A method for screening doping elements in Te-containing phase change materials based on first principles

Abstract

The present invention discloses a method for screening doping elements for Te-containing phase change materials based on first principles, belonging to the field of phase change storage technology. The present invention uses first principles calculations to analyze the local structure and bonding properties of amorphous models based on element-doped Te-containing phase change materials from different angles at the atomic scale, thereby screening and optimizing different doping elements and doping components. The physical mechanism by which element doping comprehensively improves the crystallization rate, amorphous thermal stability, and reliability of Te-containing phase change materials is obtained at a microscopic level. By optimizing the three goals of improving amorphization stability, crystallization rate, and cycle characteristics, and comparing and analyzing multiple indicators for each optimization goal, the present invention achieves screening and optimization of different doping elements and doping concentrations, providing guidance for the research of element-doped Te-containing phase change materials and devices, thereby greatly reducing experimental costs, shortening the development cycle of materials and devices, and improving R&D efficiency.

Description

Technical Field

[0001] The present invention belongs to the technical field of phase change storage, and more specifically, relates to a method for screening doping elements of Te-containing phase change materials based on first principles. Background Art

[0002] Phase change materials are mainly sulfide compound materials, among which compounds composed of three elements: Ge, Sb, and Te are the most common. Optimizing the performance of phase change materials is the key to improving the performance of phase change memory, and the microstructure of phase change materials determines their macroscopic properties. At present, the main means of optimizing the performance of phase change materials is element doping. Taking the fourth main group element doped Sb-Te phase change material as an example, after doping, a tetrahedral structure can be formed with itself as the center. In the amorphous state, the strongly bonded tetrahedral clusters are quite different from the structure of the crystal (octahedron), which hinders the spontaneous crystallization of the phase change material, thereby improving its stability and data retention ability. However, while the above-mentioned doping improves the amorphous stability of the Sb-Te phase change material, it will inevitably reduce its crystallization speed, thereby affecting the SET speed of the memory device.

[0003] Therefore, it is necessary to screen the doping elements and optimize the doping concentration so that the doping elements can form strong bonds with the phase change material matrix elements to improve the thermal stability of the material while retaining the basic ring structure during the amorphization process, thereby increasing the number of original rings and improving the crystallization speed of the material. It is expected to achieve high-speed and high-reliability phase change memory, so that it can be used as a commercial phase change memory material.

[0004] Currently, the research on doping material screening and doping concentration optimization is generally completed through experiments, which is not only costly but also has a long R&D cycle and many uncontrollable factors. Summary of the Invention

[0005] In view of the defects of the prior art, the purpose of the present invention is to provide a method for screening doping elements of Te-containing phase change materials based on first principles, aiming to solve the problems of long R&D cycle and high cost of the prior art.

[0006] To achieve the above objectives, the present invention provides a method for screening doping elements of Te-containing phase change materials based on first principles, comprising:

[0007] S1. Select doping elements: Select an element from the periodic table whose electronegativity difference with Te is greater than 1 as the M element and proceed to S2. If all elements have been screened, proceed to S12.

[0008] S2. Constructing a control crystalline model and an optimal crystalline model: 1) establishing a crystalline model of an undoped Te-containing phase change material as a control crystalline model; 2) establishing crystalline models for different doping sites of M atoms, and the crystalline model with the lowest formation energy is used as the optimal crystalline model;

[0009] S3. Constructing a control amorphous model and an optimal amorphous model: 1) using first principles to simulate the randomization, melting and quenching process of the control crystalline model to obtain a control amorphous model; 2) using first principles to simulate the randomization, melting and quenching process of the optimal crystalline model to obtain an optimal amorphous model;

[0010] S4. Calculate the electron localization function of the optimal crystalline model. According to the distribution of the electron localization function, obtain the bond strength between the M atom and the non-Te atom. If the bond strength of the M atom is higher than that of the non-Te atom, proceed to S5; otherwise, proceed to S1.

[0011] S5. Calculate the pair distribution function of the M-Te chemical bond in the optimal amorphous model, and then determine the cutoff radius of the M atom and non-Te atoms as the basis for the calculation of S6-S7;

[0012] S6. Analyze the Peierls-like distortion degree of the control amorphous model and the optimal amorphous model respectively. If the Perierls-like distortion degree of the optimal amorphous model is stronger than that of the control amorphous model, mark the M element as meeting the first indicator. Otherwise, proceed to S1.

