Automatic crystal structure refinement analysis method and system based on random one-time refinement

Abstract

The invention discloses an automatic crystal structure refinement analysis method and system based on random one-time refinement, and belongs to the technical field of material structure characterization and X-ray diffraction. According to the method, Python script automatic execution and random one-time refinement strategies are combined, the working efficiency is greatly improved, and time consumption and personal errors caused by manual operation are effectively reduced; according to the method, random variation is introduced to serve as a fine trimming parameter, multi-round one-time fine trimming is conducted, the parameter space can be explored more comprehensively, and therefore a better crystal structure model is found; a data acquisition module, an initial fine trimming module, a random change fine trimming module, a result selection and evaluation module and the like are designed in the system, so that the whole-process automation of crystal structure analysis is realized, and the accuracy and efficiency of analysis are improved; the method is suitable for various crystal materials, automatic refinement can be carried out only through diffraction patterns, instrument parameters and structure parameters of the crystal materials, and accurate and efficient data support is provided for the scientific research and industrial fields.

Description

Technical Field

[0001] The present invention belongs to the technical field of material structure characterization and X-ray diffraction, and specifically relates to an automated crystal structure refinement analysis method and system based on random one-time refinement. Background Art

[0002] Crystal structure analysis is an important part of materials science research. The traditional Rietveld refinement method is highly dependent on manual operation, which is not only time-consuming but also requires a high level of experience from researchers. Even when using existing refinement software such as GSASII and FullProf, a lot of manual intervention is still required when dealing with complex crystal structures, which greatly reduces the efficiency of analysis. In addition, the traditional method requires researchers to repeatedly check and adjust parameters to ensure the accuracy of the refinement results, which further increases the workload and time cost. A one-time refinement method based on random initial values ​​has been proposed in some studies, but the initial data of this method is based on the results of manual refinement, which limits its automation and general applicability. Existing studies have not formed an automatic refinement method based on multiple strategies, and cannot fully utilize the advantages of different strategies to improve the efficiency and accuracy of refinement. There is insufficient discussion on the performance of this method under different initial value errors, which makes it difficult for researchers to evaluate its reliability and stability in practical applications.

[0003] To solve the problems of much manual intervention, low efficiency and human bias in the crystal structure analysis process, patent CN114972185A discloses a full spectrum line fitting method of polycrystalline powder X-ray diffraction spectrum based on statistical modeling. This method quantitatively analyzes the diffraction signal in reciprocal space through fast Fourier transform and Savitzky-Golay filtering, thereby obtaining the background distribution function including Compton scattering, fluorescence, multi-order diffraction and other factors, and realizes the functions of "statistical background calculation", "accurate determination of component lattice constant", "quantitative determination of component volume fraction", etc., but the degree of automation is low; CN114705704A discloses a method for measuring austenite content in steel, using the Rietveld method for steel samples with preferred orientation, and using TOPAS refinement software to refine the structure of the X-ray diffraction spectrum, the purpose is to improve the accuracy of the test results when calculating the austenite content in steel. However, this method still requires manual intervention, and there is a problem of more manual intervention.

[0004] In summary, the existing refinement methods in the field of crystal structure analysis still need to be continuously improved and optimized to achieve high automation, intelligence and efficiency. Summary of the invention

[0005] In view of the fact that the traditional Rietveld refinement method is highly dependent on manual operation, time-consuming and requires high experience of researchers. Although the existing improvement scheme introduces automated tools, it still requires a lot of manual intervention when processing complex structures, which is inefficient and lacks the technical status of combining multiple strategies and exploring performance stability, the present invention aims to provide an automated crystal structure refinement analysis system and method based on random one-time refinement.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The present invention provides a method for automated crystal structure refinement analysis based on random one-time refinement, comprising:

[0008] S1, obtaining diffraction data, structural parameters and instrument parameters of the crystal material to be analyzed;

[0009] S2, designing different refinement strategies based on the structural parameters and instrument parameters of the crystal material to be analyzed, performing initial refinement on the diffraction data of the crystal material to be analyzed based on the different refinement strategies, obtaining initial refinement results including refinement fitting factors, and selecting an optimal result 1 from the initial refinement results, wherein the optimal result 1 is a refinement result with the smallest initial refinement fitting factor and a correct crystal structure;

[0010] S3, introducing random changes to the atomic fraction coordinates and other parameters in the optimal result 1, performing a second refinement on the diffraction data in the optimal result 1, obtaining a second refinement result including a second refinement fitting factor, and selecting the optimal result 2 from the second refinement result; the optimal result 2 is a refinement result with the smallest second refinement fitting factor and a correct crystal structure;

[0011] S4, compare the best result 1 and the best result 2, select the refinement result with the smallest refinement fitting factor as the final result, and output the crystal structure information.

[0012] In S1, the structural parameters include space group, unit cell parameters, atomic fractional coordinates, atomic types and symmetry operations, which are saved in the structural model file; the instrument parameters include the wavelength of X-rays and the target material type, which are saved in the instrument parameter file.

[0013] S2 specifically includes:

[0014] S21, importing the diffraction data, crystal structure data and instrument parameters of the crystal material to be analyzed into the X-ray powder diffraction refinement engine software;

[0015] S22, designing different refinement strategies based on crystal structure data and instrument parameter data, importing X-ray powder diffraction refinement engine software, performing initial refinement on the diffraction data of the crystal material to be analyzed based on different refinement strategies, and obtaining initial refinement results including initial refinement fitting factors;

[0016] S23, calculating the initial refinement fitting factor, checking the refined structural parameters, and selecting the result with the smallest initial refinement fitting factor and correct crystal structure as the optimal result 1; the initial refinement result includes the initial refinement fitting factor, the initial refined structural parameters, the initial refined diffraction pattern and the visual representation of the initial refined structural parameters.

[0017] The X-ray powder diffraction refinement engine software is any one of GSAS-II, FULLPROF and TOPAS.

