Simulation software method for virtual production line of semiconductor chip

Abstract

The invention discloses a semiconductor chip virtual production line simulation software method which comprises an etching module, a deposition module, a gluing module, a photoetching module, a cleaning module, a photoresist removing module, an ion implantation module, a rapid heat treatment module, a thermal oxidation module and a CMP module so as to simulate complete production simulation of a semiconductor chip. Models such as plasma density, frequency, Debye length and Poisson equation potential distribution are introduced into the etching module to carry out modeling simulation on an etching process, the simulation reduction capacity of the models is enhanced in a correction term compensation mode, and the deposition module comprises a plasma enhanced chemical vapor deposition (PECVD) module, a low-pressure chemical vapor deposition (LPCVD) module and a physical vapor deposition (PVD) module. According to the method, the deposition behavior is modeled, correction terms are introduced to compensate non-ideal factors, the adaptive capacity of the model to a non-ideal process environment is improved, finally, wafer morphology graphs and multi-physical field thermodynamic diagrams of all modules are output, and the production effect of wafers under different process parameters is visually presented.

Description

Technical Field

[0001] This invention relates to the field of semiconductor chip technology, and in particular to a software method for simulating a virtual production line for semiconductor chips. Background Technology

[0002] As the core of the information technology industry, the semiconductor industry has experienced explosive growth in recent years. As the core product of the semiconductor industry, chips have increasingly complex manufacturing processes, which has placed higher demands on the quantity and quality of professional talents. However, the traditional talent training model is difficult to meet the needs of the rapid development of the chip industry, especially in the practical teaching aspect, where there are obvious shortcomings.

[0003] Traditional chip manufacturing education mainly relies on theoretical lectures, process demonstrations, and limited visits to actual production lines. Theoretical lectures are abstract and difficult to understand, while process demonstrations lack interactivity. For example, when explaining key processes such as etching and photolithography in chip manufacturing, students can only watch videos or look at pictures, making it difficult for them to understand the complex principles and process details of chip manufacturing. They are unable to combine complex theoretical knowledge with actual production processes. On-site visits are limited by factors such as factory production safety and equipment operation. Students can only observe and cannot truly participate in the chip manufacturing process. As a result, when faced with actual production problems, students lack the ability and experience to solve practical problems. Summary of the Invention

[0004] This invention aims to overcome at least one of the defects of the prior art and provide a semiconductor chip virtual production line simulation software method to solve the technical problems of insufficient interactive feedback and lack of equipment operation experience in chip manufacturing teaching in the prior art.

[0005] This invention provides a method for simulating a virtual production line for semiconductor chips, comprising: The etching module is used to simulate etching the wafer, remove the material on the wafer surface, and output the etched topography of the wafer. The deposition module is used to simulate the formation of deposited materials on wafers and output a deposition morphology map of the wafers; The photoresist coating module is used to simulate coating photoresist onto the wafer surface and output a coating coverage topography image. The photolithography module is used to simulate transferring a mask pattern onto the wafer surface and output the outline of the pattern structure formed by photolithography. The cleaning module is used to simulate the removal of contaminants generated on the wafer surface, ensuring the normal operation of subsequent modules and outputting an image of the wafer's morphology after cleaning.

[0006] Etching is the removal of unwanted material from the surface of a wafer. Different chip structures and functions require different etching depths and profiles. The etching module needs to precisely construct the chip's microstructure to ensure that it removes only the target material without damaging surrounding structures. The etching module achieves accurate simulation and prediction of the wafer etching process through a series of modeling processes. The steps of the etching module include: S101. Obtain input parameters and default parameters. The input parameters include etching gas group selection, etching gas flow rate setting, power, bias voltage, pressure, and time. The default parameter of this module is the gas composition corresponding to the etching gas group. Incorporating all input parameters and default parameters into the etching modeling process ensures that the model can comprehensively consider various factors affecting the etching effect.

[0007] S102. After receiving the input parameters and default parameters, the etching module performs plasma environment modeling. By modeling the plasma environment involved in the etching process, we can gain a deeper understanding of the distribution and characteristics of plasma in the etching environment.

[0008] S103. Based on plasma environment modeling, plasma energy and diffusion modeling is performed on the diffusion behavior of plasma in the environment. The focus is on plasma energy and diffusion process, which helps to analyze how plasma interacts with the wafer surface and how energy is transferred and distributed during the etching process.

[0009] S104. Based on the plasma environment model and plasma energy and diffusion model, the etching behavior is modeled. This model can simulate the actual etching behavior of the wafer during the etching process and predict the etching effect.

[0010] S105. Based on the etching behavior model, output the etching morphology pattern and multiphysics thermal map of the wafer. S106. The output results include two-dimensional or three-dimensional morphological evolution diagrams of the wafer etching process and plasma thermograms under corresponding process parameters, providing engineers with intuitive information on the etching process. This facilitates engineers in analyzing the etching effect, evaluating the rationality of process parameters, and optimizing the process to meet the stringent requirements of semiconductor manufacturing for etching processes.

[0011] The etching module is used to simulate inductively coupled plasma (ICP) and capacitively coupled plasma (CCP) etching processes. By integrating plasma physics formulas and etching behavior models, it achieves high-precision digital modeling and real-time dynamic simulation of the wafer etching process. It also intuitively outputs two-dimensional / three-dimensional wafer morphology evolution diagrams and various plasma physical field thermograms. Furthermore, it introduces an extensible correction term mechanism to compensate for physical phenomena that the standard model may not fully cover in actual complex working conditions, especially in terms of material differences, ion incident angle effects, and electric field / temperature / density non-uniformity. By setting specific correction coefficients or function forms, the software can dynamically adjust etching rate, selectivity, sputtering yield, charge distribution, etc., based on experimental data or typical process characteristics. This significantly improves the engineering consistency and generalization ability of the simulation results without changing the model structure.

