Method for testing surface roughness of silicon-based solar cells based on molecular dynamics simulation

Abstract

The present invention discloses a method for testing the surface roughness of silicon-based solar cells based on molecular dynamics simulation. The method uses the "deposit" command in the Lammps software to deposit SiH3 molecules onto a smooth crystalline silicon substrate. The "deposit" command in the molecular dynamics simulation serves as a loading tool. According to Newton's second law, the velocities of the atoms that make up the molecules change after deposition, and the forces between the atoms cause the SiH3 molecules to grow on the crystalline silicon, resulting in the growth of a hydrogenated amorphous silicon film on the crystalline silicon. A mixed potential function is used to describe the atomic-to-atomic interactions between the crystalline silicon and the SiH3 molecules. This method, based on the open-source software Lammps and the visualization tool Ovito, combines the roughness formula with the visualization results to calculate the surface roughness of hydrogenated amorphous silicon films grown on crystalline silicon, providing theoretical guidance for the conditions required for practical film production.

Description

Technical Field

[0001] The present invention relates to the field of thin film preparation, and in particular to a method for testing the surface roughness of a silicon-based heterojunction solar cell passivation layer (hydrogenated amorphous silicon) during growth based on molecular dynamics simulation. Background Art

[0002] Thin film growth techniques include chemical vapor deposition (CVD), plasma chemical vapor deposition (PCVD), and atomic layer deposition (ALD). During the deposition process, uneven thermal expansion and momentum transfer from impinging atoms to the deposited film surface can lead to poor film quality. The film's surface morphology affects the film's optical, electrical, corrosion, and oxidation resistance. The density of the passivation layer (i.e., film quality) during growth in silicon-based heterojunction solar cells impacts the cell's conversion efficiency. Atomic distribution significantly influences the bond strength between the film and the substrate. Thin film growth is an extremely complex, nonequilibrium molecular dynamics process. Surface roughness is a macroscopic manifestation of the microstructure during film growth. Only by understanding the film formation mechanism at the atomic and molecular levels and revealing the growth mechanism of surface roughness during film growth can we fundamentally control film surface roughness. Experimental techniques only allow the study of a few integral parameters that characterize thin film structural properties. Detailed study of thin film structural properties remains a major challenge in modern thin film optical technology.

[0003] Studying the microscopic mechanisms of thin film growth at the atomic level is of great scientific significance for understanding the physical nature of thin film growth, understanding the influence of surface atomic microscopic physical and chemical processes on film growth patterns, surface interface morphology, and microscopic states, and establishing the intrinsic connection between the microscopic state and macroscopic properties of thin films. Therefore, molecular dynamics simulation is an indispensable theoretical tool for studying the microscopic mechanisms of thin film growth. Summary of the Invention

[0004] In response to the deficiencies of the above-mentioned prior art, the present invention provides a method for testing the surface roughness of the growth of the passivation layer of a silicon-based heterojunction solar cell based on molecular dynamics simulation, specifically a method for testing the surface roughness of SiH3 molecules after deposition on smooth crystalline silicon using molecular dynamics simulation. This method can be based on the open source software Lammps and the visualization software Ovito, combined with the roughness formula, to calculate the roughness of the grown thin film. This method combines theory with practical operation, overcomes the drawbacks of the cumbersome theory and the inability to intuitively survey, and provides a basis for the experimental preparation of thin films. The present invention takes the preparation of the passivation layer of a silicon-based heterojunction solar cell as an example, that is, the growth of a hydrogenated amorphous silicon film on a smooth crystalline silicon surface as an example.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A method for testing the surface roughness of silicon-based solar cells based on molecular dynamics simulation mainly includes the following steps:

[0007] The present invention uses the "deposit" deposition command in Lammps software to deposit SiH3 molecules downward as deposited molecules onto a smooth crystalline silicon substrate. The structure of a silicon-based heterojunction solar cell consists of a crystalline silicon layer (c-Si), a passivation layer (hydrogenated amorphous silicon (a-Si:H)), and a doping layer. The present invention primarily simulates the surface roughness changes after the hydrogenated amorphous silicon layer is deposited on the crystalline silicon layer. The hydrogenated amorphous silicon layer is formed by depositing SiH3 molecules onto the smooth crystalline silicon, forming an interface between the crystalline silicon and the hydrogenated amorphous silicon. Because the deposition process results in a loose hydrogenated amorphous silicon film, poor film quality, and surface roughness, it is necessary to explore the optimal surface roughness to improve film quality and enhance the efficiency of silicon-based solar cells. More specifically, deposition is used as a loading tool in molecular dynamics simulations. According to Newton's second law, the speed of atoms changes, and the forces between atoms cause atoms to grow on the crystalline silicon. Among them, the present invention adopts a mixed potential function (LJ potential function and Tersoff potential function) to describe the interaction between atoms in crystalline silicon and SiH3 group molecules. The use of the mixed potential function is more suitable for the deposition of molecular models and crystalline silicon systems, so the numerical simulation of the roughness of the passivation layer film growth is more accurate.

[0008] The method comprises the following steps:

[0009] Step 1: Use Materials Studio (MS) software to establish the SiH3 group molecular structure and crystalline silicon substrate model. Use the Lammps software toolkit msi2lmp program to convert the coordinate file of the established SiH3 group molecular structure and crystalline silicon substrate model into a data file that can be recognized by Lammps software. Construct a simulation box of 0.8nm*1.5nm*5nm. The number of crystalline silicon atoms in the simulation box is 416. The crystalline silicon uses the most stable (111) surface, which is consistent with the experiment. The vacuum layer height is 3nm, which serves as the spatial region for the deposition of molecules. Use the program written in Lammps software to realize the molecular dynamics simulation calculation of the growth of the passivation layer structure (hydrogenated amorphous silicon) in silicon-based solar cells.

[0010] Step 2: Use crystalline silicon as the substrate, with a vacuum layer of a few nanometers on top. SiH3 molecules serve as the deposition material, emerging from the top of the vacuum layer and then deposited onto the substrate. First, set the simulation system of crystalline silicon in Lammps software. Set the initial simulation parameters and use the three-dimensional simulation mode. Select the metal format for the measurement unit. Set the boundaries of the three directions of the simulation space to periodic boundaries. Set the atom type to full format, the bond type between atoms to harmonic format, and set the masses of silicon and hydrogen atoms. Set the initial coordinates and initial velocity of the simulation system and select the simulation time step.

[0011] Step 3: Select the interatomic interaction potential and use a hybrid potential function, namely the many-body Tersoff potential and the two-body LJ potential, to describe the interactions between silicon atoms and silicon atoms and silicon atoms and hydrogen atoms. For this, set the pair_style type to hybrid tersofflj / cut 10. Use the finite-difference Velocity-Verlet algorithm for simulations, with a time step of 0.2 femtoseconds. Reducing the time step can increase simulation accuracy.

[0012] Step 4: fix is ​​a command in Lammps software, and its function is to apply a certain operation to a group of particles. Selecting the fix deposit command is the most commonly used command for molecular or atomic deposition. Every 10,000 steps (i.e. 0.002ns), a molecule is inserted into the vacuum layer of the simulation box, and a total of 600 molecules are inserted. However, since molecules may detach from the substrate and fly into the vacuum layer during the simulation, a deletion area is set at the top of the vacuum layer of the simulation box to quickly delete the molecules that detach from the substrate to ensure that they do not affect the subsequent molecular deposition; this is achieved by the thermo_modify command in Lammps software. It should be noted that since molecules will be continuously inserted during the simulation, the running speed of Lammps software will drop significantly towards the end of the simulation, and the performance requirements of the server are relatively high.