[0018] The different refinement strategies are designed based on the structural parameters and the instrument parameters and are obtained by adjusting the order of the structural parameters and the instrument parameters.

[0019] The second refinement specifically includes:

[0020] A random change is introduced into the atomic fraction coordinates in the optimal result 1 to obtain the atomic fraction coordinates after the random change is introduced, and a random change percentage is introduced into other parameters in the optimal result 1 to obtain the values ​​of other parameters after the random change percentage is introduced; based on the atomic fraction coordinates after the random change is introduced and the values ​​of other parameters after the random change percentage is introduced, the diffraction data in the optimal result 1 is refined for the second time, and the k is refined. i times, obtaining a second refinement result including a refinement fitting factor; the other parameters include background factor, unit cell parameter, micro strain, grain size, temperature factor and instrument parameter;

[0021] The atomic fraction coordinate after introducing random changes = optimal result + atomic fraction coordinate + random changes, where random changes = random.uniform(-Δx i ,Δx i );

[0022] The other parameter values ​​after introducing the random change percentage = optimal result + other parameter values ​​× (1 + random change percentage); where random change percentage = random.uniform (-Δp i ,Δp i );

[0023] Where i represents the number of random changes, Δx represents the random changes introduced by the atomic fraction coordinates, and Δx i represents the i-th random change introduced by the atomic fraction coordinates, Δp represents the percentage of random changes introduced by other parameters; Δpi represents the i-th random variation percentage introduced by other parameters; the k i represents the number of refinements to introduce the i-th random variation, Δx i <0.02,Δp i <2%, k i The value is 1000 to 10000, and i is 1 to 8.

[0024] The refined fitting factor is the peak shape fitting factor R p , weight fitting factor R wp , goodness of fit factor GOF and expected fit factor R exp Any one of .

[0025] Furthermore, the peak shape fitting factor R p The calculation formula is: Among them, Y o,i is the ith experimental observation, Y c,i is the i-th calculated fit value, and n is the number of data points.

[0026] Furthermore, the weight fitting factor R wp The calculation formula is: Among them, w i Represents the weight of each data point, w i =1 / Y o,i ; Y o,i is the experimental observation value (the count value corresponding to the i-th 2θ), Y c,i is the calculated fitted value, and n is the number of data points.

[0027] Furthermore, the expected fitting factor R exp The calculation formula is as follows: Where N is the total number of data points and P is the number of fitting parameters.

[0028] Furthermore, the calculation formula of the goodness of fitness factor GOF (Good of fitness) is as follows:

[0029] Furthermore, the chi-square value χ 2 It is often used to evaluate the degree of fit between the model and experimental data. The calculation formula is as follows: Among them, Y o,i is the ith experimental observation, Y c,i is the i-th calculated fit value, and n is the number of data points.

[0030] The present invention provides a system for realizing the above-mentioned automated crystal structure refinement analysis method based on random one-time refinement, comprising:

[0031] A data acquisition module, used to acquire diffraction data, structural model data and instrument parameter data of the crystal material to be analyzed;

[0032] The initial refinement strategy design and execution module performs initial refinement on the diffraction data of the crystal material to be analyzed based on different refinement strategies to obtain initial refinement results, and selects the initial refinement result with the smallest initial refinement fitting factor as the optimal result 1;

[0033] A random variation and second refinement module is used to introduce random variation to perform a second refinement on the optimal result 1, obtain a second refinement result including a second refinement fitting factor, and select the refinement result with the smallest second refinement fitting factor as the optimal result 2;

[0034] The result selection and evaluation module is used to evaluate the first refinement result and the second refinement result, compare the best result 1 and the best result 2, obtain the refinement result with the lowest refinement fitting factor and the correct crystal structure as the final result, and output the crystal structure information.

[0035] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned automated crystal structure refinement analysis method based on random one-time refinement when executing the computer program.

[0036] The present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned automated crystal structure refinement analysis method based on random one-time refinement are implemented.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The automated crystal structure refinement analysis method based on random one-time refinement provided by the present invention can not only improve work efficiency and analysis accuracy, but also ensure the reliability and scalability of the results, and has significant technical advantages and application prospects. Through the automated refinement process, the time and human deviation of manual trial and error are significantly reduced. The combination of Python scripts allows the refinement process to be automatically executed, greatly improving work efficiency; by introducing random changes as initial parameters for multiple rounds of one-time refinement, more parameter spaces can be explored to find a better crystal structure model; by comparing the fitting factors and crystal structure correctness under different refinement strategies, the advantages and disadvantages of different models can be evaluated more objectively, avoiding the deviation caused by subjective judgment.

[0039] Furthermore, a variety of different refinement strategies can be designed for different data qualities, material properties and instruments. The use of multiple strategies enables this method to adapt to a variety of complex situations, improve the accuracy and reliability of refinement results, reduce human intervention, and achieve highly automated crystal structure analysis; by introducing random changes in different ranges to the atomic fraction coordinates and other parameters in the crystal structure and performing multiple one-time refinements, this method can more fully explore the parameter space, find better refinement results, and improve the accuracy and reliability of the analysis.

[0040] The system for automatic crystal structure refinement and analysis based on random one-time refinement provided by the present invention includes a data acquisition module, an initial refinement module, a random variation refinement module, and a result selection and evaluation module. The method for automatic crystal structure refinement and analysis is realized through modular design, the efficiency and accuracy of the analysis are improved, and it is easy to promote and use in practical applications. It can be applied to a variety of different crystal materials and has wide applicability and expansibility. Whether it is an inorganic material, an organic material or a composite material, as long as its diffraction pattern and structural model data can be obtained, the present invention can be used for refinement and analysis, so that the present invention has a wide range of application prospects and can provide reliable data support for scientific research and industrial fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Flow chart of the automated crystal structure refinement analysis method based on random one-time refinement of the present invention, wherein A is a flow chart and B is a detailed flow chart;

[0042] Figure 2 A random value of Δx=0.0001 is introduced into the atomic fraction coordinates of the PbSO4 structural parameters in Example 1 of the present invention, and a random variation of Δp=0.01% is introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is R wp Figure,(b) is a local view;