[0012] The plasma environment modeling includes: Plasma density Model:

[0013] Among them, plasma density This indicates the number of electrons in a unit volume of plasma. A high electron density means more active particles participate in the etching reaction, thus improving etching efficiency. This represents the current applied to the plasma source. This indicates the frequency of the alternating current applied to the plasma source, and the current... As the plasma source power increases, more energy is input into the plasma, and electrons gain higher energy, thereby increasing the collision frequency and ionization frequency between electrons and gas molecules, and increasing the plasma density. ,frequency Increasing the frequency may shorten the acceleration time of electrons in the electric field, reducing the probability of collisions between electrons and gas molecules, which is detrimental to plasma generation. Increasing the thickness may also reduce the thickness of the plasma sheath, which is beneficial for the acceleration and etching of ions to the wafer surface; Through current With frequency Analysis and prediction of plasma density Distribute and optimize process parameters to achieve the desired etching effect.

[0014] The plasma density model introduces a density correction term.

[0015]

[0016] in, This indicates the corrected plasma density. This represents the plasma density calculated theoretically. By introducing a density correction term, the influence of source aging or power fluctuations on the density distribution is compensated, thereby enhancing the simulation stability.

[0017] plasma frequency Model:

[0018] plasma frequency , representing the natural frequency of the collective oscillation of electrons in the plasma, reflects the rate at which electrons in the plasma return to equilibrium after being disturbed. These are fundamental physical constants that represent the amount of charge on an electron. The fundamental physical constant, representing the vacuum permittivity. This represents the mass of an electron. The lighter the electron, the greater its acceleration and the higher the frequency of its collective oscillation under the same external force. When plasma is disturbed, electrons deviate from their equilibrium positions, resulting in charge separation. Due to the Coulomb force between positive and negative charges, the electrons experience a restoring force, causing them to move back to their equilibrium positions. However, electrons possess inertia, causing them to overtake their equilibrium positions and continue moving in the opposite direction, thus forming a periodic oscillating motion, namely the collective oscillation of electrons. By analyzing the collective oscillation behavior of electrons in plasma, the plasma frequency... The model can quantitatively describe the frequency of this oscillation, helping staff understand the basic laws of collective electron oscillation in plasma, thereby comprehensively predicting plasma state and dynamic characteristics, predicting dynamic characteristics such as plasma oscillation amplitude and phase change, and providing a basis for judging precise etching conditions.

[0019] Debye length Model:

[0020] Describes the scale of charge interaction in plasma, representing the distance over which the electric field energy generated by any charge in the plasma can act, reflecting the charge shielding effect of the plasma. When the spatial scale is greater than the Debye length... At this time, the plasma is considered to be electrically neutral as a whole, its spatial scale is smaller than the Debye length, and charge separation is obvious. It is the dielectric constant, a physical quantity that describes the ability of a medium to respond to an electric field. It is Boltzmann's constant, used to describe the relationship between the thermal energy of microscopic particles and temperature. It is the electron temperature, representing the average thermal energy of electrons in the plasma; By calculating the sheath thickness and local electric field distribution, the electric field environment inside the etching cavity is further simulated in greater detail, thus enhancing the accuracy of the simulation.

[0021] In the model, the sheath thickness d is derived by combining the electrostatic Poisson equation and the quasi-neutral approximation. s :

[0022] in, T is the length of the Debye line. e n e For electron temperature and density, This represents the sheath potential drop.

[0023] Inside the sheath, the electric field increases approximately linearly, using the one-dimensional Poisson-Boltzmann approximation:

[0024] This electric field determines the acceleration behavior of ions near the surface and has a direct impact on the incident angle distribution and etching / deposition directionality.

[0025] The Debye length model introduces an electric field enhancement correction term γ·Φ b :

[0026] in, E eff Indicates the effective electric field strength. It is the electric field enhancement correction factor. This represents the bias voltage, corrected by introducing an electric field enhancement term. This improves the accuracy of etching directionality simulation by correcting the bias voltage's effect on sheath thickness and electric field gradient.

[0027] Poisson equation potential distribution model:

[0028] It is electric potential The Laplace operator describes the second-order rate of change of electric potential in space, reflecting the curvature information of the electric potential field. This represents space charge density, which is the amount of charge per unit volume within the etching cavity. It reflects the density of charge distribution in space. It is the vacuum permittivity, which characterizes the vacuum's ability to "accommodate" an electric field and plays a key role as a medium in the relationship between electric field and charge. By using the Poisson equation potential distribution model, the corresponding potential distribution can be calculated based on the known space charge density distribution. Conversely, if the boundary conditions and some information of the potential are known, the charge distribution can also be inferred. This accurately depicts the charge distribution and spatial potential changes at various locations within the etching cavity, helps adjust and correct the plasma density distribution, improves the accuracy of simulation results, and helps optimize etching process parameters, thereby enhancing the quality and consistency of etched products.

[0029] Poisson equation model introduces a shielding correction term :

[0030] By introducing a shielding correction term, the shielding effect of the cavity material boundary on the spatial potential is reflected, thereby improving the accuracy of plasma modeling.

[0031] The plasma energy and diffusion modeling includes: Boltzmann energy Distribution model:

[0032] This indicates that the particle energy is At that time, the probability density function of particles within the plasma describes the distribution of particles in different energy states. It is a normalization constant used to ensure that the integral of the probability density is 1 over the entire energy range. Temperature is a physical quantity that represents the intensity of plasma thermal motion and is used to measure the degree of intensity of plasma thermal motion. By accurately calculating the energy distribution of particles within the plasma, which is closely related to temperature, a thermogram can visually display the temperature distribution of the plasma at different locations. The calculated particle energy distribution data provides the basic data for thermogram simulation, which can be further converted into the input information required for thermogram simulation. This enables the simulation to more realistically reflect the actual temperature environment, ensuring the accurate simulation and stability of the etching process temperature environment.

