Methane catalyst and screening method and preparation method thereof

Abstract

The present invention provides a methane catalyst and its screening method and preparation method. The screening method includes: obtaining initial data: preparing a first methane catalyst using a preset first dosage table, and performing a methane catalytic combustion test to obtain a first T 50 Data; Data iteration: According to the first usage table and the first T 50 The data was iterated to prepare the second methane catalyst and the second T was obtained through the methane catalytic combustion test. 50 Data; initial screening: for the second T 50 The data is initially screened to identify a candidate set of optimized methane catalysts; validation is performed on the candidate set; and final screening is performed on the candidate set to determine the optimal methane catalyst. This invention utilizes machine learning to screen methane catalysts, which is more time-efficient and labor-efficient than the existing cross-sectional method and can yield methane catalysts with both excellent catalytic activity and stability.

Description

Technical Field

[0001] The present application relates to the field of catalyst technology, and in particular, to a method for screening a methane catalyst, a method for preparing a methane catalyst, and the corresponding methane catalyst obtained. Background Art

[0002] The performance of a catalyst is closely related to its composition and content. In the prior art, the compatibility of catalyst components is primarily determined using a cross-fertilization method. This involves first introducing one component and then conducting extensive experiments to determine the optimal ratio under given conditions. Based on the experimental results, another component is then introduced and the experiment repeated to further optimize the ratio.

[0003] Researchers have found that for multi-component catalysts, the substrate molecules selectively adsorbed and activated by different components are also very different, and the synergistic effect between the components will change with the component ratio. Only when the catalyst's synergistic adsorption and activation ability for all substrate molecules is adapted can the catalytic activity of the catalyst reach the optimal level. Therefore, when a new component is introduced, the balance is broken, and the initial component ratio is no longer the optimal result, and a new round of experimental screening of all components is required. In this case, as the number of catalyst components increases, the total number of cumulative experiments will increase exponentially. For example, when there are two components, 10 ratios are used to screen the optimal ratio. When one component is added to become three components, the third component must be adjusted at each of the above 10 ratios. Assuming that the third component also needs to be screened at 10 ratios, then 100 experiments are required, and similarly. Four groups require 1,000 experiments, and five components require tens of thousands of experiments to obtain the optimal ratio, resulting in excessively high catalyst research and development costs and extremely low efficiency.

[0004] In addition, the catalytic effect of methane catalysts in the prior art is still not very good.

[0005] The contents of the background technology section are merely the technologies known to the inventors and do not necessarily represent the existing technologies in this field. Summary of the Invention

[0006] The first object of the present invention is to provide a method for screening a methane catalyst, which comprises:

[0007] Obtain initial data: Prepare the first methane catalyst using the preset first dosage table, and obtain the first T through the methane catalytic combustion test. 50 Data; where T 50 is the temperature corresponding to a methane conversion rate of 50%; wherein the first dosage table includes the dosages of at least two groups of metal salts required to prepare the first methane catalyst;

[0008] Data iteration: According to the first usage table and the first T 50 The data was iterated to prepare the second methane catalyst and the second T was obtained through the methane catalytic combustion test. 50 data;

[0009] Initial screening: for the second T 50 The data were preliminarily screened to determine the optimal group of methane catalysts to be determined;

[0010] Verification: performing a verification operation on the pending optimization group; and

[0011] Final screening: Perform final screening on the pending optimization group to determine the optimized methane catalyst.

[0012] In some embodiments of the present invention, the method for preparing the first methane catalyst comprises:

[0013] respectively adding Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 into the heated potassium bromide solution to obtain a first reaction solution;

[0014] adding aqueous ammonia to the first reaction solution to obtain a second reaction solution;

[0015] heating the second reaction liquid, and adding Al2O3 to the heated second reaction liquid to obtain a third reaction liquid; and

[0016] The third reaction liquid is cooled, centrifuged, washed and calcined.

[0017] In some embodiments of the present invention, the concentration of the potassium bromide solution is 4 g / L to 25 g / L, the concentration of the ammonia water is 0.1 g / L to 0.3 g / L, and the mass ratio of the potassium bromide solution, the ammonia water, Na2PdCl4, and Al2O3 is (20000 to 50000): (5000 to 20000): (0 to 15): (100 to 500);

[0018] Among them, the molar dosages of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 are set to A, B, C, D, and E, respectively, with C≥0.06mmol, A+B=0.05mmol, C+D=0.2mmol, and E / (C+D+E)<0.1; the first dosage table includes at least two sets of data of A~D values ​​and E / (C+D+E) values.

[0019] In some embodiments of the present invention, the data iteration step includes:

[0020] According to the first usage table, the first T 50Data table for data setting iteration;

[0021] The data in the iterative data table is input into a preset Bayesian optimization model to obtain a post-iteration data table, wherein the post-iteration data table includes the predicted metal salt dosage, the predicted T 50 value and first variance value;

[0022] According to the prediction T 50 The predicted metal salt dosage is selected in the order of the first variance value from small to large and the first variance value from large to small to obtain a second dosage table;

[0023] According to the second dosage table, preparing a second methane catalyst in the same manner as preparing the first methane catalyst; and

[0024] The second methane catalyst was subjected to a methane catalytic combustion test to obtain the second T 50 data.

[0025] In some embodiments of the present invention, the primary screening step comprises:

[0026] Determine the second T 50 Is there T in the data? 50 The value is less than the set value;

[0027] If yes, then set the T value including the value less than the set value. 50 The group of values ​​is the pending optimization group.