[0043] Figure 3 A random value of Δx=0.001 is introduced into the atomic fraction coordinates of the PbSO4 structural parameters in Example 1 of the present invention, and a random variation of Δp=0.1% is introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure, where (a) is R wp Figure; (b) is a local diagram;

[0044] Figure 4 A random value of Δx=0.020 is introduced into the atomic fraction coordinates of the PbSO4 structural parameters in Example 1 of the present invention, and a random variation of Δp=0.2% is introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wpFigure, where (a) is R wp Figure; (b) is a local diagram;

[0045] Figure 5 A random value of Δx=0.050 is introduced into the atomic fraction coordinates of the PbSO4 structural parameters in Example 1 of the present invention, and a random variation of Δp=0.5% is introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is R wp Figure,(b) is a local view;

[0046] Figure 6 A random value of Δx=0.090 was introduced into the atomic fraction coordinates of the PbSO4 structural parameters in Example 1 of the present invention, and a random variation of Δp=0.9% was introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is R wp Figure,(b) is a local view;

[0047] Figure 7 The measured value, fitted value and difference diagram of the random one-time refinement result of PbSO4 in Example 1 of the present invention;

[0048] Figure 8 This is a structural diagram of the refined PbSO4 result of Example 1 of the present invention, drawn using VESTA;

[0049] Fig. 9 A random value of Δx=0.0001 was introduced into the atomic fraction coordinates of the Ca5F(PO4)3 structural parameters in Example 2 of the present invention, and a random variation of Δp=0.01% was introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is the number of runs and the R obtained wp Figure,(b) is a local view;

[0050] Fig.10 A random value of Δx=0.001 was introduced into the atomic fraction coordinates of the Ca5F(PO4)3 structural parameters in Example 2 of the present invention, and a random variation of Δp=0.1% was introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is the number of runs and the R obtained wp Figure,(b) is a local view;

[0051] Fig.11 A random value of Δx=0.02 is introduced into the atomic fraction coordinates of the Ca5F(PO4)3 structural parameters in Example 2 of the present invention, and a random variation of Δp=0.2% is introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wpFigure; (a) is the number of runs and the R obtained wp Figure,(b) is a local view;

[0052] Fig.12 A random value of Δx=0.005 was introduced into the atomic fraction coordinates of the Ca5F(PO4)3 structural parameters in Example 2 of the present invention, and a random variation of Δp=0.5% was introduced into other parameters as initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is the number of runs and the R obtained wp Figure,(b) is a local view;

[0053] Fig.13 A random value of Δx=0.009 was introduced into the atomic fraction coordinates of the Ca5F(PO4)3 structural parameters in Example 2 of the present invention, and a random variation of Δp=0.9% was introduced into other parameters as the initial parameters for the number of runs and R of one-time refinement. wp Figure; (a) is the number of runs and the R obtained wp Figure,(b) is a local view;

[0054] Fig.14 The measured values, fitted values, and difference values ​​of the random one-time refinement results of Ca5F(PO4)3 in Example 2 of the present invention;

[0055] Fig.15 This is a structural diagram of the refined Ca5F(PO4)3 result of Example 2 of the present invention, drawn using VESTA;

[0056] Fig.16 In Example 3 of the present invention, three background function coefficients are used. Based on the results of strategy 1 refinement, a random value of Δx=0.002 is introduced into the atomic fraction coordinates of the PbSO4 crystal structure, and a random change of Δp=0.2% is introduced into other parameters as the initial parameters for one-time refinement. The number of runs and R wp Figure; (b) is a local picture.

[0057] Fig.17 In Example 3 of the present invention, 7 background function coefficients are used. Based on the results of strategy 4 refinement, a random value of Δx=0.002 is introduced into the atomic fraction coordinates of the PbSO4 crystal structure, and a random change of Δp=0.2% is introduced into other parameters as the initial parameters for the one-time refinement operation times and R wp Figure; (b) is a local picture.

[0058] Fig.18In Example 3 of the present invention, 7 background function coefficients are used. Based on the results of strategy 5 refinement, a random value of Δx=0.005 is introduced into the atomic fraction coordinates of the PbSO4 crystal structure, and a random variation of Δp=0.5% is introduced into other parameters as the initial parameters for the one-time refinement operation times and R wp Figure; (b) is a local picture. DETAILED DESCRIPTION

[0059] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme 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 only 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 should fall within the scope of protection of the present invention.

[0060] Optimal result 1: During the crystal structure refinement process, by comparing the initial refinement results obtained under different refinement strategies, the one with the smallest refinement fit factor (such as R wp ) and the results meet the requirements of crystal structure correctness;

[0061] Optimal result 2: introduce random changes to the atomic fraction coordinates and other parameters in the optimal result 1, perform a second refinement on the diffraction data in the optimal result 1, and select the result with the smallest refinement fitting factor and correct crystal structure from all the second refinement results as the optimal result 2;

[0062] Final result: Compare the fitting factors of the best result 1 and the best result 2, and select the result with the smaller fitting factor as the final result;

[0063] Example 1

[0064] See attached Figure 1 This embodiment takes the crystal material PbSO4 as an example to provide an automated crystal structure refinement analysis method based on random one-time refinement.

[0065] S1, obtaining powder X-ray diffraction data of the crystalline material PbSO4 to be analyzed, and obtaining the powder X-ray diffraction pattern, structural parameters and instrument parameters of the crystalline material PbSO4 to be analyzed through the obtained powder X-ray diffraction data;

[0066] S11, select a structural model, and the structural parameters are saved in the structural model file. For details, see Table 1. Table 1 lists the space group, lattice parameters (a, b, c, α, β, γ) and atomic and atomic fraction coordinates of the PbSO4 structural parameters in the structural model.

