In-situ online infrared spectroscopy monitoring equipment, system and method for crystallization process control

Abstract

The present invention provides an in-situ online infrared spectrum monitoring device, system and method for controlling a crystallization process, including an optical fiber probe, an infrared host, a temperature control device, a crystallizer, a computer and related software modules: an infrared spectrum acquisition module, a concentration measurement module and a model building module. The optical fiber probe includes an input and output optical fiber, a probe tube and an attenuated total reflection crystal. The probe tube is located at the front end of the incident light fiber and the output light fiber, and the attenuated total reflection crystal is arranged at the end of the probe tube; the spectrum data collected by the infrared host is transmitted to the computer through a USB data cable and displayed through the spectrum acquisition module. The spectrum data is combined with a chemometric intelligent algorithm, and the concentration value of the sample to be tested is calculated in real time through the concentration measurement module; the computer adjusts the temperature control device temperature rise and fall program to realize the process of heating dissolution and cooling crystallization of the sample. This system can realize the temperature control of the crystallization process and the real-time in-situ online measurement of the solution concentration.

Description

Technical Field

[0001] The present invention relates to the technical field of online measurement process control, and in particular to an in-situ online infrared spectrum monitoring device, system and method for crystallization process control. Background Art

[0002] In recent years, with the accelerated development of refined production, process analytical technology (PAT) and production "visualization" have been valued by domestic and foreign companies. Online PAT technology plays an important role in reducing production time, improving product quality, and improving the degree of automation. The rapid development of process analytical technology has realized the real-time measurement of key process parameters and multiple product quality indicators, and has become an important tool for strengthening process control, especially in the pharmaceutical process. Crystallization is an extremely important process in pharmaceutical production. More than 90% of the active ingredients of drugs exist in the form of crystals. In the crystallization process, the concentration of the solution is one of the most important variables. It is the main driving force for crystal nucleation and growth. The early online concentration measurement methods mainly include density method, conductivity method, calorimetry, and refractive index method. The density method determines the concentration by measuring the density of a solution without crystals. However, in the industrial crystallization process, the solution often contains impurities of different concentrations. It is necessary to take a suspension of crystals, filter the crystals, and measure the density of the liquid. The process is relatively complicated and difficult to operate, such as clogging the sieve and temperature fluctuations during sampling. Therefore, the density method will be inaccurate. Although the conductivity method does not require external sampling, its limitations are that it is not applicable to organic systems in most cases (because organic systems are usually not conductors) and is very sensitive to temperature and impurities. Although calorimetry has been proven to be a very effective online sensor in fields such as chemical reactions, it is less widely used due to the weak thermal effect of the crystallization process. Although this method is effective when the refractive index changes with the concentration of the solution, it is sensitive to external light and air bubbles. In addition, multiple substances with different refractive indices may appear in the crystallization system.

[0003] Infrared spectroscopy combined with chemometric models can achieve concentration measurement, but the traditional Fourier transform infrared spectroscopy (FTIR) is to guide light into the measured medium, but it cannot be directly used for the measurement of crystal suspensions, because the crystal will block the light beam, resulting in the detector not receiving the signal. The emergence of attenuated total reflection technology provides a possible solution to this problem and is an important promoter in the field of modern crystallization process control. Attenuated total reflection technology has the advantages of simple sample preparation and cleaning, no need for pretreatment, no damage to the sample, and can eliminate the influence of factors such as light scattering. The measured infrared spectrum has high quality, no interference fringes, and clear and undistorted characteristic bands. In the process of a beam of light being transmitted by an incident optical fiber through the ATR crystal and returning to the detector, the beam will not be directly reflected back at the inner surface of the ATR crystal in contact with the measured medium, but will generate an attenuated wave. This attenuated wave will extend into the medium to a penetration depth of about 1-10um, and then be reflected back. The reflected beam will be affected by the medium in which the attenuated wave is located, and the chemical information of the medium can be obtained. Based on this principle, the present invention has developed a set of in-situ online infrared spectrum monitoring equipment, system and method for crystallization process control. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and to provide an in-situ online infrared spectroscopy monitoring device, system and method for controlling a crystallization process, so as to achieve real-time online measurement of the concentration of the crystallization process.