[0028] In some embodiments of the present invention, the verifying step includes:

[0029] According to the amount of metal salt and T 50 Data table for value setting verification;

[0030] The data in the verification data table is input into the Bayesian optimization model to obtain a verification data table, wherein the verification data table includes verification of metal salt dosage, verification of T 50 value and second variance value;

[0031] Verify T as described 50 The corresponding verification metal salt dosage is selected in the order of the values ​​from small to large and the second variance values ​​from large to small to obtain a third dosage table;

[0032] According to the third usage table, preparing a third methane catalyst in the same manner as preparing the first methane catalyst; and

[0033] The third methane catalyst was subjected to a methane catalytic combustion test to obtain a third T 50 data.

[0034] In some embodiments of the present invention, the final screening step includes:

[0035] Determine the third T 50 Is there T in the data? 50 The value is less than the set value,

[0036] If not, the undetermined optimization group is set as the optimization group, wherein the methane catalyst corresponding to the optimization group is the optimized methane catalyst.

[0037] In some embodiments of the present invention, if the second T 50 No T in the data 50 The value is less than the set value, the screening method further includes:

[0038] Updating the iterative data table, and performing the data iteration step and the primary screening step again according to the updated iterative data table until the undetermined optimization group is screened out;

[0039] The updated iterative usage table includes the first usage table, the first T 50 Data, the predicted metal salt dosage obtained in the remaining data iteration steps except the last data iteration step and the corresponding second T 50 data.

[0040] In some embodiments of the present invention, if the third T 50 There is T in the data 50 The value is less than the set value, the screening method further includes:

[0041] This will include the T that is less than the set value 50 The group with the best value is added to the undetermined optimization group, and the verification step and the final screening step are performed again until the optimized methane catalyst is screened out.

[0042] A second object of the present invention is to provide a method for preparing a methane catalyst, which comprises:

[0043] respectively adding Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 into the heated potassium bromide solution to obtain a first reaction solution;

[0044] adding aqueous ammonia to the first reaction solution to obtain a second reaction solution;

[0045] heating the second reaction liquid, and adding Al2O3 to the heated second reaction liquid to obtain a third reaction liquid; and

[0046] Cooling, centrifuging, washing and calcining the third reaction liquid to obtain the catalyst;

[0047] The concentration of the potassium bromide solution is 4 g / L to 25 g / L, the concentration of the ammonia water is 0.1 g / L to 0.3 g / L, and the mass ratio of the potassium bromide solution, the ammonia water, Na2PdCl4, and Al2O3 is (20000 to 50000): (5000 to 20000): (0 to 15): (100 to 500);

[0048] The molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4 and Y(NO3)3 are set to A, B, C, D and E respectively, C≥0.06mmol, A+B=0.05mmol, C+D=0.2mmol, E / (C+D+E)<0.1.

[0049] In some embodiments of the present invention, the preparation method further comprises:

[0050] dissolving potassium bromide in deionized water to obtain the potassium bromide solution; and

[0051] The potassium bromide solution is heated at 60° C. to 100° C. for 10 min to 30 min.

[0052] In some embodiments of the present invention, heating the second reaction liquid comprises:

[0053] The second reaction solution is heated at 60° C. to 100° C. for 1 to 2 hours.

[0054] The third object of the present invention is to provide a methane catalyst obtained by any of the above screening methods; or

[0055] The method is obtained by any of the above-mentioned preparation methods.

[0056] In some embodiments of the present invention, 0.03567 mmol≤A≤0.04567 mmol, 0.00443 mmol≤B≤0.01433 mmol, 0.1381 mmol≤C≤0.1581 mmol, 0.0419 mmol≤D≤0.0619 mmol, and 0.00413≤E / (C+D+E)≤0.02413.

[0057] In some embodiments of the present invention, 0.0416 mmol≤A≤0.0516 mmol, 0≤B≤0.0084 mmol, 0.18881 mmol≤C≤0.20881 mmol, 0.00019 mmol≤D≤0.00219 mmol, and 0.05845≤E / (C+D+E)≤0.07845.

[0058] The present invention uses machine learning to screen methane catalysts, which saves time and effort compared to the cross-cutting method in the prior art, and can obtain methane catalysts with excellent catalytic activity and stability.

[0059] The preparation method provided by the present invention can obtain a methane catalyst with excellent catalytic activity and stability under specific raw materials and dosage.

[0060] The methane catalyst provided by the present invention is obtained by the above screening method or preparation method and has very excellent performance.

[0061] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The accompanying drawings, which constitute a part of the present disclosure, are used to provide a further understanding of the present disclosure. The exemplary embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation to the present disclosure.

[0063] Figure 1 A method for screening a methane catalyst provided by one embodiment of the present invention is shown.

[0064] Figure 2 The invention shows a method for preparing a methane catalyst according to an embodiment of the invention.

[0065] Figure 3 A data distribution diagram of the Bayesian optimization model output provided by an embodiment of the present invention is shown.

[0066] FIG4( a ) shows a transmission electron microscope image of a methane catalyst provided in one embodiment of the present invention.

[0067] FIG4(b) shows the element distribution diagram of the methane catalyst shown in FIG4(a).

[0068] FIG4( c ) shows the XRD spectrum of the methane catalyst shown in FIG4( a ).

[0069] Figure 5 A transmission electron microscope image of a methane catalyst provided by another embodiment of the present invention is shown. DETAILED DESCRIPTION

[0070] Hereinafter, only certain exemplary embodiments are briefly described. As will be appreciated by those skilled in the art, the described embodiments may be modified in various ways without departing from the spirit or scope of the present invention. Therefore, the drawings and description are to be considered as illustrative in nature and not restrictive.

[0071] The disclosure below provides many different embodiments or examples for implementing the present invention. In order to simplify the disclosure of the present invention, the components and settings of specific examples are described below. Of course, they are merely examples and are not intended to limit the present invention. In addition, the present invention may repeat reference numbers and / or reference letters in different examples. Such repetition is for the purpose of simplicity and clarity and does not in itself indicate the relationship between the various embodiments and / or settings discussed. In addition, the present invention provides examples of various specific processes and materials, but those of ordinary skill in the art will recognize the application of other processes and / or the use of other materials.