[0013] S7. Count the ring structure distributions of the control amorphous model and the optimal amorphous model respectively. If the number of four-membered rings in the optimal amorphous model is higher than that in the control amorphous model, proceed to S8; otherwise, proceed to S1.

[0014] S8. Based on the reference crystalline model, simulate the nucleation process and calculate the lattice orbital Hamiltonian population of the chemical bond between non-Te atoms and Te to obtain the bond strength and system stability of the chemical bond between non-Te atoms and Te. Similarly, obtain the M-Te bond strength and system stability in the optimal crystalline model. If the system stability of the optimal crystalline model is higher than that of the reference crystalline model, mark the M element as meeting the second indicator. Otherwise, proceed to S1.

[0015] S9. Count the mean square displacements of Te atoms in the control amorphous model and the optimal amorphous model respectively. If the mean square displacement of Te atoms in the optimal amorphous model is lower than that in the control amorphous model, proceed to S10; otherwise, proceed to S1.

[0016] S10. Calculate the electron localization functions of the control amorphous model and the optimal amorphous model respectively, and obtain the lone pair electron distribution as the calculation basis of S11;

[0017] S11. Further determine the void ratio in the low electron density distribution model based on the lone pair electron distribution of the control amorphous model and the optimal amorphous model. If the void ratio in the optimal amorphous model is lower than that in the control amorphous model, mark the M element as meeting the third indicator. Otherwise, proceed to S1.

[0018] S12. All M elements that meet the three criteria are selected as optimal doping elements.

[0019] Preferably, the Te-containing phase change material is a Ge-Sb-Te system, Sb-Te system, or Ge-Te system phase change material.

[0020] Preferably, in step S2, the system formation energy of different doping sites of the M atoms is obtained by static calculation, thereby determining whether the doping type of the M atoms is substitutional doping, vacancy doping, or interstitial doping.

[0021] Preferably, in step S4, the bond strength between the M atom and the non-Te atom is determined as follows:

[0022] The optimal crystalline model was subjected to structural optimization and static self-consistent calculations to obtain the electron localization function distribution map around M atoms and non-Te atoms on a specified two-dimensional plane.

[0023] The lower the degree of electron localization, the stronger the atom's ability to lose electrons and the higher the bond strength.

[0024] Preferably, in step S5, the cutoff radius is a value corresponding to the first trough in the distribution function.

[0025] Preferably, in step S6, atomic pairs with a distance greater than the cutoff radius are not bonded, all bond angles in the model are calculated, and bond angles greater than 150° are screened. The corresponding two bond lengths are recorded as r1 and r2, and then the distribution function of the correlation pair of r1 and r2 is calculated;

[0026] The most important bond length distributions of r1 and r2 correspond to the peaks. The farther the peaks are from the diagonal, the greater the difference between r1 and r2 and the greater the degree of Peierls-like distortion.

[0027] Preferably, in step S7, atomic pairs with a distance greater than the cutoff radius are not bonded to each other, and different types and numbers of rings are formed by atomic bonding in the statistical model.

[0028] Preferably, in step S8, the bonding strength and system stability are determined as follows:

[0029] Perform static self-consistent calculations on the optimal crystalline model and output the wave function file WAVECAR, which is then post-processed using the LOBSTER software package to calculate the COHP data between any atomic pairs.

[0030] The part of -COHP less than 0 represents the contribution of the antibonding state, and the part of -COHP greater than 0 represents the contribution of the bonding state. The less the contribution of the antibonding state below the Fermi level, the more stable the chemical bond and the stronger the bonding strength.

[0031] Preferably, in step S10, the lone pair electron distribution is determined as follows:

[0032] Calculate the electron localization function (ELF) and then display the distribution of lone pair electrons in the model by setting the isosurface to 0.85.

[0033] Preferably, when establishing the amorphous model in step S3, for the same M element, amorphous models of the M-doped Te-containing phase change material with different doping concentrations are established, thereby screening out the optimal doping concentration.