[0005] The present invention is achieved in that:

[0006] In the first aspect, the present invention provides an in-situ online infrared spectrum monitoring device for controlling a crystallization process, comprising an optical fiber probe, an infrared host, a temperature control device, a crystallizer and a computer, an infrared host, the infrared host comprising an infrared light source, an interferometer beam splitter, a moving mirror, a fixed mirror, an MCT detector, a semiconductor electronic refrigeration device, an infrared reflector, a laser and a power supply. An optical fiber probe, the optical fiber probe comprising an incident light optical fiber, an outgoing light optical fiber, a probe tube and an attenuated total reflection crystal, the probe tube is placed at the front end of the incident light optical fiber and the outgoing light optical fiber, the attenuated total reflection crystal is placed at the end of the probe tube, the infrared host is connected to the optical fiber probe through an optical fiber coupler SMA interface, the infrared host is connected to the computer for communication, the computer is connected to the temperature control device for communication, and the temperature control device is used to control the temperature of the crystallizer.

[0007] The Fourier infrared host is developed based on the principle of Fourier transform of infrared light after interference. It consists of an infrared light source, an interferometer (zinc selenide (ZnSe)) beam splitter, a moving mirror, a fixed mirror, a permanent alignment, a double retroreflector design, an MCT detector with four-stage semiconductor electronic cooling, and various infrared reflectors, lasers, control circuit boards and power supplies. The infrared host can detect infrared light in the range of 2-12um (5000-800cm -1 ).

[0008] There are two optical fibers for incident and outgoing light. The optical fiber materials are divided into sulfide glass fiber, which can transmit infrared wavelengths in the range of 1-6um (10000-1600cm-1), and silver halide polycrystalline glass fiber, which can transmit infrared wavelengths in the range of 3-18um (3300-550cm-1). The optical fiber protective cover is made of spiral stainless steel flat wire, with a silicone rubber sheath on the outer layer. The front section near the crystal is wrapped with a 20-100cm long probe made of Hastelloy material, which is used for concentration measurement of crystallization process in reaction vessels of different capacities. According to the absorption range of infrared light of the substance to be measured, the appropriate optical fiber material is selected. This system selects silver halide polycrystalline glass fiber in the mid-infrared spectrum range. The end of the optical fiber is directly screwed into the spectrum host through the optical fiber coupler.

[0009] Attenuated total reflection (ATR) crystals are mainly made of diamond, silicon, germanium, zinc sulfide (ZnS), zinc selenide (ZnSe), etc. Diamond is more sensitive than zinc selenide, but it is limited at 2000cm -1 The transmittance at the left and right sides is low, and it is resistant to strong acid and alkali chemical properties, but it is expensive. Silicon has a low attenuation to light and is not resistant to strong acid and alkali. Zinc selenide can pass a wider spectral range than diamond and has good resistance to acid and alkali, but it is not resistant to halogenated reagents or strong hydrogenation. According to the different requirements of industrial users, choose the corresponding crystal that can adapt to various environments. The size of the ATR crystal probe is reasonably designed to make as much infrared light as possible produce an attenuation wave on the surface of the ATR crystal, and then be reflected back, so as to obtain the chemical information of the medium. This system uses diamond crystal, and diamond has a reflection angle of 45°.

[0010] The immersed attenuated total reflection infrared fiber optic probe is composed of incident and output optical fibers, an external Hastelloy tube sleeve, and an ATR crystal. It can be directly inserted into the liquid sample, making it plug-and-play, without the need for sample preparation, and can achieve in-situ online real-time monitoring.

[0011] Computer is used as the carrier of the in-situ on-line infrared spectrum monitoring system for crystallization process control.

[0012] In a second aspect, the present invention provides an in-situ online infrared spectroscopy monitoring system for crystallization process control, including an infrared spectroscopy acquisition module, a concentration prediction module, and a model building module;

[0013] The infrared spectrum acquisition module has the following main functions: automatic background subtraction, spectrum acquisition speed of 1 second / time, real-time display of infrared spectra of different wave numbers and different times, and can achieve continuous automatic storage, forming a three-dimensional image with X-axis as wave number (or wavelength), Y-axis as absorbance (or attenuation) and Z-axis as time, which can intuitively show the changes of infrared spectrum during the reaction process; resolution 4cm -1 and 8cm -1 Can switch automatically; can view saved local files in real time.