[0067] Table 1: PbSO4 structural parameters and atomic fraction coordinates

[0068]

[0069] S12, obtain instrument parameters, which are saved in the instrument parameter file. The instrument parameters include the wavelength of the X-ray. The target material type is Cu.

[0070] S2, use Python to call the comprehensive structure analysis system GSAS-II; set the background function type to "chebyschev-1", the number of background function coefficients to 7, and the number of refinement cycles to 12; load the obtained structural parameters, instrument parameters, and diffraction data, and the others are GSASII default values.

[0071] S21, importing the diffraction data, structural parameters and instrument parameters of PbSO4 into X-ray powder diffraction refinement engine software;

[0072] S22, design different refinement strategies based on structural parameters and instrument parameters, import X-ray powder diffraction refinement engine software, and perform initial refinement on the diffraction data of the crystal material to be analyzed based on different refinement strategies;

[0073] S23, calculating the initial refinement fitting factor, checking the refined structural parameters, comparing the refinement results obtained by different refinement strategies, and selecting the result with the smallest refinement fitting factor and the correct crystal structure as the optimal result 1; the initial refinement result includes the initial refinement fitting factor, the initial refined structural parameters, the initial refined diffraction pattern and the visual representation of the initial refined structural parameters.

[0074] The powder X-ray diffraction pattern was fitted and refined based on the crystal structure information. Six different refinement strategies were used to obtain the weighted fitting factors R of the refinement results of different refinement strategies. wp , record the results of the initial refinement of each strategy, the 6 strategies used are shown in Table 2, and the initial refinement results of the PbSO4 strategy are shown in Table 3.

[0075] Table 2: Six refinement strategies used

[0076]

[0077] Table 3: PbSO4 strategy refinement results (the background function has 7 coefficients)

[0078]

[0079]

[0080] From the data in Table 3, we can see that the Rwp The smallest factor value is strategy 1, R wp The value is 9.73955.

[0081] The strategy 1 result (R wp =9.73955), recorded as the optimal result 1; if the optimal result of the preliminary refinement does not meet the requirements of the crystal structure correctness, continue to check the results of other strategies until a result that meets the requirements is found; find the optimal result of the preliminary refinement results of the strategy, recorded as the optimal result 1; the structural parameters of the optimal result 1 and the corresponding R wp The factors are shown in Table 4.

[0082] Table 4: Structural parameters of the initial refinement results of PbSO4 (the number of background function coefficients is 7)

[0083]

[0084] The best result is determined by: Among the refined results of several strategies, the one with the smallest fitting factor (R wp The result is that the minimum value is the smallest and the crystal structure is correct.

[0085] The correctness of the crystal structure includes but is not limited to: Refinement result spectrum: Check the degree of agreement between the refined result spectrum and the observed spectrum at each peak. The higher the agreement, the more reliable the refinement result. Crystal structure information: Check the rationality of crystal structure parameters such as bond length, bond angle, atomic fraction coordinates, whether they conform to the symmetry of the crystal, and whether the values ​​of other refinement parameters (such as temperature factor, occupancy, etc.) are within a reasonable range. Refinement fit factor: Check R wp , R p , R exp , GOF and χ 2 For equal fitting factor values, select the smallest fitting factor.

[0086] S3, using the refinement parameters of the optimal result 1 as the starting value, introducing random changes in different ranges as refinement parameters for multiple rounds of one-time refinement, and recording the results of each refinement. Based on the optimal result 1, a random value of Δx = 0.0001 was introduced to the atomic fraction coordinates in the structural parameters, and a random change of Δp = 0.01% was introduced to other parameters as the initial refinement parameters, and a one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved, and repeated 5000 times; the red dotted line in the figure is the R of the optimal result 1 obtained by strategic refinement of the PbSO4 spectrum wp The refined fitting factor R of the 2986th result of this one-time refinement wp =9.73944 is the lowest, and the crystal structure is correct. Other fitting factors and other results are shown in Table 5.

[0087] Table 5: One-time refinement results of PbSO4 with different random variation ranges (7 background function coefficients)

[0088]

[0089] Based on the optimal result 1 obtained from the initial refinement of the PbSO4 spectrum, a random value of Δx = 0.001 was introduced into the atomic fraction coordinates in the structural parameters, and a random change of Δp = 0.1% was introduced into other parameters as the initial refinement parameters. A one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved. The process was repeated 5000 times. The weight fitting factors and the number of runs of the refinement results are shown in Figure 3 As shown in (a), Figure 3 (b) is a local image. The refined fitting factor R of the 4186th result of this one-time refinement wp =9.73948 is the lowest, and the crystal structure is correct. The results of other fitting factors are shown in Table 5.

[0090] Based on the optimal result 1 obtained from the initial refinement of the PbSO4 spectrum, a random value of Δx = 0.002 was introduced into the atomic fraction coordinates in the structural parameters, and a random change of Δp = 0.2% was introduced into other parameters as the initial refinement parameters. A one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved. The process was repeated 5000 times. The weight fitting factors and the number of runs of the refinement results are shown in Figure 4 As shown in (a), Figure 4 (b) is a local image. The refined fitting factor R of the 2168th result of this one-time refinement wp =9.73945 is the lowest, and the crystal structure is correct. The results of other fitting factors are shown in Table 5. Based on the optimal result 1 obtained by the initial refinement of the PbSO4 spectrum, a random value of Δx=0.005 is introduced to the atomic fraction coordinates in the structural parameters, and a random change of Δp=0.5% is introduced to other parameters as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved, and repeated 5000 times; the weight fitting factors and running times of the refinement results are shown in Figure 5 As shown, Figure 5 (b) is a local image; the refined fitting factor R of the 4729th result of this one-time refinement wp=9.73955 is the lowest, and the crystal structure is correct; other fitting factors and other results are shown in Table 5. Based on the optimal result 1 obtained by the initial refinement of the PbSO4 spectrum, a random value of Δx=0.009 is introduced into the atomic fraction coordinates in the structural parameters, and a random change of Δp=0.9% is introduced into other parameters as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved, and repeated 5000 times; the weight fitting factors and running times of the refinement results are shown in Figure 6 As shown, Figure 6 (b) is a local image; the refined fitting factor R of the 30th result of this one-time refinement wp =9.73957 is the lowest, and the crystal structure is correct; other fitting factors and other results are shown in Table 5. Based on the optimal result 1 obtained by the initial refinement of the PbSO4 spectrum, random changes of (Δx=0.01, Δp=1.0%), (Δx=0.015, Δp=1.5%) and (Δx=0.015, Δp=1.5%) were introduced into the atomic fraction coordinates and other refinement parameters in the structural parameters as the initial refinement parameters, and a one-time refinement was performed. The refinement results (including refinement parameters and refinement fitting factors) were recorded and saved, and repeated 5000 times; the weight fitting factors and other results of the refinement results are shown in Table 5.