[0034] In general, the above technical solutions conceived by the present invention have the following beneficial effects compared with the prior art:

[0035] The present invention provides a method for screening doping elements for Te-containing phase change materials based on first principles. The first principles calculations adopted can analyze the local structure and bonding properties of amorphous models based on element-doped Te-containing phase change materials from different angles at the atomic scale, realize the screening and optimization of different doping elements and doping components, and obtain the physical mechanism of element doping to comprehensively improve the crystallization rate, amorphous thermal stability and reliability of Te-containing phase change materials from a microscopic level; through the three optimization goals of improving amorphization stability, improving crystallization rate and improving cycle characteristics, each optimization goal comprehensively compares and analyzes multiple indicators to realize the screening and optimization of different doping elements and doping concentrations, provide guidance for the research of element-doped Te-containing phase change materials and devices, thereby greatly reducing experimental costs, shortening the development cycle of materials and devices, and improving R&D efficiency; at the same time, the first principles calculations can avoid various uncontrollable experimental variables in the experiment and obtain accurate performance change laws. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a flow chart of a method for screening doping elements for Te-containing phase change materials based on first principles provided by the present invention.

[0037] Figure 2 It is the formation energy of the Rb-doped Sb2Te3 phase change material with different doping sites provided by the present invention.

[0038] Figure 3It is a two-dimensional planar distribution diagram of the electron localization function around the Rb atoms and Sb atoms in model 1 provided by the present invention.

[0039] Figure 4 It is a three-body correlation function distribution diagram of the restricted angle of Model 2 and Model 3 provided by the present invention.

[0040] Figure 5 This is a statistical diagram of the ring structure distribution of Model 2 and Model 3 provided by the present invention.

[0041] Figure 6 The present invention provides a lattice orbital Hamiltonian population calculation (COHP) diagram for Sb-Te and Rb-Te bonds based on Model 2 and Model 3, simulating the nucleation process.

[0042] Figure 7 This is a statistical diagram of the mean square displacement of Te atoms in Model 2 and Model 3 at 300K provided by the present invention.

[0043] Figure 8 It is a diagram of the void ratio in Model 2 and Model 3 provided by the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0045] The present invention provides a method for screening doping elements of Te-containing phase change materials based on first principles, wherein the Te-containing phase change materials are Ge-Sb-Te system, Sb-Te system or Ge-Te system phase change materials. Figure 1 As shown, the method includes:

[0046] S1. Select doping elements: Select an element from the periodic table whose electronegativity difference with Te is greater than 1 as the M element and proceed to S2. If all elements have been screened, proceed to S12.

[0047] S2. Construct a reference crystalline model and an optimal crystalline model: 1) Establish a crystalline model of an undoped Te-containing phase change material as a reference crystalline model; 2) Establish crystalline models for different doping sites of M atoms, and take the crystalline model with the lowest formation energy as the optimal crystalline model.

[0048] In step S2, the system formation energy of different doping sites of the M atoms is obtained by static calculation, thereby determining whether the doping type of the M atoms is substitutional doping, vacancy doping, or interstitial doping.

[0049] Materials Studio software was used to model the doping of M element atoms into Te-containing phase change materials. Crystalline models were established for different doping sites of M atoms. Structural optimization and static calculations were performed on the above-mentioned crystalline models in sequence to obtain the formation energy of M-doped Te-containing phase change materials at different doping sites. The crystalline model with the lowest formation energy was selected as the optimal crystalline model. At the same time, the crystalline model of undoped Te-containing phase change material was established using Materials Studio software, and structural optimization was performed. The obtained model was used as the control model.

[0050] S3. Construct a control amorphous model and an optimal amorphous model: 1) Use the first principles to simulate the randomization, melting and quenching process of the control crystalline model to obtain a control amorphous model; 2) Use the first principles to simulate the randomization, melting and quenching process of the optimal crystalline model to obtain an optimal amorphous model.

[0051] S4. Calculate the electron localization function of the optimal crystalline model, and obtain the bond strength between the M atom and the non-Te atom based on the distribution of the electron localization function. If the bond strength of the M atom is higher than that of the non-Te atom, proceed to S5; otherwise, proceed to S1.