[0014] The concentration prediction module can predict the concentration in real time according to the selected model, generate the reaction component concentration change curve, and display it in a window at the same time, and update it automatically; it has a real-time display list of concentration and sample information, and displays it in chronological order; it has the function of setting control lines, and issues warning prompts when the predicted indicators exceed the control standards; it has the function of distinguishing abnormal spectra and saving the product batch number information corresponding to the spectrum; when the module is closed, all data can be automatically saved, and the result report can be printed.

[0015] The key to using the concentration prediction module is to establish a quantitative analysis model based on the chemometric algorithm. The model building module has partial least squares PLS, principal element analysis PCA, principal element regression PCR, neural network NN and support vector machine SVM linear and nonlinear modeling techniques; including first-order derivatives, second-order derivatives, smoothing and other signal preprocessing functions; it has a special function of automatic selection of characteristic bands, including genetic algorithms, interval partial least squares, elimination of uninformative variables and other methods. All data and modeling methods are integrated.

[0016] In a third aspect, the present invention provides an in-situ online infrared spectroscopy monitoring method for crystallization process control, comprising the following steps:

[0017] S1. Build a cooling crystallization experimental device, set up a jacketed crystallizer, and equip it with a condenser to prevent volatilization. Insert the optical fiber probe into the crystallizer. The optical fiber probe is directly connected to the infrared spectrometer. The infrared host collects the signal and transmits it to the computer, and displays it in real time on the computer display screen. After selecting the calibrated cooling crystallization model of the crystallization sample, the concentration measurement module can predict the concentration value in the crystallization process in real time. The temperature is controlled by the temperature control device, and the stirrer with adjustable stirring speed is used to conduct the experiment to realize the concentration change of the crystallization sample during the crystallization process;

[0018] S2. Collect samples for model calibration, establish a model of infrared spectrum and corresponding concentration, take concentration points at different temperatures every 5°C within the set temperature range, prepare n groups of samples with different concentrations, where concentration refers to the mass ratio of the crystallization sample to the solvent, pour the solvent into the crystallizer, turn on the temperature control device to keep the temperature of the solvent in the crystallizer at the set value, add the crystallization sample, stir continuously, and collect the infrared spectrum of the specified concentration at the set temperature through the infrared spectrum collection module, collect three spectra continuously, and take the average spectrum for standby use;

[0019] S3. Then, the temperature is raised to the next set temperature point, and the solute of the crystallized sample is weighed to make the concentration reach the second set concentration value. After the solute is completely dissolved, infrared spectra are continuously collected and recorded and saved, and the average spectrum is taken for later use. This step is repeated so that the corresponding infrared spectra are collected for all concentration points designed in S2, thereby obtaining infrared spectra of the crystallized sample at different temperatures and different concentrations.

[0020] S4. Arrange the acquired infrared spectrum data, randomly select several groups of samples as verification data, and the rest as training data to establish a partial least squares model.

[0021] S5. Selection of preprocessing methods for training data. A partial least squares model is established for the training set through methods including comparative smoothing, baseline removal, and first-order derivative. The prediction ability of the model is evaluated using indicators including training set correlation coefficient, training set root mean square error, leave-one-out cross-validation correlation coefficient, leave-one-out cross-validation root mean square error, validation set correlation coefficient, and validation set root mean square error. The closer the correlation coefficient is to 1, the better. The smaller the validation root mean square error and cross-validation root mean square error, the better. The final preprocessing method selected is a combination of 1D+Detrend.

[0022] S6. After selecting an effective spectral preprocessing method, select an effective band according to the position of the infrared absorption peak of the crystalline sample to establish the best partial least squares model. Similarly, the correlation coefficient is close to 1, and the smaller the root mean square error, the better, and the most preferred model is obtained.

[0023] S7. After the model is established, it is saved in the concentration prediction module and applied to the real-time monitoring of the concentration during the cooling and crystallization process of the crystallization sample.

[0024] S8. Concentration prediction of cooling crystallization process: reconfigure the supersaturated crystallization sample solution, raise the temperature to the specified value, maintain for a period of time, wait for the crystallization sample to completely dissolve, and then cool it down to the specified value. The whole process is stirred at a certain rate. The change process of the crystallization sample dissolving and the concentration value increasing in the heating process can be measured in real time, and the concentration decreases after the crystals appear in the cooling process.