[0091] Among the results of one refinement with different random variation ranges, the fitting factor is the smallest (R wp The result with the smallest crystal structure and the correct crystal structure is recorded as the best result 2. After inspection, a random value of Δx = 0.0001 was introduced to the atomic fraction coordinates in the crystal structure, and a random change of Δp = 0.01% was introduced to other parameters as the initial parameters for a one-time refinement of the 1692nd result. The crystal structure is correct and the refined fitting factor R wp =9.73947 lowest.

[0092] S4, compare the best result 1 and the best result 2, and select the best fit factor (R wp The result with the lowest ) and more reasonable refinement parameters is taken as the final refinement result. The crystal structure parameters of the best result 1 and the best result 2 are shown in Table 4.

[0093] The best way to find the result: refine the fitting factor (R wp ) is the lowest, and the results of the refined parameters are reasonable or more reasonable; according to the minimum R wp The crystal structure is the most likely structure based on the given experimental data. The best result 2 is selected as the final result. The refinement result map is shown in Figure 7 .

[0094] By the attached Figure 7The data show that the refined fitting values ​​by the method of the present invention are in good agreement with the observed values, which indicates that the model of the crystal structure parameters is relatively accurate. wp =9.73947, GOF = 1.97513, the fitting effect is good. The structure diagram of PbSO4 obtained by refinement is as follows Figure 8 As shown, it can be seen that the bond length and bond angle are reasonable, the crystal structure model is accurate, and a high-precision PbSO4 crystal structure model is obtained by introducing random changes for refinement.

[0095] Example 2

[0096] This embodiment takes the crystal material Ca5F(PO4)3 as an example to provide an automated crystal structure refinement analysis method based on random one-time refinement.

[0097] S1, obtaining powder X-ray diffraction data of the crystalline material Ca5F(PO4)3 to be analyzed, structural parameters of the crystalline material Ca5F(PO4)3 to be analyzed, and instrument parameters, and obtaining a powder X-ray diffraction pattern of the crystalline material Ca5F(PO4)3 to be analyzed through the obtained Ca5F(PO4)3 powder X-ray diffraction data;

[0098] S11, select a structural model, and the structural parameters are saved in the structural model file, and the structural parameters include space group, unit cell parameters, atomic species, atomic fraction coordinates, symmetry operations, etc. Table 6 lists the space group, lattice parameters (a, b, c, α, β, γ) and atomic and atomic fraction coordinates of the Ca5F(PO4)3 structural model in the structural model.

[0099] Table 6: Structural parameters and atomic fraction coordinates of the Ca5F(PO4)3 structural model

[0100]

[0101] S12, obtain instrument parameters, which are saved in the instrument parameter file. The instrument parameters include the wavelength of the X-ray. The target material type is Cu.

[0102] S2, use Python to call the comprehensive structure analysis system GSAS-II for refinement. Set the background function type to "log interpolate", the number of background function coefficients to 10, and the number of refinement cycles to 12; load the obtained structural parameters, instrument parameters, and diffraction data. The rest are GSASII default values. Based on the crystal structure information, the powder X-ray diffraction pattern is fitted and refined. Six different refinement strategies are used to obtain the weighted fitting factor R of the refinement results of different refinement strategies. wp, record the results of each strategy refinement; the initial refinement results of the Ca5F(PO4)3 spectrum refined using the 6 strategies in Table 1 are shown in Table 7.

[0103] Table 7: Ca5F(PO4)3 strategy refinement results

[0104]

[0105] From the data in Table 7, we can see that the R wp The smallest factor value is strategy 3, R wp The value is 9.36697. The strategy 3 result (R wp =9.36697) as the basis. If the crystal structure is incorrect or the refinement fitting factor is large, consider whether the structural model of the crystal material is correct and select the correct crystal structure model. Find the optimal result of the strategy refinement result and record it as the optimal result 1. The structural parameters of the optimal result 1 are shown in Table 8.

[0106] Table 8: Results of main parameters of Ca5F(PO4)3 refinement

[0107]

[0108] The best result is determined by: Among the refined results of several strategies, the one with the smallest fitting factor (R wp The crystal structure is correct. The crystal structure is correct, including but not limited to: Refinement result spectrum: Check the consistency between the refined result spectrum and the observed spectrum at each peak. Crystal structure information: Check the rationality of crystal structure parameters such as bond length, bond angle, atomic fraction coordinates, and other refined parameters within a reasonable range. Refinement fit factor: Check R wp , R p , R exp The values ​​of fitting factors such as and GOF are selected, and the smallest fitting factor is selected.

[0109] S3, taking the refinement parameters of the optimal result 1 as the starting values, introducing random changes in different ranges as refinement parameters to perform multiple rounds of one-time refinement (second refinement), and recording each refinement result.

[0110] Based on the optimal result 1 obtained from the initial refinement of the Ca5F(PO4)3 spectrum, a random value of Δx = 0.0001 was introduced to the atomic fraction coordinates in the crystal structure, and a random change of Δp = 0.01% was introduced to other parameters as the initial refinement parameters. A one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved. The process was repeated 5000 times. The weight fitting factors and the number of runs of the refinement results are shown in Fig. 9 As shown, attached Fig. 9The refined fitting factor R of the 3744th result of this one-time refinement is wp =9.36568 is the lowest, and the crystal structure is correct; other fitting factors and other results are shown in Table 9.