[0052] In step S4, the bond strength between the M atom and the non-Te atom is determined as follows:

[0053] The optimal crystalline model was subjected to structural optimization and static self-consistent calculations. The resulting ELFCAR file was then processed using VESTA visualization software to generate a distribution diagram of the electron localization function around the M and non-Te atoms on a specified two-dimensional plane. A higher degree of electron localization indicates a stronger electron-acquisition capacity. Conversely, a higher degree of electron delocalization indicates a stronger electron-loss capacity. The bond strength of atoms is expressed by comparing their electron-loss capacities; a stronger electron-loss capacity indicates a stronger bond strength.

[0054] S5. Calculate the pair distribution function of the M-Te chemical bond in the optimal amorphous model, and then determine the cutoff radius of the M atoms and non-Te atoms as the calculation basis for S6~S7.

[0055] In step S5, the cutoff radius is the value corresponding to the first trough in the distribution function.

[0056] S6. Analyze the Peierls-like distortion degree in the control amorphous model and the optimal amorphous model respectively. If the Perierls-like distortion degree of the optimal amorphous model is stronger than that of the control amorphous model, mark the M element as meeting the first indicator. Otherwise, proceed to S1.

[0057] In step S6, all bond angles in the model are calculated (atomic pairs with a spacing greater than the cutoff radius are not bonded), and bond angles greater than 150° are screened. The corresponding two bond lengths are recorded as r1 and r2, and then the distribution function of the r1 and r2 correlation pairs is calculated. The most important r1 and r2 bond length distributions correspond to the peaks. The further the peak is from the diagonal, the greater the difference between r1 and r2, that is, the greater the degree of Peierls-like distortion.

[0058] The higher the bonding strength with Te atoms and the greater the Peierls-like distortion, the higher the amorphous stability of the material and the better the thermal stability of the device.

[0059] S7. Count the ring structure distributions of the control amorphous model and the optimal amorphous model respectively. If the number of four-membered rings in the optimal amorphous model is higher than that in the control amorphous model, proceed to S8; otherwise, proceed to S1.

[0060] In step S7, the statistical model calculates the number and types of rings formed by atomic bonding (where pairs of atoms separated by a distance greater than the cutoff radius do not form bonds). A four-membered ring structure is typically formed by connecting two octahedra. Four-membered rings are considered the basic unit rings for material crystallization, so the number of four-membered rings can be used as a reference indicator: a greater number of four-membered rings indicates a greater likelihood of crystallization.

[0061] S8. Based on the control crystal model, simulate the nucleation process, calculate the lattice orbital Hamiltonian population of the chemical bond formed by non-Te atoms and Te, and obtain the bond strength and system stability of the chemical bond formed by non-Te atoms and Te; similarly, obtain the M-Te bond strength and system stability in the optimal crystal model. If the system stability of the optimal crystal model is higher than that of the control crystal model, mark the M element as meeting the second indicator; otherwise, enter S1.

[0062] The optimal crystalline model was subjected to static self-consistent calculation using VASP software, and the wave function file WAVECAR was output. The LOBSTER software package was then used for post-processing to calculate the COHP data (lattice orbital Hamiltonian population) between any atomic pairs. COHP can describe the bonding and antibonding contributions of chemical bonds, thereby reflecting the bonding strength of the material and the stability of the system. The part where -COHP is less than 0 represents the contribution of the antibonding state, and the part where -COHP is greater than 0 represents the contribution of the bonding state. The less the antibonding contribution below the Fermi level, the more stable the chemical bond and the stronger the bonding strength, the stronger the ability to capture Te atoms, the higher the bonding strength, and the easier it is to nucleate.

[0063] The more four-membered rings there are, the stronger the capture ability is, which means the faster the material crystallizes and the faster the device erases and writes.

[0064] S9. Count the mean square displacements of Te atoms in the control amorphous model and the optimal amorphous model respectively. If the mean square displacement of Te atoms in the optimal amorphous model is lower than that in the control amorphous model, proceed to S10; otherwise, proceed to S1.

[0065] S10. Calculate the electron localization functions of the control amorphous model and the optimal amorphous model respectively to obtain the lone pair electron distribution as the calculation basis for S11.