[0072] In addition, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs. It will also be understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and the present invention, and will not be interpreted in an idealized or overly formal sense unless expressly defined as such in this article.

[0073] As used herein, "about" or "approximately" is inclusive of the stated value and means within an acceptable range of deviation from the particular value as determined by one skilled in the art, taking into account the measurement in question and the errors associated with the measurement of the particular quantity (i.e., the limitations of the measurement system). For example, "about" can mean within one or more standard deviations, or within ±30%, ±20%, ±10%, or ±5% of the stated value.

[0074] The following description of the embodiments of the present invention is provided in more detail with reference to the accompanying drawings and examples to provide a better understanding of the present invention and its advantages in various aspects. However, the embodiments and examples described below are for illustrative purposes only and are not intended to limit the present invention.

[0075] Figure 1 A method for screening a methane catalyst provided by an embodiment of the present invention is shown, which includes an initial data acquisition step, a data iteration step, a primary screening step, a verification step, and a final screening step.

[0076] The step of obtaining the initial data includes: preparing the first methane catalyst using the preset first usage table, and obtaining the first T through the methane catalytic combustion test. 50 Data. The first dosage table includes the dosage of at least two groups of metal salts required to prepare the first methane catalyst. 50 The temperature corresponding to the methane conversion rate of 50% is shown in Table 1. The lower the temperature, the better the effect of the corresponding methane catalyst.

[0077] The present invention selects the temperature corresponding to the methane conversion rate of 50% (T 50 ) was used as the target indicator, and the subsequent Bayesian optimization screening was carried out based on the amount of metal salt.

[0078] Figure 2 The preparation method of the methane catalyst provided by one embodiment of the present invention is shown, which includes the following steps S1 to S4. The above-mentioned step of obtaining the initial data can be adopted Figure 2 The preparation method shown is used to prepare the first methane catalyst.

[0079] S1: Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 are respectively added to a heated potassium bromide solution to obtain a first reaction solution.

[0080] Alternatively, potassium bromide is dissolved in deionized water to obtain a potassium bromide solution, and the potassium bromide solution is then heated at 60° C. to 100° C. for 10 min to 30 min. The metal salt is then added.

[0081] Optionally, the concentration of the potassium bromide solution is 4 g / L to 25 g / L.

[0082] In the present invention, the molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4 and Y(NO3)3 are set to A, B, C, D and E respectively, C≥0.06mmol, A+B=0.05mmol, C+D=0.2mmol, and E / (C+D+E)<0.1.

[0083] To ensure spontaneous redox reactions and the formation of a spherical core-shell structure, the amount of Ce(NO3)3 is set at more than twice the amount of precious metals. By setting the molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 within the above ranges, a methane catalyst with relatively good performance can be obtained.

[0084] S2: adding aqueous ammonia to the first reaction solution to obtain a second reaction solution.

[0085] In this step, the metal salt and potassium bromide undergo spontaneous redox reaction under alkaline conditions to generate precious metal PtPd alloy as the core and small particles of CeZrYO x Core-shell structured nanospheres with solid solution tightly wrapped around the shell.

[0086] In this step, the pH value of the second reaction solution must be greater than 10 to ensure an effective reaction.

[0087] Optionally, the concentration of the ammonia water used is 0.1 g / L to 0.3 g / L.

[0088] S3: heating the second reaction liquid, and adding Al2O3 to the heated second reaction liquid to obtain a third reaction liquid.

[0089] The core-shell nanospheres obtained in this step can be evenly dispersed on the surface of the Al2O3 carrier.

[0090] Optionally, heating the second reaction liquid includes: heating the second reaction liquid at 60° C. to 100° C. for 1 hour to 2 hours.

[0091] S4: cooling, centrifuging, washing and calcining the third reaction liquid.

[0092] In this step, the third reaction liquid is cooled, centrifuged, washed and calcined to obtain the desired methane catalyst.

[0093] The third reaction liquid can be cooled to room temperature and then centrifuged. The product after centrifugation can be repeatedly washed with water and ethanol. Calcination can be performed under air at a temperature of about 840°C to about 850°C for about 2.5 to about 3.5 hours, preferably at about 850°C for about 3 hours.

[0094] Wherein, the mass ratio of potassium bromide solution, ammonia water, Na2PdCl4 and Al2O3 in the above steps can be (20000-50000):(5000-20000):(0-15):(100-500).

[0095] The preparation method provided by the present invention can obtain a methane catalyst with excellent catalytic activity and stability under specific raw materials and dosage.

[0096] When the above steps of obtaining initial data are adopted Figure 2 When preparing the first methane catalyst in the preparation method shown, the first usage table in the initial data acquisition step includes at least two sets of data: A to D values ​​and E / (C+D+E) values. In this case, the initial data acquisition step may specifically include:

[0097] The molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 are set to A, B, C, D, and E, respectively, where C ≥ 0.06 mmol, A+B = 0.05 mmol, C+D = 0.2 mmol, and E / (C+D+E) < 0.1;

[0098] Determining a first usage table, wherein the first usage table includes at least two sets of data of A-D values ​​and E / (C+D+E) values;

[0099] According to the A to D values ​​and E / (C+D+E) values ​​in the first dosage table, Figure 2The preparation method shown is used to prepare a methane catalyst to obtain a first methane catalyst; and

[0100] The first methane catalyst was tested for methane catalytic combustion and the first T 50 data.

[0101] In the present invention, the data iteration step includes: according to the first usage table and the first T 50 The data was iterated to prepare the second methane catalyst and the second T was obtained through the methane catalytic combustion test. 50 data.