[0013] Step 5: Divide the simulation system of the crystalline silicon substrate from top to bottom into three parts: free zone, constant temperature zone, and fixed zone. Use the region command to fix the fixed zone by imposing boundary constraints. In the free zone, atoms are allowed to move freely under the laws of Newtonian mechanics. Use the group command to group the crystalline silicon substrate and deposited molecules for accurate identification of the deposition process.

[0014] Step 6: Apply random initial velocity to the crystalline silicon substrate model and perform energy minimization calculation to determine the energy optimal structure of the crystalline silicon substrate model, that is, the most energy stable structure.

[0015] Step 7: The NVT ensemble is used to relax the entire crystalline silicon substrate model. The temperature control method during relaxation adopts the Nose-Hoover method to control the temperature at 475K (consistent with the experimental growth conditions) to ensure that the temperature of growing hydrogenated amorphous silicon on the crystalline silicon substrate is the same as that in the experiment.

[0016] Step 8: Based on the above steps 4, 5, 6, and 7, obtain the final data of the simulation system, import the output file containing the atomic coordinate information into the Ovito visualization software, and obtain snapshots of the film deposition at different times and the final morphology of the film growth.

[0017] Step 9: Based on the data extracted in step 8, it is necessary to post-process the statistical information and obtain the number of growing molecular atoms and atomic coordinate parameters in Ovito software; using the roughness RMS formula, The number of atoms on the surface of hydrogenated amorphous silicon film after growth, i.e., the number of atoms that form amorphous silicon film after SiH3 deposition, the average height value of the growing atoms, and the z-axis value of the growing atoms, i.e., the number of atoms that form amorphous silicon film after SiH3 deposition, are counted to calculate the roughness of the surface of hydrogenated amorphous silicon film. n represents the number of atoms deposited on the substrate, z represents the number of atoms that form amorphous silicon film after SiH3 deposition. i represents the z coordinate of the i-th atom, represents the average z-coordinate of all deposited atoms.

[0018] Compared with the existing technology, the beneficial effects of the present invention are: it realizes the detection of surface morphology during film growth, surveys the growth status of the film by changing the deposition conditions, provides a theoretical basis for experimental preparation of thin films, and realizes a computational simulation method for controlling the surface smoothness of the film.

[0019] The present invention mainly has the following advantages:

[0020] 1) The molecular dynamics simulation provided by the present invention for calculating the surface roughness of thin film growth can not only effectively reduce the cost and consumption caused by experiments, but also simulate experiments that cannot be accurately performed and completed due to limitations of related equipment and materials.

[0021] 2) Materials Studio software can be used to establish a defective substrate crystalline silicon model (e.g., a convex crystalline silicon substrate or a concave substrate, or a substrate model with roughness). By changing the deposition time, deposition incident energy, etc., the surface roughness can be repaired. The present invention can also obtain the surface roughness of the film after molecular deposition growth based on the molecular dynamics simulation method, providing a theoretical basis for thin film preparation.

[0022] 3) Materials Studio software can be used to establish different molecular structures and simulate various molecular structures such as organic macromolecules, organic long-chain molecules, and small molecules to deposit thin films on different substrates. The surface roughness of the film after growth can also be obtained through molecular dynamics simulation methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A molecular dynamics simulation flow chart provided for a preferred embodiment of the present invention.

[0024] Figure 2 This graph shows the evolution of molecules deposited onto a substrate at different times. At t = 0 ps, ​​the initial structure is a 0.8 nm x 1.5 nm x 5 nm simulation box with a 3 nm vacuum layer, serving as the initial substrate model. At t = 2 ps, molecules appear in the vacuum layer. These molecules are automatically inserted and generated at a predetermined position 0.5 nm from the top of the vacuum layer, and then deposited onto the silicon substrate using the Deposition command.

[0025] Figure 3 This is a visual representation of the molecules of the present invention when deposited onto a substrate.