[0025] S9, verify the accuracy of the online infrared model concentration measurement, set up two groups of experiments, the solution ratio is the same as S8, in the first group of experiments, after heating to the specified temperature, the solute is completely dissolved and stirring is stopped. After a period of maintenance, the upper clear liquid is taken with a syringe and put into a weighing bottle with a known mass m1 to weigh m2. After it is completely dried, it is weighed again to obtain m3. The calculated concentration is: In the second group of experiments, the cooling process continued to cool to the specified value, crystals appeared and nucleated. After cooling to the next specified value, stirring was stopped. After a period of time, the upper clear liquid was taken with a syringe and placed in a flask of known mass. The concentration c2 was calculated, and the concentration value measured by the in-situ online infrared spectroscopy monitoring system at the corresponding time was found. The in-situ online infrared spectroscopy monitoring system was verified by calculating the relative error, and the accuracy of the concentration measurement was controlled during the cooling and crystallization process.

[0026] The beneficial effects of the present invention are:

[0027] The data is transmitted to the computer through the communication line and displayed by the spectrum acquisition module. The spectral data is combined with the chemometric algorithm to calculate the concentration value of the sample to be tested in real time and displayed on the concentration display interface. According to the material to be crystallized, the temperature control equipment's heating and cooling program is adjusted to increase and decrease the temperature of the solution in the crystallizer to achieve the cooling and crystallization process of the sample. This system can realize in-situ online measurement of the concentration change of the solution during the crystallization process. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0029] Figure 1 The device diagram of the in-situ online infrared spectroscopy monitoring system for crystallization process control;

[0030] Figure 2 Schematic diagram of an immersed infrared fiber probe with an attenuated total reflection (ATR) crystal;

[0031] Figure 3 This is the interface diagram of the infrared spectrum acquisition module for the cooling and crystallization process of the aspirin-ethanol system;

[0032] Figure 4 This is the 3D infrared spectrum of the cooling crystallization process of aspirin-ethanol system;

[0033] Figure 5 Infrared spectra of potassium aluminum sulfate aqueous solutions with different concentrations at 30°C;

[0034] Figure 6 This is the real-time online prediction module interface of potassium aluminum sulfate concentration during the cooling and crystallization process of potassium aluminum sulfate;

[0035] Figure 7 To build the module for the model, the partial least squares (PLS) method;

[0036] Figure 8 Building modules for the model, neural network (ANN) approach;

[0037] Fig. 9 The infrared absorption of styrene C=C double bond in ethanol solvent in different concentrations and polymerization experiments;

[0038] Fig.10 Real-time concentration changes in the process of producing polystyrene from styrene monomer;

[0039] Fig.11 Flow chart of the in-situ online infrared spectroscopy monitoring method for crystallization process control.

[0040] In the figure: 1. infrared host; 2. computer; 3. optical fiber probe; 301. probe tube; 302. attenuated total reflection crystal; 303. incident light optical fiber; 304. outgoing light optical fiber; 4. mechanical stirring device; 5. thermometer; 6. temperature control equipment; 7. crystallizer. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0044] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0045] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0046] The in-situ online infrared spectrum monitoring system for controlling the crystallization process of the present invention comprises an optical fiber probe, an infrared host, a temperature control device, a crystallizer, a computer and related software acquisition modules: infrared spectrum acquisition and concentration measurement, and model establishment are composed of three modules.

[0047] ATR crystal is made of diamond to adapt to complex, harsh and severe industrial site conditions.

[0048] Optical fiber, the transmission range is 3-18um (3300-550cm -1 )'s silver halide polycrystalline glass fiber optical fiber.

[0049] Example 1: This implementation case measures the concentration of aspirin-ethanol system during the cooling crystallization process in situ and monitors the concentration change in the solution in real time. Using the aspirin-ethanol system, the relationship between infrared spectrum and concentration is studied, and a concentration prediction model is established using a chemometric method based on the Lambert-Beer law. The specific steps are as follows:

[0050] Step (1) build a cooling crystallization experimental device, such as Figure 1As shown. A 250mL jacketed crystallizer is equipped with a condenser to prevent volatilization. The fiber optic probe is inserted into the crystallizer. The other end of the fiber optic probe is directly connected to the infrared spectrometer host. The infrared host collects the signal and transmits it to the computer via a USB2.0 data cable, and displays it in real time in the spectrum acquisition module interface. After selecting the calibrated aspirin cooling crystallization model, the concentration measurement module can predict the concentration value during the crystallization process in real time. The temperature control equipment of JULABO company is used, the temperature is controlled by a circulating water bath, and the magnetic stirrer with adjustable stirring speed of IKA company is used. The drugs aspirin (acetylsalicylic acid) and ethanol are analytically pure from Aladdin Co., Ltd.