[0102] In some embodiments of the present invention, the data iteration step includes:

[0103] According to the first dosage table, the first T 50 Data table for data setting iteration;

[0104] The data in the iterative data table are input into the Bayesian optimization model to obtain the iterative data table, wherein the iterative data table includes the predicted metal salt dosage, the predicted T 50 value and first variance value;

[0105] According to the prediction T 50 Select the corresponding predicted metal salt dosage in the order of small to large values ​​and large to small first variance values ​​to obtain a second dosage table;

[0106] According to the second dosage table, preparing a second methane catalyst in the same manner as preparing the first methane catalyst; and

[0107] The second methane catalyst was tested for methane catalytic combustion to obtain the second T 50 data.

[0108] When using Figure 2 When preparing the first methane catalyst according to the preparation method shown in the figure, the iterative data table includes the A to D values ​​and E / (C+D+E) values ​​in the first dosage table and the first T 50 The data, the predicted amount of metal salt is the predicted A ~ D value and E / (C + D + E) value, the second methane catalyst also uses Figure 2 Prepared according to the preparation method shown.

[0109] The data iteration step in the present invention can be performed once or twice or more. When two or more data iterations are performed, the iteration data table used in each iteration needs to be updated. The updated iteration usage table includes the first usage table, the first T 50 Data, the predicted metal salt dosage obtained in the remaining data iteration steps except the last data iteration step and the corresponding second T 50That is, when the data iteration step is performed once, the iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first usage table and the first T 50 When the data iteration step is performed twice, the first iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first usage table and the first T 50 The second iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first usage table, the first T 50 Data, the predicted A-D value and E / (C+D+E) value of the iterative data table obtained after the first data iteration, the second T of the second catalyst obtained after the first data iteration 50 When the data iteration step is performed three times, the first iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first usage table and the first T 50 The second iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first usage table, the first T 50 Data, the predicted A-D value and E / (C+D+E) value of the iterative data table obtained after the first data iteration, the second T of the second catalyst obtained after the first data iteration 50 The third iteration data table includes the A to D values ​​and E / (C+D+E) values ​​in the first dosage table, the first T 50 Data, the predicted A-D value and E / (C+D+E) value of the iterative data table obtained after the first data iteration, the second T of the second catalyst obtained after the first data iteration 50 data, the predicted A-D values ​​and E / (C+D+E) values ​​of the iterative data table obtained after the second data iteration, and the second T of the second catalyst obtained after the second data iteration. 50 When the data iteration step is performed four times or more, the same process is repeated and no further explanation is given.

[0110] The Bayesian optimization model involved in the present invention mainly includes two components: a probabilistic proxy model of the black box function f(X) and an acquisition function.

[0111] The probabilistic proxy model of the black box function f(X): X*=arg min f(x), X∈Ω, where X* represents the amount of raw metal salt used, that is, the amount of catalyst component, Ω represents the design space of interest, and f(X) represents the sampling state point, which is equivalent to T 50 The probabilistic proxy model of the black box function f(X) is used to provide predictions of unsampled state points f(X) and their uncertainties. Mathematically speaking, its goal is to find the global minimum of the unknown target function f(X), that is, T 50 Minimum value.

[0112] Acquisition function: arg max α(X) = arg min f(X), where α(X) is the acquisition function. The selection criteria for the next query design point are determined by combining the prediction and its uncertainty. Therefore, minimizing f(X) is replaced by maximizing the acquisition function α(X).

[0113] By maximizing the acquisition function α(X), the target catalyst composition X can be selected at locations where uncertainty is large (called exploration) and model prediction is low (called utilization).

[0114] The Bayesian optimization model adjusts the catalyst component X in an iterative manner to optimize f(X). At the beginning, it is necessary to give certain catalyst components and their target performance values ​​as initial data. Unlike conventional machine learning methods that require a large amount of data, the Bayesian optimization model has very low requirements for initial data. It can provide a small amount of random or even individual data, or it can provide a specified portion of data. In each iteration, the Gaussian process regression model (GPR) is used to analyze the known data and predict the target performance of any catalyst component. Due to the uniqueness of Gaussian process regression, while obtaining the predicted value m(X), the uncertainty corresponding to the predicted value can also be quantified by the variance s(X). Afterwards, the lower confidence bound algorithm (LCB) is used as the acquisition function. The acquisition function GP-LCB can be defined to comprehensively consider the predicted value and uncertainty, hoping that the predicted value is better while the uncertainty is smaller, and finally maximize the acquisition function to determine the catalyst component that needs to be experimented in this iteration. The acquisition function GP-LCB, the full name of which is the confidence lower limit based on Gaussian process regression, has the form of a function: α GP-LCB = -m(X)+w×s(X), which is a commonly used acquisition function in Bayesian optimization. It uses the given weight coefficient w as the weight to achieve a trade-off between the predicted value and uncertainty.

[0115] If the acquisition function prioritizes improving predictions, it's called "exploitation," meaning the catalyst component selection relies more on the component with the best target performance in the known data. This means the selected component is close to the optimal one. If the acquisition function prioritizes minimizing uncertainty, it's called "exploration," meaning the catalyst component selection relies more on exploring the unknown component space. This means the selected component is far from the optimal one.

[0116] In some embodiments of the present invention, a dynamic acquisition function is used, and instead of giving a weight coefficient w as in conventional models, this coefficient is also used as an optimization variable, and is changed in each iteration according to the data situation. In this way, when the amount of data is small, for example, when the data table after iteration only includes 4 groups of data, each iteration can use the "exploration" scheme to obtain 3 groups of data to quickly discover the optimal target performance of the entire space, and the "utilization" scheme to obtain 1 group of data to converge to the optimal target performance near the known data. As the iteration proceeds, the overall uncertainty decreases, the prediction will be more accurate, and at the same time, the area with better target performance in the component space can be obtained. At this time, the "utilization" scheme can be used more frequently to ultimately determine the component with the best target performance.