[0033] Heat conduction equation model:

[0034] Indicates temperature Regarding time The partial derivative describes the rate of change of plasma temperature with time. The thermal diffusivity is a comprehensive indicator of the thermal conductivity and heat capacity of a plasma. A higher thermal diffusivity indicates that heat travels faster within the plasma. It is temperature The Laplace operator is used to describe the second-order rate of change of temperature in space, that is, whether the temperature is uniformly distributed at each point in space and how it changes. The heat conduction equation model details the spatial and temporal changes in plasma temperature during the etching process, through... It allows for precise understanding of the plasma temperature variation trend at every moment during etching. For example, at the beginning of etching, the temperature may rise rapidly, but as etching progresses, the rate of temperature increase may gradually slow down, or even stabilize or decrease. This model can reflect temperature variations across different locations in space. Within the etching chamber, temperatures may differ at different locations. This model helps determine the shape of the temperature distribution, such as which areas have higher temperatures, which have lower temperatures, and the magnitude and direction of temperature gradients. These can all be quantitatively analyzed using this model. Furthermore, the model supports refined temperature control management and thermal field analysis. Based on a precise description of temperature changes over time and space, more sophisticated temperature control strategies can be developed. For example, the power of heating or cooling devices can be adjusted in real time according to the rate of temperature change to maintain the plasma temperature within a set range, ensuring the stability and consistency of the etching process. This allows for refined temperature control management. Moreover, the model facilitates in-depth analysis of the thermal field distribution during the etching process, thereby analyzing the impact of the thermal field on the etching effect. For instance, uneven temperature can lead to inconsistent etching rates and unsatisfactory etching profiles, providing a theoretical basis for optimizing the etching process.

[0035] The heat conduction equation model and the Boltzmann energy distribution model introduce a correction term T. eff :

[0036] Where f(w) represents the value at particle energy of 0. At that time, the probability density function of particles within the plasma, The equivalent temperature is introduced to reflect the thermal effects of excited-state particles and collisions, thereby improving the realism of thermal field simulation.

[0037] Plasma diffusion model:

[0038] Represents plasma particle density Regarding time The partial derivatives describe the plasma particle density. Over time rate of change, The diffusion coefficient reflects the strength of plasma particle diffusion ability in the etching cavity medium. The larger the diffusion coefficient, the faster the particles diffuse in the medium. It is particle density The Laplace operator describes the second-order rate of change of particle density in space, reflecting the curvature information of the particle density field in space, that is, whether the distribution of particle density at each point in space is uniform and how it changes. This model allows for a detailed understanding of the diffusion of plasma particles at different locations and times within the etching chamber. For example, it can analyze the diffusion process of particles from high-density areas to low-density areas, as well as the changes in particle density distribution during the diffusion process. It provides in-depth analysis of the diffusion behavior and density change trends of plasma particles within the etching chamber. By adjusting parameters such as gas flow rate, pressure, and temperature within the etching chamber, the diffusion and density distribution of particles can be controlled, thereby improving the uniformity, selectivity, and etching rate of the etching process, and ultimately enhancing the quality of the etched products.

[0039] The etching behavior modeling includes: Sputtering output Model:

[0040] This is the sputtering yield, representing the number of target atoms sputtered per incident ion. It reflects the material removal rate; the higher the value, the higher the sputtering efficiency. This refers to the incident particle energy. When the ion energy is low, increasing the energy can significantly improve the sputtering yield. However, once the ion energy exceeds a certain threshold, further increasing the energy will lead to a decrease in sputtering yield. It refers to the incident angle. As the incident angle increases, the ion collision cascades closer to the surface, increasing the probability of atom release. When the angle is too large, the residence time of ions on the surface decreases, leading to a drop in yield. The energy index reflects the degree to which ion energy affects sputtering yield. The value is typically 0.5-2, depending on the target material and the type of ions. The angle index reflects the degree to which the incident angle affects sputtering yield. The value is typically 1-3, depending on the target material and the type of ion. , The results were obtained by fitting experimental data under different material and process conditions. , The values ​​may differ; The sputtering yield model can calculate the sputtering yield under different ion energies and incident angles by correcting the model parameters α and β. This can make the simulation results closer to the actual process conditions, providing a reliable predictive basis for the simulation and optimization of etching processes.

[0041] The sputtering output model introduces a correction term to the angle output function Y(θ):

[0042] in, Y0 represents the sputtering yield at an incident angle of θ. Y0 is perpendicular incident, i.e., the sputtering yield when the incident angle θ = 0. θ is the angle between the incident particle and the normal to the target surface. n is an empirical index used to adjust the degree of influence of the angle on the sputtering yield. The value of n is determined by experimental or theoretical analysis based on the specific sputtering system to accurately describe the variation of sputtering yield with the incident angle.

[0043] Since the incident angle of the etching particles affects the yield, the ability of the model to simulate and reproduce structures such as sidewall etching can be enhanced by introducing a correction term angle yield function.

[0044] Etching rate Model:

[0045] This indicates the etching rate, which is the thickness of wafer material etched per unit time. This indicates the etching depth, which is the actual thickness of material removed during the etching process. Indicates the duration of the etching operation; In the semiconductor manufacturing field, because wafers contain various tiny structures such as transistors and wires, their dimensional accuracy directly affects the performance and reliability of the chip. By monitoring various parameters during the etching process, the etching rate can be adjusted in real time. and time To ensure etching accuracy, for example, if the etching rate is too fast, the etching depth may exceed the expected depth. In this case, the etching rate can be reduced by lowering the radio frequency power or reducing the flow rate of the reaction gas.

[0046] The etching rate model introduces a correction term for the material response coefficient. :

[0047] in, This represents the actual etching rate after considering the material response. This represents the baseline etching rate without considering material differences. By introducing this correction term, the etching response rates of different materials under plasma are corrected, ensuring accurate differentiation of multi-material etching processes.