[0066] In step S10, the lone pair electron distribution is determined as follows:

[0067] ELF is a dimensionless constant with a range of 0-1. ELF values ​​closer to 1 indicate a higher degree of electron localization. ELF values ​​closer to 0 indicate a higher degree of electron delocalization, with lone pairs having a high degree of electron localization. An ELF value of 0.85 indicates the formation of lone pairs. The Electron Localization Function (ELF) can be calculated and then the distribution of lone pairs in the model can be visualized by setting the isosurface to 0.85.

[0068] S11. Based on the lone pair electron distribution of the control amorphous model and the optimal amorphous model, further determine the void ratio in the low electron density distribution model. If the void ratio in the optimal amorphous model is lower than that in the control amorphous model, mark the M element as meeting the third indicator. Otherwise, proceed to S1.

[0069] The lower the mean square displacement of Te atoms and the smaller the void volume ratio, the higher the material reliability and the better the cycle durability of the device.

[0070] S12. All M elements that meet the three criteria are selected as optimal doping elements.

[0071] The method further includes:

[0072] When establishing the amorphous model in step S3, for the same M element, amorphous models of the M-doped Te-containing phase change material with different doping concentrations are established, thereby screening out the optimal doping concentration.

[0073] Example

[0074] In this embodiment, the doping element is Rb and the parent phase change material is Sb2Te3. The specific steps are as follows:

[0075] 1. In the periodic table, by comparing the electronegativity difference between Te and Rb, if the result is greater than 1, proceed to the next step of calculation.

[0076] Second, Materials Studio software was used to model the Rb element atoms doped into the Sb2Te3 phase change material. A crystalline model was established for the different doping sites of the Rb atoms. The structure optimization and static calculation of the above crystalline model were carried out in turn to obtain the formation energy of the Rb-doped Te-containing phase change material at different doping sites, as shown in the following example: Figure 2 As shown in the figure, compared with the system formation energy of other doping sites, the system formation energy is the lowest when the Rb atom is vacancy doped. The crystalline model of Rb in vacancy doping is taken as the optimal crystalline model, recorded as Model 1. At the same time, the crystalline model of undoped Sb2Te3 phase change material is established using Materials Studio software, and the structure is optimized. The obtained model is used as the control model, recorded as Model 0.

[0077] 3. Using the first principles, the randomization, melting and quenching processes of model 0 and model 1 with a certain doping concentration are simulated to obtain amorphous models, which are respectively recorded as model 2 and model 3 for subsequent simulation calculations.

[0078] 4. Calculate the electron localization function (ELF) for model 1, as follows: Figure 3 As shown, according to the distribution of electron localization function, it can be seen that compared with Sb atom, the degree of electron delocalization of Rb atom is higher, indicating that the ability of Rb atom to lose electrons is higher than that of Sb atom, which indicates that the bond strength between Rb atom and Te atom is higher than the chemical bond between Sb atom and Te, so the next calculation is carried out.

[0079] 5. Calculate the distribution function of the Rb-Te chemical bond in model 3 and determine the cutoff radius of the Rb and Sb atoms in model 3, which are and Used for subsequent calculations.

[0080] 6. Based on step 5, the angle-restricted three-body correlation function (ALTBC) is used to analyze the degree of Peierls-like distortion in models 2 and 3. The specific process is: calculate all the bond angles in the model (atomic pairs with a spacing greater than the cutoff radius are not bonded), select the bond angles greater than 150°, and record the corresponding two bond lengths as r1 and r2, and then calculate the distribution function of the r1 and r2 correlation pairs. The results are as follows: Figure 4 As shown in Figure 3, compared with model 2, the peaks corresponding to r1 and r2 in model 3 are farther away from the diagonal, indicating that the greater the difference between r1 and r2, the stronger the Perierls-like distortion of model 3 is than that of model 2, indicating that Rb element doping improves the amorphous thermal stability of Sb2Te3 phase change material, which is consistent with indicator 1.