[0117] The primary screening step of the present invention includes: 50 The data were preliminarily screened to identify a candidate optimization group of methane catalysts.

[0118] In some embodiments of the present invention, the initial screening step includes:

[0119] Determine the second T 50 Is there T in the data? 50 The value is less than the set value;

[0120] If yes, set the T value including the one that is less than the set value. 50 The group of values ​​is the pending optimization group.

[0121] Alternatively, for a methane catalyst, the set point may be set to, for example, 335°C.

[0122] Optionally, if the second T 50 No T in the data 50 The value is less than the set value, the screening method also includes:

[0123] Update the iterative data table, and perform the data iteration step and the initial screening step again according to the updated iterative data table until the pending optimization group is screened out;

[0124] The updated iterative usage table includes the first usage table, the first T 50 Data, the predicted metal salt dosage obtained in the remaining data iteration steps except the last data iteration step and the corresponding second T 50 data.

[0125] The process of updating the iterative data table can be the same as the process of updating the iterative data table when performing two or more data iteration steps as described above, and will not be repeated here.

[0126] In the present invention, the verification step includes: performing a verification operation on the pending optimization group.

[0127] In some embodiments of the present invention, the verifying step includes:

[0128] According to the metal salt dosage and T 50 Data table for value setting verification;

[0129] The data in the verification data table are input into the Bayesian optimization model to obtain the verification data table, wherein the verification data table includes the verification metal salt dosage, verification T 50 value and second variance value;

[0130] According to the verification T 50 The values ​​are arranged from small to large and the second variance values ​​are arranged from large to small, and combined with the "utilization" and "exploration" rules mentioned above, the corresponding verification metal salt dosage is selected to obtain the third dosage table;

[0131] According to the third usage table, a third methane catalyst is prepared in the same manner as the first methane catalyst; and

[0132] The third methane catalyst was tested for methane catalytic combustion and the third T 50 data.

[0133] When using Figure 2 When the first methane catalyst and the second methane catalyst are prepared by the preparation method shown, the third methane catalyst is also prepared by Figure 2 At this time, the verification data table includes the A to D values ​​and E / (C+D+E) values ​​in the first dosage table and the first T 50 Data, A to D values ​​and E / (C+D+E) values ​​in the second dosage table and the second T 50 The metal salt dosage is verified by the predicted A-D values ​​and E / (C+D+E) values ​​output by the Bayesian optimization model in this step.

[0134] In the present invention, the final screening step includes: performing final screening on the to-be-optimized group to determine the optimized methane catalyst.

[0135] In some embodiments of the present invention, the final screening step includes:

[0136] Judging the Third T 50 Is there T in the data? 50 The value is less than the set value,

[0137] If not, the pending optimization group is set as the optimization group, and the methane catalyst corresponding to the optimization group is set as the optimized methane catalyst.

[0138] Optionally, if the third T 50 There is T in the data 50 The value is less than the set value, the screening method further includes: including the T less than the set value 50The group with the highest value is added to the pending optimization group, and the verification step and final screening step are carried out again until the optimized catalyst is screened out.

[0139] In some embodiments of the present invention, by limiting and adopting the proportions of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 components, a Bayesian optimization model is used to iteratively search for methane catalysts (PtPd@CeZrYO x , referred to as PPCZY) T 50 The initial data and historical data collected by the optimization iteration will be used for the next prediction. In each iteration, the proxy model is updated, which is updated as the historical data changes, and then T can be obtained from the proxy model. 50 The prediction and its uncertainty in the entire variable space. The model can then use the acquisition function to combine the prediction and its uncertainty to select the target catalyst composition. Finally, the real T 50 Data, and update the data table for iteration.

[0140] The present invention adopts the Bayesian optimization algorithm in machine learning to screen methane catalysts, which saves time and labor compared with the cross-cutting method in the prior art and can obtain methane catalysts with excellent catalytic activity and stability.

[0141] The present invention further provides Figure 1 The screening method shown or Figure 2 The methane catalyst is obtained by the preparation method shown.

[0142] In some embodiments of the present invention, when 0.03567mmol≤A≤0.04567mmol, 0.00443mmol≤B≤0.01433mmol, 0.1381mmol≤C≤0.1581mmol, 0.0419mmol≤D≤0.0619mmol, 0.00413≤E / (C+D+E)≤0.02413; or when 0.0416mmol≤A≤0.0516mmol, 0≤B≤0.0084mmol, 0.18881mmol≤C≤0.20881mmol, 0.00019mmol≤D≤0.00219mmol, 0.05845≤E / (C+D+E)≤0.07845, the T of the obtained methane catalyst is 50 The value is less than 335°C, which is much higher than the performance of the methane catalyst obtained in the prior art.

[0143] The present invention will be described below with reference to specific examples. The numerical values ​​of the process conditions in the following examples and comparative examples are exemplary, and their possible numerical ranges are as shown in the above summary of the invention. For process parameters not otherwise specified, conventional techniques can be used. Unless otherwise specified, the reagents and instruments used in the technical solutions provided by the present invention can be purchased from conventional channels or on the market. The measurement methods of the present invention are all measurement methods well known in the industry.

[0144] Preparation Example 1-1

[0145] This preparation example prepares a methane catalyst. First, dissolve 500 mg of potassium bromide in 50 mL of deionized water, heat to 60°C, and hold for 30 minutes. Then, add Na₂PdCl₄, K₂PtCl₄, Ce(NO₃)₃, Zr(NO₃)₄, and Y(NO₃)₃ according to the A–D values ​​and E / (C+D+E) values ​​in Table 1. Then, add 8 mL of dilute ammonia (prepared by adding 0.3 mL of 25% ammonia to 20 mL of deionized water). The reaction mixture is heated at 60°C for 2 hours, and 0.1 g of Al₂O₃ is added as a support. Stir for 30 minutes. After cooling to room temperature, the mixture is centrifuged, washed several times with water and ethanol, and calcined in air at 850°C for 3 hours.