[0051] Step (2) After the experimental instrument is set up, test whether the equipment is working properly, and turn on the water bath circulation, magnetic stirring, and infrared spectrometer host according to normal operations. The ATR crystal fiber probe is placed in the air, and the spectrum acquisition module is turned on to collect and subtract the background spectrum to make the infrared spectrum smooth and reduce noise.

[0052] Step (3) Model calibration sample design. The key to the application of in-situ online infrared spectrometer is to establish a model of infrared spectrum and corresponding concentration. In this case, the concentration points at different temperatures are taken every 5°C within the temperature range of 0 to 40°C. The concentration is the mass ratio of aspirin to ethanol. A total of 30 groups of samples are prepared according to the concentration list shown in Table 1:

[0053] Table 1

[0054]

[0055] Step (4) Into a 250 ml jacketed beaker, introduce 150 ml of ethanol, turn on the water bath temperature control to keep the temperature of the ethanol in the beaker at 0°C, add 14.4 g of aspirin, start stirring and continue water bath circulation at the same time, after the aspirin is completely dissolved, turn on the infrared spectrum acquisition module, collect an infrared spectrum at 0°C with a concentration of 0.12 g / g, and continuously collect three spectra to take the average spectrum for standby use.

[0056] Step (5) The temperature is then raised to the next set temperature point, and the aspirin solute is weighed and added to the crystallizer to make the concentration reach the second set concentration value. After the solute is completely dissolved, the infrared spectrum is continuously collected, recorded and saved, and the average spectrum is taken for later use. Repeat this step so that the corresponding infrared spectra are collected for all concentration points designed in step 3. The infrared spectrum of the aspirin tablet-ethanol system is as follows: Figure 3 As shown in the figure, the infrared absorption peak of aspirin in ethanol solvent is at 834 cm -1 、916cm -1 、1194cm -1 、1608cm -1 、1720cm-1 The absorption peak is particularly obvious at 1194cm -1 .

[0057] Step (6) organizes the acquired infrared spectrum data, wherein a total of 8 groups of samples in Table 2 are used as verification data and the rest are used as training data to establish a partial least squares model.

[0058] Table 2

[0059] temperature Aspirin / ethanol concentration ratio (g / g) 5℃ 0.12 10℃ 0.13 15℃ 0.15 20℃ 0.17 25℃ 0.2 30℃ 0.25 35℃ 0.31 40℃ 0.36

[0060] Step (7) selects a preprocessing method for the training data, compares smoothing (Detrend), baseline removal (Baseline), first-order derivative (1st Derivative) and other methods, establishes a partial least squares (PLS) model for the training set, and uses the training set correlation coefficient (R 2 c ), training set root mean square error (RMSEC), leave-one-out cross-validation correlation coefficient (R 2 cv ), leave-one-out cross-validation root mean square error (RMSECV), validation set correlation coefficient (R 2 p ), validation set root mean square error (RMSEP) and other indicators to evaluate the prediction ability of the model. The closer the correlation coefficient is to 1, the better it is, and the smaller the root mean square error is, the better it is. The final preprocessing method selected is a combination of first-order derivative + smoothing (Detrend).

[0061] After selecting an effective spectral preprocessing method in step (8), the optimal partial least squares model is established by selecting an effective band according to the position of the infrared absorption peak of aspirin, and the correlation coefficient is close to 1 and the smaller the root mean square error, the better. Then, 8 sets of verification data are input into the model. The optimal model result is: RMSEC = 0.0024, R 2 c =0.9997, RMSECV=0.003, R 2 cv =0.9996, RMSEP = 0.0021, R 2 p =0.999.

[0062] After the model of step (9) is established, it is saved in the concentration measurement module and applied online to the real-time monitoring of the concentration during the cooling and crystallization process of the aspirin-ethanol system.