[0048] Material selection ratio Control Model:

[0049] The selectivity ratio is an indicator used to measure the difference in etching rates between two different materials during the etching process. A higher selectivity ratio indicates a greater difference in etching rates between the target material and the protective layer material. This means the etching of the target material is more precise, and the protective layer material is better protected, resulting in effective etching of the target material. It is the etching rate of the target material. This indicates the etching rate of the protective layer material; for example, if the target etching depth is... The thickness of the protective layer is When the material selection ratio When the size is large enough, even if the etching time is long, the protective layer will not be completely etched away, thus ensuring the accuracy of the final etched structure. In semiconductor manufacturing etching processes, the target materials for etching are typically silicon, silicon oxide, and silicon nitride, while the protective layer materials are usually photoresist masks and adjacent circuit structures. Different etching gases exhibit different etching characteristics for different materials. By adjusting the type, proportion, and flow rate of the etching gases and optimizing the gas combination, the etching properties can be altered. and Thus optimizing the selection ratio For example, when etching silicon, using a mixture of fluorine-containing gas and oxygen may alter the etching rate of silicon and photoresist compared to using a fluorine-containing gas alone. Through experiments and simulations, suitable gas combinations can be found to achieve the desired results. The ideal value has been achieved.

[0050] Material selection ratio model introduces a correction term, selection ratio correction factor. :

[0051] in, Indicates the selection ratio. It is the selection ratio correction factor. Indicates the etching rate of the target material. This indicates the etching rate of the protective layer material. By introducing this correction term, the etching suppression performance of protective layers such as photoresist is corrected, making the prediction of protective layer residue closer to reality.

[0052] The etching module employs an scalable correction term system to compensate for key influencing factors such as material differences, angle effects, and electric field distribution. This design helps enhance the stability and accuracy of the standard model in real-world process scenarios, especially providing more relevant results for typical problems such as multi-material etching and deep trench structure modeling.

[0053] The deposition module includes: a plasma-enhanced chemical vapor deposition (PECVD) module, a low-pressure chemical vapor deposition (LPCVD) module, and a physical vapor deposition (PVD) module; The input parameters for the plasma-enhanced chemical vapor deposition (PECVD) module are: material selection, deposition gas flow rate setting, time, temperature, and pressure. The default parameters for this module are: a list of materials to choose from and the corresponding gas types for each material. The module outputs: two-dimensional or three-dimensional morphological evolution diagrams of the wafer deposition process and plasma thermograms under the corresponding process parameters.

[0054] The input parameters for the low-pressure chemical vapor deposition (LPCVD) module are: material selection, deposition gas flow rate setting, time, temperature, and pressure. The default parameters for this module are: a list of materials to choose from and the corresponding gas types for each material. The module outputs: two-dimensional or three-dimensional morphology evolution diagrams during wafer deposition and flow field diagrams under corresponding process parameters; The input parameters for the physical vapor deposition (PVD) module are: target selection, time, power, and pressure. The default parameter for this module is the target material list. The module outputs: two-dimensional or three-dimensional morphological evolution diagrams of the wafer deposition process and plasma thermograms under process conditions.

[0055] The specific steps of the plasma-enhanced chemical vapor deposition (PECVD) module include: S201. After receiving the input parameters and default parameters, the PECVD module performs plasma behavior modeling. S202. Based on the plasma behavior model, model the particle flux and transport of plasma to determine the particle flux and particle transport characteristics reaching the material surface. S203. Based on the modeling results of particle flux and particle transport characteristics of the flux and transport model, establish a surface reaction and deposition behavior model. S204, the surface reaction and deposition behavior model simulation outputs deposition morphology graphics, showing the morphology and distribution of the deposited material on the wafer surface, and also outputs multiphysics field thermograms, showing the distribution of multiphysics fields during the deposition process.

[0056] Plasma-enhanced chemical vapor deposition (PECVD) simulates the process and combines theories such as plasma physics models, statistical distributions, and transport mechanisms to digitally model and dynamically simulate deposition behavior in real time. The final output is a two-dimensional / three-dimensional wafer deposition morphology and thermal maps of various physical field distributions. In actual PECVD process simulations, relying solely on idealized basic models such as Maxwell's distribution, Langevin's equations, and Boltzmann relations is insufficient to fully cover the variable process conditions and complex material-plasma interactions. Therefore, to improve the physical realism and simulation accuracy of the model, this module introduces a series of correction terms to compensate for various non-ideal factors, including but not limited to material reactivity, structural geometry, local field changes, and particle transport characteristics.

[0057] Specifically, the plasma-enhanced chemical vapor deposition (PECVD) module includes: Electric drift model:

[0058] in, Represents ion mass. This formula represents the change of ion velocity v with time t. It describes the process of ions being accelerated by the electric field in the plasma sheath and helps to predict the ion trajectory and the directionality of the deposited film. After considering collisions, the Langevin equation model is as follows:

[0059] in, It is the amount of charge carried by a charged particle. It describes the damping effect of gas molecule collisions on charged particles. It is the collision frequency, which reflects the number of collisions between charged particles and gas molecules per unit time. It refers to the velocity of charged particles. The magnitude of the velocity affects the momentum exchange during collisions and other factors. It describes the damping effect of gas molecule collisions on charged particles as they are accelerated in an electric field, and can more realistically predict the motion behavior of particles in actual process environments.

[0060] Maxwell's distribution model:

[0061] in, The velocity of the particle is represented by e≈2.71828. Using this formula to calculate the energy distribution of particles in the plasma helps to predict the probability of particles participating in the reaction, which directly affects thin film deposition.

[0062] Maxwell's distribution model introduces an energy distribution correction term. :

[0063] in, This represents the corrected energy distribution function. This represents the energy distribution function based on the Maxwell distribution. This represents the correction factor. Represents particle energy. Represents the energy scale parameter. The term represents the exponential term, describing the distribution of electron energy in the non-thermal tail as a function of energy. The variation of the form, by introducing the influence of non-thermal tail electrons, improves the modeling accuracy of high-energy electrons in excitation and ionization reactions.

[0064] Debye shielding model:

[0065] The effects of sheath thickness and local electric field are calculated to clarify the potential relationship between plasma and substrate, and to accurately simulate the spatial distribution of electric field during deposition.

[0066] Debye shielding model introduces electric field correction term :

[0067] in, This indicates the corrected Debye length. Indicates the original Debye length. This represents the correction factor. Sheath potential, by introducing an electric field correction term Considering the influence of sheath potential on the spatial electric field, especially under high power and fast modulation, it is used to correct the sheath width and field gradient.