[0146] The prepared methane catalyst was subjected to a methane catalytic combustion test to obtain T 50 For details, see Table 1.

[0147] Preparation Examples 1-2 to 1-5

[0148] Preparation Examples 1-2 to 1-5 differ from Preparation Example 1-1 only in the A to D values ​​and the E / (C+D+E) value.

[0149] The prepared methane catalyst was subjected to a methane catalytic combustion test to obtain T 50 For details, see Table 1.

[0150] Table 1

[0151] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 Preparation Example 1-1 0.0475 0.0025 0.16 0.04 0.01 378 Group 2 Preparation Example 1-2 0.0475 0.0025 0.16 0.04 0.02 382 Group 3 Preparation Examples 1-3 0.0475 0.0025 0.16 0.04 0.03 372 Group 4 Preparation Examples 1-4 0.0475 0.0025 0.16 0.04 0.04 375 Group 5 Preparation Examples 1-5 0.0475 0.0025 0.16 0.04 0.05 369

[0152] Example

[0153] This example screens the optimized methane catalyst and sets the target T 50 The value is 335℃. The specific steps are as follows:

[0154] First iteration:

[0155] Input the data in Table 1 into the Bayesian optimization model and perform the first iteration to obtain the data table after the first iteration, where the data table after the first iteration includes the A to D values ​​and E / (C+D+E) values ​​predicted for the first time, the T 50value and the first variance value;

[0156] According to the first prediction T 50 The corresponding A~D values ​​and E / (C+D+E) values ​​of the first prediction are selected in the order of small to large values ​​and the first first variance values ​​from large to small. According to the "utilization" and "exploration" rules mentioned above, the first and second usage table is determined, see Table 2.

[0157] According to the A to D values ​​and the E / (C+D+E) value in Table 2, a second methane catalyst was prepared in the same manner as in Preparation Example 1-1.

[0158] The second methane catalyst prepared was subjected to a methane catalytic combustion test to obtain the actual T 50 The data are shown in Table 2. Table 2 includes 4 groups of A~D values ​​and E / (C+D+E) values, as well as the corresponding actual T 50 data.

[0159] Second iteration:

[0160] The data in Table 1 and Table 2 are input into the Bayesian optimization model and the second iteration is performed to obtain the data table after the second iteration, wherein the data table after the second iteration includes the second predicted A to D values ​​and E / (C+D+E) value, the second predicted T 50 value and the second first variance value;

[0161] According to the second prediction T 50 The corresponding A~D values ​​and E / (C+D+E) values ​​of the second prediction are selected in the order of small to large values ​​and the second first variance values ​​from large to small. According to the "utilization" and "exploration" rules mentioned above, the second second usage table is determined, see Table 3.

[0162] According to the A to D values ​​and the E / (C+D+E) value in Table 3, a second methane catalyst was prepared in the same manner as in Preparation Example 1-1.

[0163] The second methane catalyst prepared was subjected to a methane catalytic combustion test to obtain the actual T 50 The data are shown in Table 3. Table 3 includes 4 groups of A~D values ​​and E / (C+D+E) values, as well as the corresponding actual T 50 data.

[0164] Third iteration:

[0165] The data in Tables 1 to 3 are input into the Bayesian optimization model for the third iteration to obtain the data table after the third iteration, wherein the data table after the third iteration includes the A to D values ​​and E / (C+D+E) values ​​predicted for the third time, the T 50 value and the third first variance value;

[0166] According to the third prediction T 50 The corresponding A~D values ​​and E / (C+D+E) values ​​of the third prediction are selected in the order of small to large values ​​and the third first variance values ​​from large to small. According to the "utilization" and "exploration" rules mentioned above, the third second usage table is determined, see Table 4.

[0167] According to the A to D values ​​and the E / (C+D+E) value in Table 4, a second methane catalyst was prepared in the same manner as in Preparation Example 1-1.

[0168] The second methane catalyst prepared was subjected to a methane catalytic combustion test to obtain the actual T 50 The data are shown in Table 4. Table 4 includes 4 groups of A~D values ​​and E / (C+D+E) values, as well as the corresponding actual T 50 data.

[0169] Initial screening:

[0170] From Tables 2 to 4, we can see that the actual T 50 The value was less than 335°C, so a verification step was performed.

[0171] First verification:

[0172] The data in Tables 1 to 4 were input into the Bayesian optimization model to obtain the first validation data table, which included the first validation A to D values ​​and E / (C+D+E) values, the first validation T 50 Value and first and second variance values;

[0173] According to the first verification T 50 The corresponding first verification A~D values ​​and E / (C+D+E) values ​​are selected in the order of small to large values ​​and the first second variance values ​​from large to small. According to the "utilization" and "exploration" rules mentioned above, the third usage table is determined, see Table 5.

[0174] According to the A to D values ​​and the E / (C+D+E) value in Table 5, a third methane catalyst was prepared in the same manner as in Preparation Example 1-1.

[0175] The third methane catalyst prepared was subjected to a methane catalytic combustion test to obtain the actual T 50 The data are shown in Table 5. Table 5 includes 4 groups of A~D values ​​and E / (C+D+E) values, as well as the corresponding actual T 50 data.

[0176] First final screening:

[0177] From Table 5, we can see that the actual T 50 The value is less than 335°C, so this group is also set as a pending optimization group.