[0063] Step (10) In-situ online monitoring of the cooling crystallization process of the aspirin-ethanol system: reconfigure the supersaturated aspirin-ethanol solution, pour 28.2g of aspirin solute into 100ml of ethanol, heat to 40°C, maintain for 20min, cool to 10°C at 0.3°C / min, and set the stirring speed to 300rpm during the whole process. Turn on the whole set of in-situ online infrared measurement system to monitor the cooling crystallization process of aspirin during the heating process in real time, and measure the concentration increase during the heating dissolution process, the appearance of crystals after cooling, and the concentration change of aspirin during the crystallization process.

[0064] Step (11) is to verify the accuracy of the online infrared model in predicting the concentration, and two groups of experiments are set up. The solution ratio is the same as step 10. In the first group of experiments, during the heating process, after the temperature is raised to 35°C, the solute is completely dissolved and stirring is stopped. After maintaining for 20 minutes, the upper clear liquid is taken with a syringe and placed in a weighing bottle with a known mass of m1 to obtain m2. After being completely dried, it is weighed again to obtain m3. The calculated value of its concentration is: c1 = (m3-m1) / (m2-m3). In the second group of experiments, during the cooling process, the temperature is continuously lowered to 20°C, crystals appear and nucleation explodes. After the temperature is further lowered to 15°C, stirring is stopped. After maintaining for 20 minutes, the upper clear liquid is taken with a syringe and placed in a weighing bottle with a known mass. The concentration is calculated as c2.

[0065] Step (12) Find the concentration value c predicted by the concentration module at the corresponding moment 10 and c 20 , through the formula (c 10 -c1) / c1、(c 20 -c2) / c2 calculated relative errors are 2.6% and -3.47%, respectively, which verifies the accuracy of the in-situ online infrared spectroscopy monitoring system in controlling concentration measurement during the cooling crystallization process.

[0066] Example 2: This implementation case in-situ online infrared spectroscopy monitoring system measures the concentration of potassium aluminum sulfate during cooling and crystallization. Figure 6 shown.

[0067] Example 3: This implementation case is an in-situ online infrared spectroscopy monitoring system, which calculates the reaction conversion rate by measuring the change in concentration during the reaction. Use this in-situ online infrared monitoring system to monitor the process of synthesizing polystyrene from styrene monomer in real time, and pre-test the conversion rate in real time through changes in styrene concentration. The reaction principle is as follows: the reaction solvent is anhydrous ethanol, 150g of anhydrous ethanol is placed in a three-necked flask, a certain amount of styrene monomer is added, and under nitrogen protection at 70°C, initiator AIBN is added after a certain period of time, and the reaction is carried out for 4-5 hours. During the reaction, the concentration change of styrene is monitored in real time using online infrared. The final product obtained by the reaction is polystyrene. The specific implementation steps are as follows:

[0068]

[0069] Step (1) Prepare samples with different concentration gradients. In this case, a total of 7 concentration gradients were prepared. The mass ratios (g / g) of styrene and ethanol were as follows: 0 g: 150 g, 5 g: 150 g, 10 g: 150 g, 20 g: 150 g, 30 g: 150 g, 40 g: 150 g, 50 g: 150 g. Infrared spectra of different concentrations were obtained. The absorption position of the characteristic peak of the C=C double bond was 1633.82 cm -1 , enlarged as Fig. 9 As shown, with the change of styrene concentration, the peak value of C=C double bond has obvious changes;

[0070] Step (2) establishing a partial least squares model of infrared spectrum and corresponding concentration according to the modeling method provided in Implementation Case 1, and importing the model into the concentration measurement model for online monitoring;

[0071] Step (3) monitors the process of synthesizing polystyrene from styrene monomer in real time, and the concentration of styrene monomer changes. At a certain moment, the concentration of styrene monomer is C0, and at the next moment, the concentration of styrene monomer is C t , the conversion rate is calculated by the following formula;

[0072]