[0068] Boltzmann relation model:

[0069] in, This indicates that at an electric potential of electron density at that location, This is the electron density at a reference point, which is typically taken as a location where the potential is 0. Represents the elementary charge. Representing electric potential, describing the potential level at a point in space, this model helps analyze the influence of the plasma spatial potential gradient on deposition uniformity by describing the distribution law of electron density as a function of electric potential.

[0070] Particle flux model:

[0071] Represents particle flux. Indicates the particle number density at The gradient in the direction is used to accurately assess the thickness, growth rate, and uniformity of the deposited film by calculating the net flux rate of the deposited material on the wafer surface.

[0072] The particle flux model introduces a correction term for the material reactivity coefficient η:

[0073] in, This represents the corrected particle flux. The initial particle flux is represented by the material reaction coefficient, which is corrected by introducing a term. It characterizes the actual reactivity of different materials in plasma deposition reactions. For example, Si and SiO2 have different reaction efficiencies in PECVD, and are used for material difference modeling.

[0074] The particle flux model also introduces an angle-dependent function. : Commonly used

[0075] in, This represents the particle flux after taking into account the particle incident angle. The particle flux at perpendicular incidence is represented by an angle-dependent function. To simulate the effect of the angle between the particles and the substrate surface on the deposition efficiency, the shading attenuation is considered when the angle is large, in order to enhance directional control.

[0076] Finally, the PECVD model also introduces an edge deposition correction term. :

[0077] in, This indicates the corrected edge deposition rate. This indicates the deposition rate inside the wafer or in a normal structural region. Indicates the edge deposition correction factor. Indicates edge location, suppresses deposition rate in wafer boundaries or structural edge regions, prevents non-physical sharp corners or accumulation, and improves output image quality.

[0078] Compared to traditional simulation methods that rely solely on static models, the PECVD module introduces a correction term mechanism to express the dynamic changes in the deposition process caused by factors such as temperature, electric field, and material response in a parametric manner. This method enhances the model's adaptability to non-ideal process environments, enabling it to have better generalization performance under complex conditions such as multi-material and multi-gas combinations. It has certain engineering practical value and model expansion potential.

[0079] The specific steps of the physical vapor deposition (PVD) module include: S301. Adjust the dynamic magnetic field according to the input parameters and default parameters of the PVD module; S302. After completing the dynamic magnetic field control, conduct plasma sputtering simulation to obtain the spatial distribution of particles; S303. Calculate the thin film deposition rate and dynamically update the morphology; S304 Output deposition morphology diagrams to visually display the surface morphology of the thin film.

[0080] The Physical Vapor Deposition (PVD) module vaporizes solid targets through heating or sputtering, causing them to condense and form thin films on substrate surfaces. Combining dynamic magnetic field control, plasma simulation, and particle deposition modeling, it achieves high-precision digital simulation, outputting two-dimensional / three-dimensional wafer deposition morphology and plasma physical field thermograms. Furthermore, to further enhance the engineering adaptability and accuracy controllability of PVD process simulation, a series of correction terms are introduced on top of the fundamental physical models, such as magnetic field modeling, sputtering kinetics, and deposition rate formulas. These correction terms express physical effects observed in experiments but not fully covered by standard formulas, such as edge attenuation of the actual magnetic field distribution, the influence of the target bombardment angle on sputtering efficiency, the shielding and loss of deposited particles in micro / nano structures, and flux shifts caused by temperature field or energy attenuation.

[0081] The Physical Vapor Deposition (PVD) module specifically includes: Dynamic magnetic field control model:

[0082] in, Representing three-dimensional coordinates The magnetic field strength at different locations in the target sputtering region affects the trajectory of charged particles and sputtering behavior. This can be achieved by studying the magnetic field strength... Precise control can optimize the distribution of ion sputtering regions. This is the baseline magnetic field strength, representing the magnetic field strength value without the influence of dynamic control factors. This represents the time-varying characteristics of the magnetic field, where, Indicates frequency, By introducing a time-varying term in the form of a sine function to represent time, the magnetic field can be dynamically changed over time, thereby achieving dynamic control of the magnetic field in the sputtering region of the target material. It adjusts the contribution of the space attenuation term to the magnetic field strength. The value of affects the distribution pattern of the magnetic field in space. Is In the middle, the magnetic field strength is determined by... and The decay rate of directional spatial position change, The larger the magnetic field, the faster it decays in regions far from the center.

[0083] This model enables precise control of the magnetic field in the sputtering region of the target material, dynamically optimizes the particle sputtering region distribution, and guides the movement of charged particles through dynamic magnetic field changes. This results in a more reasonable distribution of particles on the target surface, reduces uneven sputtering at the edges, significantly reduces edge effects, and effectively improves the uniformity and quality of thin film deposition. Furthermore, by using finite element simulation tools to simulate the magnetic field distribution, the model predicts the magnetic field strength and spatial distribution, provides an intuitive understanding of the magnetic field characteristics, and offers high-precision dynamic magnetic field visualization analysis to guide the optimization of actual process parameters.

[0084] The dynamic magnetic field control model introduces a marginal magnetic field correction coefficient. :

[0085] Among these improvements, the edge magnetic flux density attenuation is corrected to enhance magnetic field uniformity.

[0086] Plasma sputtering model:

[0087] in, It is the curl operator. It is the magnetic field strength vector. Indicates magnetic flux density The curl of a magnetic field reflects its circulation characteristics in space and describes one aspect of the spatial variation of the magnetic field. The permeability of vacuum is used to characterize magnetic properties in a vacuum. Representing the current density vector, it describes the amount of charge passing through a unit area per unit time, reflecting the flow of charge in the plasma. Indicates electric field strength Regarding time The partial derivative of represents the rate of change of the electric field with time; The fluid dynamics model is based on the fundamental assumption of a continuous medium, treating plasma as a continuous fluid. By solving the partial differential equations of continuity, the spatial distribution of plasma particles is obtained, which is suitable for simulating high-density plasma. By simulating the generation and distribution evolution of high-density plasma, the model analyzes the trajectory of particles under the action of electric and magnetic fields, energy distribution, and collision sputtering process with the target material, accurately reproducing the physical scene of real sputtering deposition.