[0178] Second verification:

[0179] The data in Tables 1 to 5 were input into the Bayesian optimization model to obtain the second validation data table, wherein the second validation data table includes the second validation A to D values ​​and E / (C+D+E) values, the second validation T 50 value and the second variance value;

[0180] According to the second verification T 50 The corresponding second verification A~D values ​​and E / (C+D+E) values ​​are selected in the order of small to large values ​​and the second second variance values ​​from large to small. According to the "utilization" and "exploration" rules mentioned above, the third usage table is determined, see Table 6.

[0181] According to the A to D values ​​and the E / (C+D+E) value in Table 6, a third methane catalyst was prepared in the same manner as in Preparation Example 1-1.

[0182] The third methane catalyst prepared was subjected to a methane catalytic combustion test to obtain the actual T 50 The data are shown in Table 6. Table 6 includes 4 groups of A~D values ​​and E / (C+D+E) values, as well as the corresponding actual T 50 data.

[0183] Second final screening:

[0184] As can be seen from Table 6, there is no actual T 50 The values ​​are less than 335°C, so the first group in Table 3 and the third group in Table 4 are both set as optimized groups.

[0185] in, Figure 3 This is the three-dimensional data distribution diagram of Tables 1 to 6. Figure 4(a) to Figure 4(c) They are the transmission electron microscope image, element distribution map and XRD spectrum of Group 3 in Table 4 respectively. Figure 5 This is the transmission electron microscopy image of Group 1 in Table 3.

[0186] Table 2:

[0187] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 0.04736 0.00264 0.12989 0.07011 0.08536 368 Group 2 0.04128 0.00872 0.05281 0.14719 0.08833 529 Group 3 0.05 0 0.15212 0.04788 0.01134 420 Group 4 0.01724 0.03276 0.16032 0.03968 0.09681 582

[0188] Table 3:

[0189] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 0.04789 0.00211 0.19322 0.00678 0.06798 338 Group 2 0.03842 0.01158 0.19896 0.00104 0.0635 354 Group 3 0.01322 0.03678 0.07119 0.12881 0.07992 511 Group 4 0.04025 0.00975 0.14867 0.05133 0.04133 387

[0190] Table 4:

[0191] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 0.04067 0.00933 0.14810 0.05190 0.01413 330 Group 2 0.04829 0.00171 0.11014 0.08986 0.07025 420 Group 3 0.04152 0.000848 0.14644 0.05356 0.01394 347 Group 4 0.03067 0.01933 0.12104 0.07896 0.05587 420

[0192] Table 5:

[0193] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 0.02812 0.02188 0.13091 0.06909 0.03283 470 Group 2 0.04955 0.00045 0.13492 0.06508 0.04229 404 Group 3 0.04660 0.00340 0.19881 0.00119 0.06845 329 Group 4 0.04358 0.00642 0.09164 0.10836 0.08367 371

[0194] Table 6:

[0195] A (mmol) B (mmol) C (mmol) D(mmol) E / (C+D+E) <![CDATA[T 50 (℃)]]> Group 1 0.04943 0.00057 0.1884 0.0116 0.058945 340 Group 2 0.0392 0.0108 0.2 0 0.09771 346 Group 3 0.04917 0.00083 0.16219 0.03781 0.02586 385 Group 4 0.04601 0.00399 0.17726 0.02274 0.05857 377

[0196] Comparative Example

[0197] Using the cross-cutting method, based on Pd@CeO2, while maintaining the same Ce(NO3)3 dosage, the dosages of the precious metal salts of Pd and Pt were first adjusted. After selecting the optimal Pt and Pd dosages, the corresponding metal salt dosages were fixed, Zr(NO3)4 was introduced, and the optimal ratio of Ce(NO3)3 and Zr(NO3)4 was adjusted and screened. After determining the ratio of these four substances, the Y(NO3)3 dosage was adjusted. Ultimately, the catalytic performance of the optimal sample screened by the traditional cross-cutting method was obtained.

[0198] The process of preparing methane catalyst is as follows:

[0199] First, dissolve 300mg of potassium bromide in 30mL of deionized water and heat to 60°C for 30 minutes. Then, add Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3. Then, add 5mL of dilute ammonia (prepared by adding 0.3mL of 25% ammonia to 20mL of deionized water). Heat the reaction mixture at 60°C for 2 hours, add 0.1g of Al2O3 as a support, and stir for 30 minutes. After cooling to room temperature, centrifuge, wash several times with water and ethanol, and calcine in air at 850°C for 3 hours.

[0200] The prepared methane catalyst was subjected to a methane catalytic combustion test to obtain T 50 The specific dosage and corresponding T of the optimized methane catalyst screened out 50 See Table 7 for data.

[0201] Table 7

[0202]

[0203]

[0204] From the above preparation examples, embodiments and comparative examples, it can be seen that the use of machine learning can shorten the number of experiments required for tens of thousands of times to only 25 experiments to obtain the optimal result. 50 The catalyst compositions below 335°C are located in different regions, rather than having similar performance due to similar components. This is almost impossible to discover using the traditional cross-sectional method, which generally only finds an optimal composition ratio.