[0073] Step (4) For this case, the C0 concentration of styrene monomer at the beginning of the reaction is C0 = 0.2552 g (styrene) / g (solution), and the concentration value C at the end of the reaction is t =0.2063 g (styrene) / g (solution), the conversion rate calculated according to the formula in step (3) is 19.16%.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. In-situ online infrared spectroscopy monitoring equipment for crystallization process control, characterized by: It includes an optical fiber probe (3), an infrared host (1), a temperature control device (6), a crystallizer (7), and a computer (2); A fiber optic probe (3), the fiber optic probe (3) comprising an incident light fiber (303), an outgoing light fiber (304), a probe tube (301) and an attenuated total reflection crystal (302), the probe tube (301) being sleeved at the front ends of the incident light fiber (303) and the outgoing light fiber (304), the attenuated total reflection crystal (302) being placed at the end of the probe tube (301) and connected to the outgoing light fiber and the incident light fiber; An infrared host (1), the infrared host (1) comprising an infrared light source, an interferometer beam splitter, a moving mirror, a fixed mirror, an MCT detector, a semiconductor electronic refrigeration device, an infrared reflector, a laser and a power supply; The infrared host (1) is connected to the optical fiber probe (3) via an optical fiber coupler SMA905 interface, the infrared host (1) is communicatively connected to the computer (2), the computer (2) is communicatively connected to the temperature control device (6), and the temperature control device (6) is used to control the temperature of the crystallizer (7).

2. The in-situ online infrared spectroscopy monitoring device for crystallization process control according to claim 1, characterized in that: The incident light optical fiber (303) and the output light optical fiber (304) are both polycrystalline glass fiber optical fibers, and the transmittable infrared wavelength range is 3-18 um.

3. The in-situ online infrared spectroscopy monitoring device for crystallization process control according to claim 1, characterized in that: The incident light optical fiber (303) and the outgoing light optical fiber (304) are covered with protective sheaths on the outside. The protective sheaths are made of spiral stainless steel flat wires. The outside of the protective sheaths is covered with a silicone rubber sheath.

4. The in-situ online infrared spectroscopy monitoring device for crystallization process control according to claim 1, characterized in that: The probe tube (301) is made of Hastelloy and has a length of 20-100 cm.

5. The in-situ online infrared spectroscopy monitoring device for crystallization process control according to claim 1, characterized in that: The attenuated total reflection crystal (302) is made of any one of diamond, silicon, germanium zinc sulfide and zinc selenide, and the reflection angle is 45 degrees. The attenuated total reflection crystal (302) is used to generate an attenuated wave on the surface of the attenuated total reflection crystal (302) for infrared light, and the wave is reflected back.

6. The in-situ online infrared spectrum monitoring device for crystallization process control according to claim 1, characterized in that: A thermometer (5) and a mechanical stirring device (4) are arranged inside the crystallizer (7).

7. The in-situ online infrared spectrum monitoring device for crystallization process control according to claim 1, characterized in that: The temperature control device (6) is equipped with a temperature sensor for transmitting a temperature signal to the computer (2) to control the temperature rate of increase, decrease and termination temperature.

8. An in-situ online infrared spectroscopy monitoring system for crystallization process control, applied to the in-situ online infrared spectroscopy monitoring device for crystallization process control according to any one of claims 1 to 7, characterized in that: It includes infrared spectrum acquisition module, concentration measurement module and model building module; The infrared spectrum acquisition module automatically deducts the background spectrum, and the spectrum acquisition speed is 1 second each time. It displays the infrared spectra of different wave numbers and different times in real time, and saves them automatically and continuously to form a three-dimensional image with the X-axis as the wave number, the Y-axis as the absorbance, and the Z-axis as the time, so as to see the changes of the infrared spectrum with the progress of the reaction; and view the saved local files in real time; The concentration measurement module predicts the concentration in real time according to the selected model, generates a reaction component concentration change curve, and displays it in a window on the computer display at the same time, and updates automatically, and has a real-time display list of concentration and sample information, and displays it in chronological order; it has a control line setting function, and issues a warning prompt when the concentration index exceeds the control standard; it has the function of distinguishing abnormal spectra, and saves the product batch number information corresponding to the spectrum; closing the concentration measurement module will automatically save all data; and print a result report; The model building module includes a chemometric algorithm, and a quantitative analysis model is established based on the chemometric algorithm. It integrates signal preprocessing functions including first-order derivatives, second-order derivatives, smoothing, and special functions for automatic selection of characteristic bands, and combines with a concentration measurement module to realize concentration prediction.

9. The in-situ online infrared spectroscopy monitoring system for crystallization process control according to claim 8, characterized in that The model building module includes partial least squares, principal element analysis, principal element regression, neural network and support vector machine linear and nonlinear modeling techniques.