[0088] The plasma sputtering model introduces an angle-dependent sputtering yield function correction term. :

[0089] in, This indicates the revised sputtering output. It represents the sputtering yield under vertical incidence, expresses the sputtering rate as a function of the incident angle, and improves the accuracy of angle modeling.

[0090] The plasma sputtering model also incorporates a target fatigue correction factor. :

[0091] in, This indicates the corrected effective sputtering yield. Indicates the initial sputtering yield. This indicates that the sputtering efficiency has decreased due to prolonged bombardment of the target material.

[0092] Thin film deposition rate Model:

[0093] Indicates the thin film deposition rate. This represents the flux of incident particles, specifically the number of ions incident on the target surface per unit area per unit time. Indicates the target area. This indicates deposition efficiency, taking into account various factors during sputtering and deposition that cause atomic loss. For example, some sputtered atoms may be redeposited onto the target or fail to reach the substrate due to collisions with surrounding gas molecules. The base area represents the size of the region that can receive sputtered atoms. This indicates the density of the thin film material.

[0094] This formula accurately predicts the rate of film thickness growth per unit time. By precisely setting relevant parameters, the control of the film deposition process can be optimized, improving the quality and performance of the film. Furthermore, molecular dynamics simulations are used to simulate the collision, diffusion, nucleation, and growth of deposited particles on the substrate surface at the atomic scale. This comprehensively reproduces the microscopic deposition morphology and dynamic evolution characteristics of the film, providing a deeper understanding of the microscopic mechanisms of film growth. Ultimately, precise two-dimensional and three-dimensional film deposition morphology data are output, helping researchers to better design and control the structure and performance of films.

[0095] The thin film deposition rate model introduces a correction term, the temperature enhancement factor. :

[0096] in, This indicates the corrected film deposition plastic sheet. Indicates the initial thin film deposition rate. This represents the temperature enhancement factor, which simulates the effect of temperature on chemical reactions and diffusion behavior.

[0097] The thin film deposition rate model also introduces a correction term, the geometric shading function. :

[0098] in, This represents the corrected particle flux distribution function. This represents the uncorrected particle flux distribution function. Represents the geometric occlusion function. The coordinates of the substrate surface indicate the degree to which deposits in the microstructure are obscured by the structure.

[0099] The thin film deposition rate model also introduces a correction term, the flux attenuation coefficient Φ. loss :

[0100] in, This represents the corrected effective particle flux. This represents the initial particle flux. Indicates the flux attenuation coefficient. Indicates the distance the particle travels. The attenuation factor represents the loss of deposited particles due to scattering, energy decay, etc.

[0101] This module introduces parameterized correction terms, enabling the system to flexibly adapt to deposition behavior under different materials, geometries, and process conditions without altering the original main model architecture. This significantly enhances the versatility, engineering fit, and future scalability of the simulation model.

[0102] The input parameters for the glue application module are: glue application time; Default parameters: wafer rotation speed, coating temperature, and dispensing speed; Output: Two-dimensional or three-dimensional topographic images of the wafer coated with photoresist, used to show the coverage and thickness of the photoresist.

[0103] The input parameters of the photolithography module include: mask selection; Default parameters: Mask and chip process association rules; Output: A two-dimensional or three-dimensional graphic structure outline formed by photolithography.

[0104] The input parameters for the cleaning module are: solution type and cleaning time; Default parameters: solution temperature, solution rate, and applicable materials; Output: Two-dimensional or three-dimensional changes in the surface morphology of the wafer after cleaning.

[0105] The aforementioned semiconductor chip virtual production line simulation software method further includes: The photoresist removal module simulates the removal of residual photoresist from the wafer surface and outputs a wafer structure diagram showing the removal of photoresist. The ion implantation module simulates the implantation of ions into a wafer, controls the depth and concentration of doped ions, and outputs a wafer structure diagram after implantation. The rapid thermal processing module is used to simulate the activation of implanted ions and form an oxide layer on the wafer surface, outputting a wafer structure diagram of the rapid thermal annealing process; The thermal oxidation module simulates the formation of a dense insulating layer on the wafer surface through high-temperature oxidation and outputs a wafer topography image during thermal oxidation. The CMP module is used to simulate planarization of the wafer surface and output a topographic image of the polished wafer.

[0106] The input parameter for the adhesive removal module is time; Default parameters: solution concentration, flow rate, wafer rotation speed, and temperature; Output: Two-dimensional or three-dimensional wafer structure diagram after photoresist removal, used to show the exposed areas and remaining patterns; The input parameters of the ion implantation module include: ion type; Default parameters: ion dose, ion energy; Output: Two-dimensional / three-dimensional evolution diagrams of the wafer structure after implantation.

[0107] The input parameters of the ion implantation module include: mode selection, oxide film function setting RTO, subtype diffusion RTA, and time; Default parameters: RTA time and temperature, RTO temperature; Output: Two-dimensional / three-dimensional evolution diagrams of wafer structure during rapid thermal annealing / oxidation processes.

[0108] The input parameters for the thermal oxidation module are: oxide film function settings and time; Default parameter: Temperature; Output: Two-dimensional or three-dimensional evolution diagram of the morphology of the wafer during thermal oxidation; The input parameter of the CMP module is time; Default parameters: pressure, polishing pad speed, and wafer speed; Output: Two-dimensional or three-dimensional evolution diagram of the surface morphology of the polished wafer.