[0111] Table 9: Ca5F(PO4)3 primary refinement results with different random variation ranges (10 background function coefficients)

[0112]

[0113] Based on the optimal result 1 obtained from the initial refinement of the Ca5F(PO4)3 spectrum, a random value of Δx=0.001 was introduced into the atomic fraction coordinates in the crystal structure parameters, and a random change of Δp=0.1% was introduced into other parameters as the initial refinement parameters. A one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved. The process was repeated 5000 times. The weight fitting factors and the number of runs of the refinement results are shown in Fig.10 As shown, the refined fitting factor R of the 3738th result of this one-time refinement is wp =9.36288 is the lowest, and the crystal structure is correct. The results of other fitting factors are shown in Table 9. Based on the optimal result 1 obtained by the initial refinement of the Ca5F(PO4)3 spectrum, a random value of Δx=0.002 is introduced to the atomic fraction coordinates in the crystal structure parameters, and a random change of Δp=0.2% is introduced to other parameters as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved, and repeated 5000 times. The weighted fitting factors and running times of the refinement results are shown in Fig.11 As shown, the refined fitting factor R of the 3252nd result of this one-time refinement is wp =9.36556 is the lowest, and the crystal structure is correct; other fitting factors and other results are shown in Table 9. Based on the optimal result 1 obtained by the initial refinement of the Ca5F(PO4)3 spectrum, a random value of Δx=0.005 is introduced into the atomic fraction coordinates in the crystal structure parameters, and a random change of Δp=0.5% is introduced into other parameters as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved, and repeated 5000 times; the weight fitting factors and running times of the refinement results are shown in Table 9. Fig.12 As shown, the refined fitting factor R of the 4667th result of this one-time refinement is wp=9.36595 is the lowest, and the crystal structure is correct; other fitting factors and other results are shown in Table 9. Based on the optimal result 1 obtained by the initial refinement of the Ca5F(PO4)3 spectrum, a random value of Δx=0.009 is introduced into the atomic fraction coordinates in the crystal structure parameters, and a random change of Δp=0.9% is introduced into other parameters as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved. The process is repeated 5000 times. The weighted fitting factors and the number of runs of the refinement results are shown in Table 9. Fig.13 As shown, the refined fitting factor R of the 3317th result of this one-time refinement is wp =9.36663 is the lowest, and the crystal structure is correct. The results of other fitting factors are shown in Table 9. Based on the optimal result 1 obtained by the initial refinement of the Ca5F(PO4)3 spectrum, random changes of Δp are introduced into the atomic fraction coordinates Δx and other refinement parameters in the crystal structure parameters, and the corresponding values ​​are (Δx=0.01, Δp=1.0%), (Δx=0.015, Δp=1.5%) and (Δx=0.015, Δp=1.5%), which are used as the initial refinement parameters. A one-time refinement is performed, and the refinement results (including refinement parameters and refinement fitting factors) are recorded and saved, and repeated 5000 times; the weight fitting factors of the refinement results are shown in Table 9.

[0114] Find the best result of the random refinement results and record it as the best result 2. The best result 2 is found by: Among the results of one refinement with several different random variation ranges, the one with the smallest fitting factor (R wp The result of one-time refinement with random value changes of Δx=0.001 and Δp=0.1% as initial parameters was checked. The crystal structure was correct and the refined fitting factor R wp =9.36288 lowest.

[0115] S4, compare the best result 1 and the best result 2, select the refined result with the lowest refinement fitting factor and the correct crystal structure as the final result, and output the crystal structure information. wp The crystal structure is the most likely structure based on the given experimental data. This result is selected as the final result. The refinement result map is shown in Fig.14 .Depend on Fig.14 It can be seen that the refined fitting value by the method of the present invention is consistent with the observed value, which indicates that the model of the crystal structure is accurate. wp The value is 9.36288, and GOF = 1.67452, which has a good fitting effect. The structure diagram of Ca5F(PO4)3 obtained by refinement is shown in Fig.15 As shown in the figure. From the data and structure diagram, it can be seen that the bond lengths and bond angles obtained are reasonable, and the crystal structure model is reasonable. By introducing random changes for refinement, a highly accurate crystal structure model is obtained.

[0116] Example 3

[0117] Based on Example 1, this example takes the crystal material PbSO4 as an example and adopts coefficients of three background functions to illustrate the effect of randomly changing a refinement under different refinement qualities.

[0118] Use Python to call the comprehensive structural analysis system GSAS-II. Set the background function type to "chebyschev-1", the number of background function coefficients to 3, and the number of refinement cycles to 12; load the obtained structural parameters, instrument parameters, and diffraction data; the rest are GSASII default values. Fit and refine the powder X-ray diffraction pattern based on the crystal structure information, and use 6 different refinement strategies to obtain the weighted fitting factor R of the refinement results of different refinement strategies. wp , record the initial refinement results of each strategy; the initial refinement results of the 6 strategies used are shown in Table 10, and the refinement results of the PbSO4 strategy when the number of background function coefficients is 3 are shown in Table 10.

[0119] Table 10: Initial refinement results of PbSO4 with different strategies (the coefficients of the background function are 3)

[0120]

[0121] From the data in Table 10, we can see that the R wp The smallest factor value is strategy 1, R wp The value is 10.46553. The optimal result of the strategy refinement is found and recorded as optimal result 1. The structural parameters of optimal result 1 are shown in Table 11.