[0026] Figure 4 This is a result diagram of the surface roughness calculation of the thin film grown after molecular deposition on a defective crystalline silicon substrate in the present invention. The left figure is the initial modeling structure with RMS=1nm. DETAILED DESCRIPTION

[0027] The present invention is further described below by a specific embodiment. The specific embodiment of the present invention is a method for testing the surface roughness of the passivation layer of a silicon-based solar cell based on molecular dynamics simulation. Figure 1 Flowchart of a molecular dynamics simulation of molecules deposited on a crystal substrate:

[0028] 1) Using Materials Studio (MS) software, we modeled the SiH3 molecular structure and crystalline silicon substrate. Using the msi2lmp program in the Lammps toolkit, we converted the coordinate files for the SiH3 molecular structure and crystalline silicon substrate into Lammps-readable data files. We then used Lammps to develop a program to perform molecular dynamics simulations of the growth of the solar cell passivation layer.

[0029] 2) Set the initial simulation parameters, use the three-dimensional simulation mode, select the metal format for the measurement unit, set the boundaries of the three directions of the simulation space to periodic boundaries, set the atom type to full format, the bond type to harmonic format, and set the masses of silicon atoms and hydrogen atoms; set the initial coordinates and initial velocity of the simulation system, and select an appropriate simulation time step.

[0030] 3) Select appropriate interatomic interaction potentials, using hybrid potential functions (i.e., the many-body Tersoff potential and the two-body LJ potential) to describe the interactions between silicon atoms and silicon atoms, and between silicon atoms and hydrogen atoms. The pair_style type is set to hybrid tersofflj / cut 10. During the calculation, the silicon crystal substrate model is divided into three parts: a constant temperature region, a free region, and a fixed region. The fixed region is fixed using the region command. The finite difference Velocity-Verlet algorithm is used for the simulation, so the time step is set to 0.2 femtoseconds. Reducing the time step can increase simulation accuracy.

[0031] 4) fix is ​​a command in Lammps, and its function is to apply a certain operation to a group of particles. Selecting the fix deposit command is the most commonly used command for molecular (atom) deposition. A molecule is inserted into the vacuum area of ​​the system every 10,000 steps (i.e. 0.002ns), and a total of 600 molecules are inserted. However, since molecules may detach from the substrate and fly into the vacuum layer during the simulation, a deletion area is set at the top of the vacuum layer of the simulation box to quickly delete the molecules that detach from the substrate to ensure that they do not affect the deposition of subsequent molecules. The thermo_modify command is used. It should be noted that since molecules will be continuously inserted during the simulation, the system operation speed will drop significantly in the later stages of the simulation, and the performance requirements of the server are relatively high.

[0032] 5) Divide the substrate crystalline silicon model from top to bottom into three parts: free zone, constant temperature zone, and fixed zone. Apply boundary constraints to fix the fixed zone using the region command.

[0033] 6) Apply random initial velocity to the entire model, perform energy minimization calculations, and determine the energy-optimal structure of the model.

[0034] 7) The NVT ensemble was used to relax the entire model. The Nose-Hoover method was used to control the temperature during relaxation, and the temperature was controlled at 475K (consistent with the experimental growth conditions). The simulation was not affected by temperature.

[0035] 8) Import the output file containing the atomic coordinate information into the Ovito visualization software to view and obtain snapshots of the film deposition at different times and the final morphology of the film growth. The snapshots at different times calculated in the present invention are as follows: Figure 2 And the molecular deposition intuitive diagram Figure 3 .

[0036] 9) Perform post-processing on the statistical information as needed. Obtain the number of growing molecular atoms and atomic coordinate parameters in Ovito software, and use the roughness (RMS) formula to calculate the The number of atoms on the surface of the film after growth, the average height of the growing atoms, and the z-axis value of the growing atoms are counted to calculate the roughness of the film surface. n represents the number of atoms deposited on the substrate, z i The z coordinate of the i-th atom is represented by the average z coordinate of all deposited atoms. The larger the RMS value, the higher the surface roughness of the film and the worse the quality. The lowest surface roughness and the smallest RMS value indicate the best surface quality. The surface roughness of the film grown on a non-smooth substrate (i.e., when RMS = 1) varies with the growth thickness. Figure 4 .