[0109] The beneficial effects of the semiconductor chip virtual production line simulation software method provided by this invention are as follows: This method includes an etching module, a deposition module, a resist coating module, a photolithography module, a cleaning module, a resist removal module, an ion implantation module, a rapid thermal treatment module, a thermal oxidation module, and a CMP module. Relying on digital twin technology, it realistically restores the equipment operating environment, status, and process under different control parameters such as gas pressure, temperature, power, gas composition, and flow rate, making the complex and difficult-to-understand abstract process concrete, and intuitively viewing the process that is difficult to observe in real equipment through animation and pictures; Based on the simulation algorithm engine and database of this application, it provides realistic feedback to the user's operation, and intuitively and realistically shows the changes in the process and final result corresponding to different operating parameters, thereby improving students' higher-order thinking and initiative. Virtual scenarios enhance one's understanding of tasks in the real workplace, and real-time feedback allows for feedback and alerts on the results of erroneous operations, thus avoiding the risks associated with misoperation of real equipment. By setting up an operation panel similar to that of semiconductor factory equipment, students can accumulate practical experience during the learning process and seamlessly connect with the actual production line. Attached Figure Description

[0110] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0111] Figure 1 This is a schematic diagram of the modules of the semiconductor chip virtual production line simulation software method.

[0112] Figure 2 This is a schematic diagram of the etching module steps.

[0113] Figure 3 This is a functional block diagram of the etching module.

[0114] Figure 4 This is a schematic diagram of the process parameters of the etching module.

[0115] Figure 5 It is the initial structural morphology of the sidewall before etching in the CCP mode of the etching module.

[0116] Figure 6 It is the morphological evolution result of the sidewall after etching in the CCP mode of the etching module.

[0117] Figure 7It is the initial structural morphology of the gate before etching in the ICP mode of the etching module.

[0118] Figure 8 It is the morphological evolution result after gate etching in the ICP mode of the etching module.

[0119] Figure 9 This is a three-dimensional distribution diagram of argon ion density in ICP mode of the etching module.

[0120] Figure 10 This is a two-dimensional distribution diagram of argon ion density in ICP mode of the etching module.

[0121] Figure 11 This is a three-dimensional distribution diagram of argon ion density in CCP mode of the etching module.

[0122] Figure 12 This is a two-dimensional distribution diagram of argon ion density in CCP mode of the etching module.

[0123] Figure 13 This is a schematic diagram of the steps of the PECVD module.

[0124] Figure 14 This is a schematic diagram of the process parameters of the PECVD module.

[0125] Figure 15 This is a functional block diagram of the PECVD module.

[0126] Figure 16 This is an initial structural morphology diagram of the nitride layer before deposition in the PECVD module.

[0127] Figure 17 This is a diagram showing the morphological evolution of the nitrided layer after deposition in the PECVD module.

[0128] Figure 18 This is a three-dimensional density distribution diagram of argon ions under the PECVD module.

[0129] Figure 19 This is a two-dimensional density distribution diagram of argon ions under the PECVD module.

[0130] Figure 20 This is a schematic diagram of the steps of the PVD module.

[0131] Figure 21 This is a schematic diagram of the process parameters of the PVD module.

[0132] Figure 22 This is a functional block diagram of the PVD module.

[0133] Figure 23 This is an initial structural morphology diagram of cobalt deposition before deposition in the PVD module.

[0134] Figure 24 This is a diagram showing the morphological evolution of cobalt after deposition using the PVD module.

[0135] Figure 25 This is a magnetic field distribution diagram of the magnetron sputtering module.

[0136] Figure 26 This is a cavity electron density distribution diagram of the PVD module. Detailed Implementation

[0137] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings: Example 1: Refer to Figure 1 like Figure 1 As shown, the present invention provides a semiconductor chip virtual production line simulation software method, comprising: The etching module is used to simulate etching the wafer, remove the material on the wafer surface, and output the etched topography of the wafer. The deposition module is used to simulate the formation of deposited materials on wafers and output a deposition morphology map of the wafers; The photoresist coating module is used to simulate coating photoresist onto the wafer surface and output a coating coverage topography image. The photolithography module is used to simulate transferring a mask pattern onto the wafer surface and output the outline of the pattern structure formed by photolithography. The cleaning module is used to simulate the removal of contaminants generated on the wafer surface, ensuring the normal operation of subsequent modules and outputting an image of the wafer's morphology after cleaning.

[0138] The input parameters for the glue application module are: glue application time; Default parameters: wafer rotation speed, coating temperature, and dispensing speed; Output: Two-dimensional or three-dimensional topographic images of the wafer coated with photoresist, used to show the coverage and thickness of the photoresist.

[0139] The input parameters of the photolithography module include: mask selection; Default parameters: Mask and chip process association rules; Output: A two-dimensional or three-dimensional graphic structure outline formed by photolithography.

[0140] The input parameters for the cleaning module are: solution type and cleaning time; Default parameters: solution temperature, solution rate, and applicable materials; Output: Two-dimensional or three-dimensional changes in the surface morphology of the wafer after cleaning.

[0141] The aforementioned semiconductor chip virtual production line simulation software method further includes: The photoresist removal module simulates the removal of residual photoresist from the wafer surface and outputs a wafer structure diagram showing the removal of photoresist. The ion implantation module simulates the implantation of ions into a wafer, controls the depth and concentration of doped ions, and outputs a wafer structure diagram after implantation. The rapid thermal processing module is used to simulate the activation of implanted ions and form an oxide layer on the wafer surface, outputting a wafer structure diagram of the rapid thermal annealing process; The thermal oxidation module simulates the formation of a dense insulating layer on the wafer surface through high-temperature oxidation and outputs a wafer topography image during thermal oxidation. The CMP module is used to simulate planarization of the wafer surface and output a topographic image of the polished wafer.

[0142] The input parameter for the adhesive removal module is time; Default parameters: solution concentration, flow rate, wafer rotation speed, and temperature; Output: Two-dimensional or three-dimensional wafer structure diagram after photoresist removal, used to show the exposed areas and remaining patterns; The input parameters of the ion implantation module include: ion type; Default parameters: ion dose, ion energy; Output: Two-dimensional / three-dimensional evolution diagrams of the wafer structure after implantation.

[0143] The input parameters of the ion implantation module include: mode selection, oxide film function setting RTO, subtype diffusion RTA, and time; Default parameters: RTA time and temperature, RTO temperature; Output: Two-dimensional / three-dimensional evolution diagrams of wafer structure during rapid thermal annealing / oxidation processes.