[0122] Table 11: Structural parameters of the initial refinement results of PbSO4 using different strategies (the number of background function coefficients is 3)

[0123]

[0124]

[0125] Taking the refinement parameters of the best result 1 as the starting value, introduce different ranges of random changes as the initial refinement parameters for multiple rounds of one-time refinement, and record the results of each refinement. Based on the best result 1 obtained by the initial refinement of the PbSO4 spectrum, introduce a random value of Δx to the atomic fraction coordinates in the crystal structure parameters, and introduce a random change of Δp to other parameters as the initial refinement parameters, perform a one-time refinement, record and save the refinement results (including refinement parameters and refinement fitting factors), and repeat 5000 times; the range of random changes includes (Δx = 0.0001, Δp = 0.01%), (Δ Table 12 shows the weight fitting factors of the refined results (Δx=0.001, Δp=0.1%), (Δx=0.002, Δp=0.2%), (Δx=0.05, Δp=0.5%), (Δx=0.09, Δp=0.9%), (Δx=0.01, Δp=1.0%), (Δx=0.015, Δp=1.5%) and (Δx=0.02, Δp=2.0%).

[0126] Table 12: One-time refinement results of PbSO4 with different random variation ranges (3 background function coefficients)

[0127]

[0128] The best result of the random refinement result was found and recorded as the best result 2. After checking the random refinement results (Table 12), a random value of Δx = 0.002 was introduced to the atomic fraction coordinates in the crystal structure, and a random change of Δp = 0.2% was introduced to other parameters. The result of the 3255th one-time refinement was used as the initial parameters. The crystal structure was correct and the refinement fitting factor R wp =10.43974 is the lowest. The weight fitting factor and running times of the refined results are as follows Fig.16 As shown in the figure, the red dotted line is the R of the optimal result 1 obtained by strategy refinement of PbSO4 spectrum. wp Value (Strategy 1, 3 background function coefficients). Compare the optimal result 1 and the optimal result 2, select the refinement result with the lowest refinement fitting factor and the correct crystal structure, and output the crystal structure information. Compared with Example 1, the random change one-time refinement result of Example 3 also obtains a better fit factor than the strategy refinement result, and the higher fit factor is caused by too few background function coefficients.

[0129] Example 4

[0130] This embodiment takes the crystalline material PbSO4 as an example, and adopts the coefficients of the 7 background functions related to Example 1. The difference is that the initial value of the random change is not the result of strategy 1, but strategy 4, strategy 5 and strategy 6 with slightly worse fitting results, in order to illustrate the effect of randomly changing a refinement (second refinement) under different refinement qualities and refinement strategies. Use Python to call the comprehensive structural analysis system GSAS-II. Set the background function type to "chebyschev-1", the number of background function coefficients to 7, and the number of refinement cycles to 12. Load the obtained structural parameters, instrument parameters, and diffraction data; the others are the GSASII default values. Fit and refine the powder X-ray diffraction pattern based on the crystal structure information, and adopt 6 different refinement strategies to obtain the weight fitting factor R of the refinement results of different refinement strategies. wp , record the results of each strategy refinement. The 6 strategies used are shown in Table 2, the PbSO4 strategy refinement results when the number of background function coefficients is 7 are shown in Table 10; the structural parameters of the refined results of strategies 4, 5 and 6 are shown in Tables 13-15 respectively.

[0131] Table 13: Structural parameters of PbSO4 refined based on strategy 4 (the number of background function coefficients is 7)

[0132]

[0133] Table 14: Structural parameters of PbSO4 refined based on strategy 5 (the number of background function coefficients is 7)

[0134]

[0135]

[0136] Table 15: Structural parameters of PbSO4 refined based on strategy 6 (the number of background function coefficients is 7)

[0137]

[0138] The initial refinement results of strategies 4, 5 and 6 were used as the starting values, and random changes in different ranges were introduced as initial refinement parameters for multiple rounds of one-time refinement (second refinement), and the results of each refinement were recorded. Based on the results obtained by refining the PbSO4 spectrum using strategies 4, 5 and 6, random values ​​of Δx were introduced into the atomic fraction coordinates in the crystal structure, and random changes of Δp were introduced into other parameters as initial refinement parameters, and one-time refinement was performed, and the refinement results (including refinement parameters and refinement fitting factors) were recorded and saved, and repeated 5000 times, including (Δx = 0.0001, Δp = 0.01%), (Δx = 0.001, Δp = 0.1%), (Δx = 0.002, Δp = 0.2%), (Δx = 0.005, Δp = 0.5%) and (Δx = 0.01, Δp = 1.0%). Strategy 4, Strategy 5 and Strategy 6 are the fitting factor results of the results of refining PbSO4 with random changes in the initial values, and are shown in Tables 16, 17 and 18.

[0139] Table 16: PbSO4 random variation based on strategy 4 refinement results (7 background function coefficients)

[0140]

[0141] Table 17: PbSO4 random variation based on strategy 5 refinement results (7 background function coefficients)

[0142]

[0143] Table 18: PbSO4 random variation based on strategy 6 refinement results (7 background function coefficients)

[0144]

[0145] Find the best result of the random refinement result, record it as the best result 2. From the data in Table 16, it can be seen that the best result of the random change refinement based on strategy 4 is the random change (Δx = 0.002, Δp = 0.2%). The weight fitting factor and the number of runs of the refinement result are as follows: Fig.17 As shown in Table 13, the unit cell parameters, atomic fraction coordinates, Mustrain_isotropic, Size_isotropic and fitting factors obtained by refinement are shown in Table 13. From the data in Table 17, it can be seen that the best result of random variation of strategy 5 is a single refinement with random variation (Δx = 0.005, Δp = 0.5%). The weight fitting factor and the number of runs of the refinement results are shown in Table 13. Fig.18As shown, the unit cell parameters, atomic fraction coordinates, Mustrain_isotropic, Size_isotropic and fitting factors obtained by refinement are shown in Table 14. From the data in Table 18, it can be seen that the best result of a random refinement based on strategy 6 is a random refinement (Δx=0.005, Δp=0.5%), and the unit cell parameters, atomic fraction coordinates, Mustrain_isotropic, Size_isotropic and fitting factors obtained by refinement are shown in Table 15. Compare the best results 1 and 2, and select the best result as the final refinement result. The refinement based on strategy 4 performs best when (Δx=0.002, Δp=0.2%); strategies 5 and 6 both reach the best when (Δx=0.005, Δp=0.5%). . Compared with Example 1, the random change single refinement result of this embodiment also obtained a better fitting factor than the strategy refinement result; the higher fitting factor is caused by the high fitting factor of the strategy refinement result, which shows that the initial refinement result of the random change single refinement should be as good as possible. Therefore, the method of the present invention adopts multiple strategies and selects the optimal result 1 as the initial value of the random change single refinement.