[0205] Obviously, the above embodiments are merely examples for the purpose of clearly illustrating the present invention and are not intended to limit the embodiments. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all embodiments here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for screening a methane catalyst, characterized in that: include: Obtain initial data: Prepare the first methane catalyst using the preset first dosage table, and obtain the first T through the methane catalytic combustion test. 50 Data; where T 50 is the temperature corresponding to a methane conversion rate of 50%; wherein the first dosage table includes the dosage of at least two groups of metal salts required to prepare the first methane catalyst; Data iteration: According to the first usage table and the first T 50 The data was iterated to prepare the second methane catalyst and the second T was obtained through the methane catalytic combustion test. 50 data; Initial screening: for the second T 50 The data were preliminarily screened to determine the optimal group of methane catalysts to be determined; Verification: performing verification operations on the pending optimization group; as well as Final screening: performing final screening on the undetermined optimization group to determine the optimized methane catalyst; Wherein, the method for preparing the first methane catalyst comprises: respectively adding Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 into the heated potassium bromide solution to obtain a first reaction solution; adding aqueous ammonia to the first reaction solution to obtain a second reaction solution; heating the second reaction liquid, and adding Al2O3 to the heated second reaction liquid to obtain a third reaction liquid; and Cooling, centrifuging, washing and calcining the third reaction liquid; The concentration of the potassium bromide solution is 4 g / L to 25 g / L, the concentration of the ammonia water is 0.1 g / L to 0.3 g / L, and the mass ratio of the potassium bromide solution, the ammonia water, Na2PdCl4, and Al2O3 is (20000 to 50000): (5000 to 20000): (0 to 15): (100 to 500); Among them, the molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 are set to A, B, C, D, and E, respectively, C ≥ 0.06 mmol, A+B = 0.05 mmol, C+D = 0.2 mmol, and E / (C+D+E) < 0.1; the first dosage table includes at least two sets of data of A~D values ​​and E / (C+D+E) values.

2. The screening method according to claim 1, wherein The data iteration step includes: According to the first usage table, the first T 50 Data table for data setting iteration; The data in the iterative data table is input into a preset Bayesian optimization model to obtain a post-iteration data table, wherein the post-iteration data table includes the predicted metal salt dosage, the predicted T 50 value and first variance value; According to the prediction T 50 The predicted metal salt dosage is selected in the order of the first variance value from small to large and the first variance value from large to small to obtain a second dosage table; According to the second dosage table, preparing a second methane catalyst in the same manner as preparing the first methane catalyst; and The second methane catalyst was subjected to a methane catalytic combustion test to obtain the second T 50 data; The primary screening step comprises: Determine the second T 50 Is there T in the data? 50 The value is less than the set value; If yes, then set the T value including the value less than the set value. 50 The group of values ​​is the pending optimization group; the verification step includes: According to the amount of metal salt and T 50 Data table for value setting verification; The data in the verification data table is input into the Bayesian optimization model to obtain a verification data table, wherein the verification data table includes verification of metal salt dosage, verification of T 50 value and second variance value; Verify T as described 50 The corresponding verification metal salt dosage is selected in the order of the values ​​from small to large and the second variance values ​​from large to small to obtain a third dosage table; According to the third usage table, preparing a third methane catalyst in the same manner as preparing the first methane catalyst; and The third methane catalyst was subjected to a methane catalytic combustion test to obtain a third T 50 data; The final screening step comprises: Determine the third T 50 Is there T in the data? 50 The value is less than the set value, If not, the undetermined optimization group is set as the optimization group, wherein the methane catalyst corresponding to the optimization group is the optimized methane catalyst.

3. The screening method according to claim 2, characterized in that If the second T 50 No T in the data 50 The value is less than the set value, the screening method further includes: Updating the iterative data table, and performing the data iteration step and the primary screening step again according to the updated iterative data table until the undetermined optimization group is screened out; The updated iterative data table includes the first usage table, the first T 50 Data, the predicted metal salt dosage obtained in the remaining data iteration steps except the last data iteration step and the corresponding second T 50 data.

4. The screening method according to claim 2, wherein If the third T 50 There is T in the data 50 The value is less than the set value, the screening method further includes: This will include the T that is less than the set value 50 The group with the best value is added to the undetermined optimization group, and the verification step and the final screening step are performed again until the optimized methane catalyst is screened out.

5. A method for preparing a methane catalyst, characterized in that: include: respectively adding Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4, and Y(NO3)3 into the heated potassium bromide solution to obtain a first reaction solution; adding aqueous ammonia to the first reaction solution to obtain a second reaction solution; heating the second reaction liquid, and adding Al2O3 to the heated second reaction liquid to obtain a third reaction liquid; and Cooling, centrifuging, washing and calcining the third reaction liquid to obtain the catalyst; The concentration of the potassium bromide solution is 4 g / L to 25 g / L, the concentration of the ammonia water is 0.1 g / L to 0.3 g / L, and the mass ratio of the potassium bromide solution, the ammonia water, Na2PdCl4, and Al2O3 is (20000 to 50000): (5000 to 20000): (0 to 15): (100 to 500); The molar amounts of Na2PdCl4, K2PtCl4, Ce(NO3)3, Zr(NO3)4 and Y(NO3)3 are set to A, B, C, D and E respectively, C≥0.06mmol, A+B=0.05mmol, C+D=0.2 mmol, and E / (C+D+E)<0.

1.

6. The preparation method according to claim 5, characterized in that The preparation method further comprises: dissolving potassium bromide in deionized water to obtain the potassium bromide solution; and Heat the potassium bromide solution at 60° C. to 100° C. for 10 min to 30 min; Heating the second reaction liquid comprises: The second reaction solution is heated at 60° C. to 100° C. for 1 h to 2 h.

7. A methane catalyst, characterized in that Obtained by the screening method according to any one of claims 1 to 4; or The method is obtained by the preparation method according to claim 5 or 6.

8. The methane catalyst according to claim 7, characterized in that 0.03567mmol≤A≤0.04567mmol, 0.00443mmol≤B≤0.01433mmol, 0.1381mmol≤C≤0.1581mmol, 0.0419mmol≤D≤0.0619mmol, 0.00413≤E / (C+D+E)≤0.02413; or 0.0416mmol≤A≤0.0516mmol, 0≤B≤0.0084mmol, 0.18881mmol≤C≤0.20881mmol, 0.00019mmol≤D≤0.00219mmol, 0.05845≤E / (C+D+E)≤0.07845.

Images