuvette into a temperature-controlled sample cell rack for spectrum collection. The conditions for collecting near-infrared spectra include: temperature of 25°C, and the collected wavenumber range of 3500-10000cm -1 , resolution 8cm -1, scanned 128 times. For the obtained near-infrared spectrum, the characteristic spectrum band 3880-4204cm was selected -1 and 4458-6194cm -1 .

[0053] (3) The characteristic spectral band is subjected to second-order differentiation with a window width of 21, and then divided into two segments (the wave number range of the first segment is 3880-4204 cm -1 and 4458-5164cm -1 The second wave number range is 5168-6194cm -1 ). Take the outer product of the two bands and get the three-dimensional spectrum X wj ; Combine the two spectra to obtain a three-dimensional spectrum X wh .

[0054] (4) The three-dimensional spectrum X wj , and the raffinate concentration of the raffinate were subjected to multidimensional partial least squares analysis to obtain the regression model Y wj ;

[0055] The three-dimensional spectrum X wh , and the raffinate concentration of the raffinate were subjected to multidimensional partial least squares analysis to obtain the regression model Y wh .

[0056] The three-dimensional spectrum X wj , substitute into the regression model Y wj The predicted value R is obtained wj ;

[0057] The three-dimensional spectrum X wh , substitute into the regression model Y wh The predicted value R is obtained wh .

[0058] The predicted value R wj and the predicted value R wh , assign values ​​a and b respectively, according to aR wj +bR wh , adjust the values ​​of a and b to, aR wj +bR whh The calculated results are close to the actual raffinate concentration of the raffinate, and a is 0.628 and b is 0.372;

[0059] Then we get the fusion model M = aY wj +bY wh .

[0060] (5) Take 15 raffinate solutions with unknown raffinate concentration as the validation set and perform the following processing on them:

[0061] The method in steps (1) to (3) is used to first determine the raffinate concentration by potassium permanganate titration, and collect near-infrared spectra to obtain corresponding characteristic spectral bands, and obtain a three-dimensional spectrum X wj ' and X wh '. wj ' and X wh 'Substitute into the fusion model M to obtain the raffinate concentration (calculated value).

[0062] The measured and calculated values ​​of the 15 samples in the validation set, as well as the deviation between the two (measured value minus calculated value) are shown in Table 1. In addition, the root mean square error of prediction (RMSEP) and R (correlation coefficient) are used to evaluate the performance of the fusion model, where

[0063]

[0064] Among them, n is the total number of samples in the validation set, y i is the measured value of the i-th sample, is the calculated value of the ith sample.

[0065]

[0066] Among them, n is the total number of samples in the validation set, y i is the measured value of the i-th sample, y i ^ is the calculated value of the i-th sample, y — is the average value of the actual values ​​of the validation set samples. By calculation, we can get R = 0.956.

[0067] Table 1

[0068]

[0069] In addition, a correlation diagram is drawn with the measured values ​​(x-axis) and calculated values ​​(y-axis) of the 15 samples in the validation set, see Figure 1 (where R is the correlation coefficient).

[0070] It can be seen that the method provided by the present invention can quickly, simply and accurately determine the raffinate concentration of the raffinate in the process of producing hydrogen peroxide by the anthraquinone method, and the prediction results at various concentrations are relatively accurate and have high stability.

[0071] Comparative Example 1

[0072] According to steps (1)-(2) in Example 1, a characteristic spectral band is obtained. Then, the characteristic spectral band is subjected to a second-order differential with a window width of 21, and then the band after the second-order differential and the raffinate concentration of the raffinate are subjected to a partial least squares analysis to obtain a model Y D .

[0073] Take 15 raffinate solutions with unknown raffinate concentration as the validation set, and proceed to "obtain the corresponding characteristic spectral band" according to step (5) in Example 1, and perform the second-order differential of the characteristic spectral band with a window width of 21, and substitute the band obtained at this time into the model Y D In the above equation, we get model Y D The results are shown in Table 2.

[0074] Calculate model Y in the same way as in Example 1 D The root mean square error of prediction (RMSEP) and the bias of the obtained results are shown in Table 2.

[0075] Table 2

[0076]

[0077] RMSEP can characterize the accuracy of the method. The smaller the RMSEP, the more accurate the result. It can be seen that the method provided by the present invention has a smaller deviation from the actual measured value of the standard method, and the RMSEP is also smaller. The method provided by the present invention has better accuracy.

[0078] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited thereto. Within the technical concept of the present invention, the technical solution of the present invention can be subjected to a variety of simple modifications, including the combination of various technical features in any other suitable manner, and these simple modifications and combinations should also be regarded as the contents disclosed by the present invention and belong to the protection scope of the present invention.

Claims

1. A method for rapidly detecting the raffinate concentration of raffinate in a hydrogen peroxide production process, characterized in that: The method includes: (1) obtaining a plurality of standard raffinates, wherein the raffinate concentrations of the plurality of standard raffinates are known and different; (2) obtaining a spectrum of a standard raffinate and selecting a characteristic spectral band; (3) Perform outer product and outer sum on the characteristic spectral bands to obtain the three-dimensional spectrum X wj and three-dimensional spectrum X wh ; (4) The above three-dimensional spectrum is regressed with the raffinate concentration of the standard raffinate to obtain the regression model Y wj and the regression model Y wh ; The regression model Y wj and the regression model Y wh Perform fusion to obtain a fusion model M; (5) Obtain a raffinate with an unknown raffinate concentration, and obtain the corresponding three-dimensional spectrum X according to steps (2)-(3) wj ' and three-dimensional spectrum X wh ', and X wj ' and X wh 'Substitute into the fusion model M to obtain the raffinate concentration to be measured.

2. The method according to claim 1, wherein: In step (2), the spectrum is a near-infrared spectrum or a mid-infrared spectrum, preferably a near-infrared spectrum.

3. The method according to claim 1 or 2, wherein: Before performing the outer product and outer sum processing, the method further includes: preprocessing and segmenting the characteristic spectrum band; Preferably, the characteristic spectrum band is preprocessed and divided into two sections, the wave number range of the first section includes 3880-4204cm -1 and 4458-5164cm -1 The second wave number range includes 5168-6194cm -1 .

4. The method according to claim 3, wherein: The preprocessing method is selected from the second order differential or the first order differential, and is more preferably the second order differential.

5. The method according to claim 1, wherein: In step (1), the raffinate concentration of the standard raffinate is 0.01-1 g / L, preferably 0.02-0.8 g / L.

6. The method according to claim 1 or 5, wherein: In step (1), the number of standard raffinates is not less than 28, preferably 50-210.

7. The method according to claim 2, wherein: The conditions for obtaining near-infrared spectra include: temperature of 12-52°C, preferably 20-30°C; wave number range of 3200-10800cm -1 ; Resolution 2-18cm -1 , preferably 6-10cm -1 .

8. The method according to claim 1 or 7, wherein: The wave number range of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 .

9. The method according to claim 1, wherein: The regression analysis method is multidimensional partial least squares analysis.

10. The method according to claim 1 or 9, wherein: The fusion method includes: wj Substitute into the regression model Y wj The predicted value R is obtained wj ; The three-dimensional spectrum X wh Substitute into the regression model Y wh The predicted value R is obtained wh ; The predicted value R wj and the predicted value R wh , assign values ​​a and b respectively, where a and b are respectively greater than 0 and less than 1, and a+b=1; according to aR wj +bR wh , adjust the values ​​of a and b to, aR wj +bR wh The calculated result is close to the actual raffinate concentration of the raffinate, and the fusion model M = aY wj +bY wh .

Images