[0146] In summary, the automated crystal structure refinement and analysis method based on random one-time refinement provided by the present invention can find better refinement results by introducing random changes to perform multiple rounds of one-time refinement (second refinement). The refinement results are in good agreement with the observed values ​​at most peak positions, indicating the accuracy of the crystal structure model. This method is highly automated, efficient and accurate, and is suitable for structure refinement and analysis of different crystal materials.

[0147] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. An automated crystal structure refinement analysis method based on random one-time refinement, characterized in that: include: S1, obtaining diffraction data, structural parameters and instrument parameters of the crystal material to be analyzed; S2, designing different refinement strategies based on the structural parameters and instrument parameters of the crystal material to be analyzed, performing initial refinement on the diffraction data of the crystal material to be analyzed based on the different refinement strategies, obtaining initial refinement results including initial refinement fitting factors, and selecting an optimal result 1 from the initial refinement results, wherein the optimal result 1 is a refinement result with the smallest initial refinement fitting factor and a correct crystal structure; S3, introducing random changes to the atomic fraction coordinates and other parameters in the optimal result 1, performing a second refinement on the diffraction data in the optimal result 1, obtaining a second refinement result including a second refinement fitting factor, and selecting the optimal result 2 from the second refinement result; the optimal result 2 is a refinement result with the smallest second refinement fitting factor and a correct crystal structure; S4, comparing the best result 1 and the best result 2, selecting the refined result with the smallest refinement fitting factor and the correct crystal structure as the final result, and outputting the crystal structure information.

2. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 1, characterized in that: In S1, the structural parameters include space group, unit cell parameters, atomic fractional coordinates, atomic types and symmetry operations, which are saved in the structural model file; the instrument parameter data include the wavelength of X-rays and target material type, which are saved in the instrument parameter file.

3. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 1, characterized in that: S2 specifically includes: S21, importing the diffraction data, structural parameters and instrument parameters of the crystal material to be analyzed into the X-ray powder diffraction refinement engine software; S22, design different refinement strategies based on structural parameters and instrument parameters, import X-ray powder diffraction refinement engine software, and perform initial refinement on the diffraction data of the crystal material to be analyzed based on different refinement strategies; S23, calculating the initial refinement fitting factor, checking the refined structural parameters, and selecting the result with the smallest initial refinement fitting factor and correct crystal structure as the optimal result 1; the initial refinement result includes the initial refinement fitting factor, the initial refined structural parameters, the initial refined diffraction pattern and the visual representation of the initial refined structural parameters.

4. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 3, characterized in that: The X-ray powder diffraction refinement engine software is any one of GSAS-II, FULLPROF and TOPAS.

5. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 3, characterized in that: The different refinement strategies are designed based on the structural parameters and the instrument parameters and are obtained by adjusting the order of the structural parameters and the instrument parameters.

6. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 1, characterized in that: The second refinement specifically includes: A random change is introduced into the atomic fraction coordinates in the optimal result 1 to obtain the atomic fraction coordinates after the random change is introduced, and a random change percentage is introduced into other parameters in the optimal result 1 to obtain the values ​​of other parameters after the random change percentage is introduced; based on the atomic fraction coordinates after the random change is introduced and the values ​​of other parameters after the random change percentage is introduced, the diffraction data in the optimal result 1 is refined for the second time, and the k is refined. i times, obtaining a second refinement result including a refinement fitting factor; the other parameters include background factor, unit cell parameter, micro strain, grain size, temperature factor and instrument parameter; The atomic fraction coordinate after introducing random changes = optimal result + atomic fraction coordinate + random changes, random changes = random.uniform (-Δx i ,Δx i ); The other parameter values ​​after introducing the random change percentage = optimal result + other parameter values ​​× (1 + random change percentage); random change percentage = random.uniform (-Δp i ,Δp i ); Where i represents the number of random changes, Δx represents the random changes introduced by the atomic fraction coordinates, and Δx i represents the i-th random change introduced by the atomic fraction coordinates, Δp represents the percentage of random changes introduced by other parameters; Δp i represents the i-th random variation percentage introduced by other parameters; the k i represents the number of refinements to introduce the i-th random variation, Δx i <0.02,Δp i <2%, k i The value is 1000 to 10000, and i is 1 to 8.

7. The method for automated crystal structure refinement analysis based on random one-time refinement according to claim 1, characterized in that: The refined fitting factor is any one of a peak shape fitting factor, a weight fitting factor, a goodness of fit factor and an expected fitting factor.

8. A system for implementing the automated crystal structure refinement analysis method based on random one-time refinement according to claims 1 to 7, characterized in that: include: A data acquisition module, used to acquire diffraction data, structural model data and instrument parameter data of the crystal material to be analyzed; The initial refinement strategy design and execution module performs initial refinement on the diffraction data of the crystal material to be analyzed based on different refinement strategies to obtain initial refinement results, and selects the initial refinement result with the smallest initial refinement fitting factor as the optimal result 1; A random variation and second refinement module is used to introduce random variation to perform a second refinement on the optimal result 1, obtain a second refinement result including a second refinement fitting factor, and select the refinement result with the smallest second refinement fitting factor as the optimal result 2; The result selection and evaluation module is used to evaluate the initial refinement results and the second refinement results to obtain the refinement results with the lowest refinement fitting factor and the correct crystal structure.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the automated crystal structure refinement analysis method based on random one-time refinement according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the automated crystal structure refinement analysis method based on random one-time refinement according to any one of claims 1 to 7 are implemented.

Images