[0081] 7. Based on step 5, the ring structure distribution statistics of model 2 and model 3 are carried out. Usually a four-membered ring structure is formed by connecting two octahedrons. It can be considered that the four-membered ring is the basic unit ring of the material crystallization, so the number of four-membered rings can be used as a reference indicator. The results are as follows Figure 5 As shown, the number of four-membered rings in model 3 is higher than that in model 2, so the next calculation is performed.

[0082] 8. Based on step 7, based on model 2, simulate the nucleation process, calculate the lattice orbital Hamiltonian population (COHP) of the Sb-Te bond, and obtain the Sb-Te bonding strength and system stability; based on model 3, simulate the nucleation process, calculate the lattice orbital Hamiltonian population (COHP) of the Rb-Te bond, and obtain the Rb-Te bonding strength and system stability. The results are as follows Figure 6 As shown, compared with model 2, model 3 has fewer antibonding states, indicating that the stability of the Rb-Te-Rb-Te system in model 3 is higher than that of the Sb-Te-Sb-Te in model 2, reflecting that the Rb atom capture ability is stronger than that of the Sb atom. The higher the bonding strength, the more Rb element doping improves the crystallization rate of the Sb2Te3 phase change material, which is in line with indicator 2.

[0083] 9. Based on step 5, the mean square displacement of Te atoms in model 2 and model 3 is statistically analyzed, such as Figure 7 As shown, the mean square displacement of Te atoms in model 3 is lower than that in model 2, so the next calculation is carried out.

[0084] 10. Based on step 9, perform electron localization function calculations on model 2 and model 3 respectively to obtain the lone pair electron distribution for subsequent calculations.

[0085] 11. Based on step 10, according to the lone pair electron distribution obtained by model 2 and model 3, the gap ratio in the low electron density distribution model is further used, and the results are as follows Figure 8 As shown, the void ratio in model 3 is lower than that in model 2, indicating that Rb element doping reduces the density difference between the crystalline and amorphous states of the Sb2Te3 phase change material, improves the cycle characteristics, and meets indicator 3.

[0086] 12. Based on steps 1 to 11, the amorphous stability, crystallization speed, and the impact of Rb-doped Te-containing phase-change memory materials on the data retention time, SET speed, and cycle endurance of the corresponding memory devices are finally analyzed according to the calculation results. If indicators 1, 2, and 3 are met at the same time, Rb is selected as the preferred doping element.

[0087] The present invention screens and optimizes the following three indicators: (1) improved amorphization stability (doping enhances chemical bond strength); (2) increased crystallization rate (residual ring structures in amorphization promote spontaneous crystallization); and (3) improved cyclic characteristics (doping makes the amorphous structure more compact, minimizes phase transition density differences, and pinning reduces matrix atom migration). By screening and optimizing different doping elements and doping concentrations, the physical mechanism by which element doping comprehensively improves the crystallization rate, amorphous thermal stability, and device reliability of Te-containing phase change materials is revealed at a microscopic level.