10. An in-situ online infrared spectroscopy monitoring method for crystallization process control, applied to the device and system described in claims 1-9, characterized in that: The following steps are involved: S1. Build a cooling crystallization experimental device, set up a jacketed crystallizer, and equip it with a condenser to prevent volatilization. Insert the optical fiber probe into the crystallizer. The optical fiber probe is directly connected to the infrared spectrometer. The signal collected by the infrared host is transmitted to the computer and displayed in real time on the computer monitor. After selecting the calibrated crystallization model of the crystallization sample, the concentration measurement module predicts the concentration value in the crystallization process in real time, and controls the temperature through the temperature control device. Use a stirrer with adjustable stirring speed to conduct experiments to achieve the concentration change of the crystallization process of the crystallization sample; S2. Model calibration sample collection, establish a model of infrared spectrum and corresponding concentration, take concentration points at different temperatures every 5°C within the set temperature range, prepare n groups of samples with different concentrations, the concentration refers to the mass ratio of the crystallization sample and the solvent, the spectrum collection process, pour the solvent into the crystallizer, turn on the temperature control device to keep the temperature of the solvent in the crystallizer at the set value, add the crystallization sample, stir continuously, wait for the crystallization sample to be completely dissolved, collect the infrared spectrum of the specified concentration at the set temperature through the infrared spectrum collection module, collect three spectra continuously, and take the average spectrum for standby; S3, then raising the temperature to the next set temperature point, weighing the crystallization sample solute so that the concentration reaches the second set concentration value, and after the solute is completely dissolved, continuously collecting infrared spectra, and recording and saving them, taking the average spectrum for standby, repeating this step, so that all the concentration points designed in S2 are collected to obtain the corresponding infrared spectra, thereby obtaining infrared spectra of the crystallization sample at different temperatures and different concentrations; S4, sorting the acquired infrared spectrum data, randomly selecting several groups of samples as verification data, and the rest as training data, and establishing a partial least squares model; S5. Selection of preprocessing methods for training data. A partial least squares model is established for the training set by methods including comparative smoothing, baseline removal, and first-order derivative. The prediction ability of the model is evaluated using indicators including training set correlation coefficient, training set root mean square error, leave-one-out cross-validation correlation coefficient, leave-one-out cross-validation root mean square error, validation set correlation coefficient, and validation set root mean square error. The closer the correlation coefficient is to 1, the better. The smaller the validation root mean square error and cross-validation root mean square error, the better. The final preprocessing method selected is a combination of 1D+Detrend. S6. After selecting an effective spectral preprocessing method, select an effective band according to the position of the infrared absorption peak of the crystallized sample to establish the best partial least squares model. Similarly, the correlation coefficient is close to 1, and the smaller the root mean square error, the better, and the optimal model is obtained; S7. After the model is established, it is saved in the concentration measurement module and applied to the real-time monitoring of the concentration during the cooling and crystallization process of the crystallization sample; S8, concentration prediction of the cooling crystallization process, reconfigure the supersaturated crystallization sample solution, raise the temperature to a specified value, maintain for a period of time, wait for the crystallization sample to completely dissolve, and then cool it to a specified value. The whole process is stirred at a certain rate, and the concentration value increases during the heating process, and the concentration decreases after the crystals appear during the cooling process are measured in real time; S9, verify the accuracy of the online infrared model concentration measurement, set up two groups of experiments, the solution ratio is the same as S8, in the first group of experiments, after heating to the specified temperature, the solute is completely dissolved and stirring is stopped. After a period of maintenance, the upper clear liquid is taken with a syringe and put into a weighing bottle with a known mass m1 to weigh m2. After it is completely dried, it is weighed again to obtain m3. The calculated concentration is: In the second group of experiments, the cooling process continued to cool to the specified value, crystals appeared and nucleated explosively, and after cooling to the next specified value, stirring was stopped. After maintaining for a period of time, the upper clear liquid was taken with a syringe and put into a weighing bottle, the concentration c2 was calculated, and the concentration value measured by the in-situ online infrared spectroscopy monitoring system at the corresponding moment was found. The in-situ online infrared spectroscopy monitoring system was verified by calculating the relative error, and the accuracy of the concentration measurement was controlled during the cooling and crystallization process.

Images