[0037] In summary, the above is only a specific embodiment of the present invention, but the scope of protection of the invention is not limited to this. Any modifications that can be made by any engineer or technician familiar with the technical field within the technical scope of the present invention should be considered as infringements of the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for testing the surface roughness of silicon-based solar cells based on molecular dynamics simulation, characterized in that: The method comprises the following steps: Step 1: Use Materials Studio software to create a model of SiH3 group molecules and a crystalline silicon substrate. Use the msi2lmp program in Lammps software to convert the coordinate file of the established SiH3 group molecule and crystalline silicon substrate model into a data file recognizable by Lammps software. Construct a simulation box with 416 crystalline silicon atoms in it, and select the most structurally stable surface of the crystalline silicon substrate. The vacuum layer of the simulation box is the space for depositing molecules. Step 2: A vacuum layer of several nanometers is placed on the crystalline silicon substrate. SiH3 group molecules, acting as the deposition material, are inserted from the top of the vacuum layer and then deposited. In Lammps software, the simulation system for the crystalline silicon substrate is first set up. The initial simulation parameters are then set, and the 3D simulation mode is used. The measurement unit is selected as metal format, and the boundaries of the simulation space in all three directions are set to periodic boundaries. The atom type is set to full format, the bond type between atoms is set to harmonic format, and the masses of silicon and hydrogen atoms are set. The initial coordinates and initial velocity of the simulation system are set, and the simulation time step is selected. Step 3: Select the interatomic interaction potential and use the hybrid potential function, i.e., the many-body potential tersoff potential function and the two-body potential LJ potential function, to describe the interactions between silicon atoms and silicon atoms, and between silicon atoms and hydrogen atoms. That is, set the pair_style type in the Lammps software to hybrid tersofflj / cut 10. Step 4: Select the fix deposit command in Lammps to insert a SiH3 group molecule into the vacuum layer of the simulation box every 10,000 steps, for a total of 600 SiH3 group molecules. Using the thermo_modify command in Lammps, set a delete region at the top of the vacuum layer of the simulation box to delete molecules that have detached from the crystalline silicon substrate to ensure that they do not affect the subsequent deposition of SiH3 group molecules. Step 5: Divide the simulation system of the crystalline silicon substrate from top to bottom into three parts: a free zone, a constant temperature zone, and a fixed zone. Use the region command to fix the fixed zone using boundary constraints. In the free zone, atoms are allowed to move freely under the laws of Newtonian mechanics. Use the group command to group the crystalline silicon substrate and deposited molecules for identification of the deposition process. Step 6: Apply a random initial velocity to the crystalline silicon substrate and perform energy minimization calculation to determine the energy optimal structure of the crystalline silicon substrate, that is, the most energy stable structure; Step 7: The entire crystalline silicon substrate is relaxed using the NVT ensemble. The temperature is controlled at 475K using the Nose-Hoover method during relaxation. Step 8: Based on the above steps 4, 5, 6, and 7, the final data of the simulation system is obtained. The output file containing the atomic coordinate information is imported into the Ovito visualization software. The snapshots of the film deposition at different times and the final morphology of the film growth are obtained based on the Ovito visualization software. Step 9: Based on the morphology image in step 8, obtain the number of growing molecular atoms and atomic coordinate parameters in the Ovito visualization software; use the roughness RMS formula to The number of atoms on the surface of the hydrogenated amorphous silicon film after growth, i.e., the number of atoms that form the hydrogenated amorphous silicon film after SiH3 deposition, the average height value of the growing atoms, and the z-axis value of the growing atoms, i.e., the number of atoms that form the hydrogenated amorphous silicon film after SiH3 deposition, are counted to calculate the surface roughness of the hydrogenated amorphous silicon film; n represents the number of atoms deposited on the substrate, z represents the average height value of the growing atoms, and z represents the average height value of the growing atoms. i represents the z coordinate of the i-th atom, represents the average z-coordinate of all deposited atoms.

Images