[0088] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for screening doping elements of Te-containing phase change materials based on first principles, characterized in that: include: S1. Select doping elements: Select an element from the periodic table whose electronegativity difference with Te is greater than 1 as the M element and proceed to S2. If all elements have been screened, proceed to S12. S2. Constructing a control crystalline model and an optimal crystalline model: 1) establishing a crystalline model of an undoped Te-containing phase change material as a control crystalline model; 2) establishing crystalline models for different doping sites of M atoms, and the crystalline model with the lowest formation energy is used as the optimal crystalline model; S3. Constructing a control amorphous model and an optimal amorphous model: 1) using first principles to simulate the randomization, melting and quenching process of the control crystalline model to obtain a control amorphous model; 2) using first principles to simulate the randomization, melting and quenching process of the optimal crystalline model to obtain an optimal amorphous model; S4. Calculate the electron localization function of the optimal crystalline model. According to the distribution of the electron localization function, obtain the bond strength between the M atom and the non-Te atom. If the bond strength of the M atom is higher than that of the non-Te atom, proceed to S5; otherwise, proceed to S1. S5. Calculate the pair distribution function of the M-Te chemical bond in the optimal amorphous model, and then determine the cutoff radius of the M atom and non-Te atoms as the basis for the calculation of S6-S7; S6. Analyze the Peierls-like distortion degree of the control amorphous model and the optimal amorphous model respectively. If the Perierls-like distortion degree of the optimal amorphous model is stronger than that of the control amorphous model, mark the M element as meeting the first indicator. Otherwise, proceed to S1. S7. Count the ring structure distributions of the control amorphous model and the optimal amorphous model respectively. If the number of four-membered rings in the optimal amorphous model is higher than that in the control amorphous model, proceed to S8; otherwise, proceed to S1. S8. Based on the reference crystalline model, simulate the nucleation process and calculate the lattice orbital Hamiltonian population of the chemical bond between non-Te atoms and Te to obtain the bond strength and system stability of the chemical bond between non-Te atoms and Te. Similarly, obtain the M-Te bond strength and system stability in the optimal crystalline model. If the system stability of the optimal crystalline model is higher than that of the reference crystalline model, mark the M element as meeting the second indicator. Otherwise, proceed to S1. S9. Count the mean square displacements of Te atoms in the control amorphous model and the optimal amorphous model respectively. If the mean square displacement of Te atoms in the optimal amorphous model is lower than that in the control amorphous model, proceed to S10; otherwise, proceed to S1. S10. Calculate the electron localization functions of the control amorphous model and the optimal amorphous model respectively, and obtain the lone pair electron distribution as the calculation basis of S11; S11. Further determine the void ratio in the low electron density distribution model based on the lone pair electron distribution of the control amorphous model and the optimal amorphous model. If the void ratio in the optimal amorphous model is lower than that in the control amorphous model, mark the M element as meeting the third indicator. Otherwise, proceed to S1. S12. All M elements that meet the three criteria are selected as optimal doping elements.

2. The method according to claim 1, wherein The Te-containing phase change material is a Ge-Sb-Te system, Sb-Te system, or Ge-Te system phase change material.

3. The method according to claim 1, wherein In step S2, the system formation energy of different doping sites of the M atom is obtained by static calculation, thereby determining whether the doping type of the M atom is substitutional doping, vacancy doping, or interstitial doping.

4. The method according to claim 1, wherein In step S4, the bond strength between the M atom and the non-Te atom is determined as follows: The optimal crystalline model was subjected to structural optimization and static self-consistent calculations to obtain the electron localization function distribution map around M atoms and non-Te atoms on a specified two-dimensional plane. The lower the degree of electron localization, the stronger the atom's ability to lose electrons and the higher the bond strength.

5. The method according to claim 1, wherein In step S5, the cutoff radius is the value corresponding to the first trough in the distribution function.

6. The method according to claim 1, wherein In step S6, atomic pairs with a distance greater than the cutoff radius are not bonded. All bond angles in the model are calculated, and those with a bond angle greater than 150° are selected. The corresponding two bond lengths are recorded as r1 and r2, and then the distribution function of the correlation pair of r1 and r2 is calculated; The most important bond length distributions of r1 and r2 correspond to the peaks. The farther the peaks are from the diagonal, the greater the difference between r1 and r2 and the greater the degree of Peierls-like distortion.

7. The method according to claim 1, wherein In step S7, atom pairs with a distance greater than the cutoff radius do not form bonds, and different types of rings and their numbers are formed by atomic bonding in the statistical model.

8. The method according to claim 1, wherein In step S8, the bonding strength and system stability are determined as follows: Perform static self-consistent calculations on the optimal crystalline model and output the wave function file WAVECAR, which is then post-processed using the LOBSTER software package to calculate the COHP data between any atomic pairs. The part of -COHP less than 0 represents the contribution of the antibonding state, and the part of -COHP greater than 0 represents the contribution of the bonding state. The less the contribution of the antibonding state below the Fermi level, the more stable the chemical bond and the stronger the bonding strength.

9. The method according to claim 1, wherein In step S10, the lone pair electron distribution is determined as follows: Calculate the electron localization function (ELF) and then display the distribution of lone pair electrons in the model by setting the isosurface to 0.

85.

10. The method according to any one of claims 1 to 9, characterized in that When establishing the amorphous model in step S3, for the same M element, amorphous models of the M-doped Te-containing phase change material with different doping concentrations are established, thereby screening out the optimal doping concentration.

Images