[0144] The input parameters for the thermal oxidation module are: oxide film function settings and time; Default parameter: Temperature; Output: Two-dimensional or three-dimensional evolution diagram of the morphology of the wafer during thermal oxidation; The input parameter of the CMP module is time; Default parameters: pressure, polishing pad speed, and wafer speed; Output: Two-dimensional or three-dimensional evolution diagram of the surface morphology of the polished wafer.

[0145] Example 2: Refer to Figures 2-10 Etching is the process of removing unwanted material from the surface of a wafer. Different chip structures and functions require different etching depths and profiles. The etching module needs to precisely construct the chip's microstructure to ensure that it removes only the target material without damaging surrounding structures. This etching module achieves accurate simulation and prediction of the wafer etching process through a series of modeling processes, such as... Figure 2 As shown, the etching module's steps include: S101. Obtain input parameters and default parameters. The input parameters include etching gas group selection, etching gas flow rate setting, power, bias voltage, pressure, and time. The default parameter of this module is the gas composition corresponding to the etching gas group. Incorporating all input parameters and default parameters into the etching modeling process ensures that the model can comprehensively consider various factors affecting the etching effect.

[0146] S102. After receiving the input parameters and default parameters, the etching module performs plasma environment modeling. By modeling the plasma environment involved in the etching process, we can gain a deeper understanding of the distribution and characteristics of plasma in the etching environment.

[0147] S103. Based on plasma environment modeling, plasma energy and diffusion modeling is performed on the diffusion behavior of plasma in the environment. The focus is on plasma energy and diffusion process, which helps to analyze how plasma interacts with the wafer surface and how energy is transferred and distributed during the etching process.

[0148] S104. Based on the plasma environment model and plasma energy and diffusion model, the etching behavior is modeled. This model can simulate the actual etching behavior of the wafer during the etching process and predict the etching effect.

[0149] S105. Based on the etching behavior model, output the etching morphology pattern and multiphysics thermal map of the wafer. S106. The output results include two-dimensional or three-dimensional morphological evolution diagrams of the wafer etching process and plasma thermograms under corresponding process parameters, providing engineers with intuitive information on the etching process. This facilitates engineers in analyzing the etching effect, evaluating the rationality of process parameters, and optimizing the process to meet the stringent requirements of semiconductor manufacturing for etching processes.

[0150] The etching module is used to simulate inductively coupled plasma (ICP) and capacitively coupled plasma (CCP) etching processes. By integrating plasma physics formulas and etching behavior models, it achieves high-precision digital modeling and real-time dynamic simulation of the wafer etching process. It also intuitively outputs two-dimensional / three-dimensional wafer morphology evolution diagrams and various plasma physical field thermograms. Furthermore, it introduces an extensible correction term mechanism to compensate for physical phenomena that the standard model may not fully cover in actual complex working conditions, especially in terms of material differences, ion incident angle effects, and electric field / temperature / density non-uniformity. By setting specific correction coefficients or function forms, the software can dynamically adjust etching rate, selectivity, sputtering yield, charge distribution, etc., based on experimental data or typical process characteristics. This significantly improves the engineering consistency and generalization ability of the simulation results without changing the model structure.

[0151] like Figure 3 As shown, the plasma environment modeling includes: Plasma density Model:

[0152] Among them, plasma density This indicates the number of electrons in a unit volume of plasma. A high electron density means more active particles participate in the etching reaction, thus improving etching efficiency. This represents the current applied to the plasma source. This indicates the frequency of the alternating current applied to the plasma source, and the current... As the plasma source power increases, more energy is input into the plasma, and electrons gain higher energy, thereby increasing the collision frequency and ionization frequency between electrons and gas molecules, and increasing the plasma density. ,frequency Increasing the frequency may shorten the acceleration time of electrons in the electric field, reducing the probability of collisions between electrons and gas molecules, which is detrimental to plasma generation. Increasing the thickness may also reduce the thickness of the plasma sheath, which is beneficial for the acceleration and etching of ions to the wafer surface; Through current With frequency Analysis and prediction of plasma density Distribute and optimize process parameters to achieve the desired etching effect.

[0153] The plasma density model introduces a density correction term.

[0154]

[0155] in, This indicates the corrected plasma density. This represents the plasma density calculated theoretically. By introducing a density correction term, the influence of source aging or power fluctuations on the density distribution is compensated, thereby enhancing the simulation stability.

[0156] plasma frequency Model:

[0157] plasma frequency , representing the natural frequency of the collective oscillation of electrons in the plasma, reflects the rate at which electrons in the plasma return to equilibrium after being disturbed. These are fundamental physical constants that represent the amount of charge on an electron. The fundamental physical constant, representing the vacuum permittivity. This represents the mass of an electron. The lighter the electron, the greater its acceleration and the higher the frequency of its collective oscillation under the same external force. When plasma is disturbed, electrons deviate from their equilibrium positions, resulting in charge separation. Due to the Coulomb force between positive and negative charges, the electrons experience a restoring force, causing them to move back to their equilibrium positions. However, electrons possess inertia, causing them to overtake their equilibrium positions and continue moving in the opposite direction, thus forming a periodic oscillating motion, namely the collective oscillation of electrons. By analyzing the collective oscillation behavior of electrons in plasma, the plasma frequency... The model can quantitatively describe the frequency of this oscillation, helping staff understand the basic laws of collective electron oscillation in plasma, thereby comprehensively predicting plasma state and dynamic characteristics, predicting dynamic characteristics such as plasma oscillation amplitude and phase change, and providing a basis for judging precise etching conditions.

[0158] Debye length Model:

[0159] Describes the scale of charge interaction in plasma, representing the distance over which the electric field energy generated by any charge in the plasma can act, reflecting the charge shielding effect of the plasma. When the spatial scale is greater than the Debye length... At this time, the plasma is considered to be electrically neutral as a whole, its spatial scale is smaller than the Debye length, and charge separation is obvious. It is the dielectric constant, a physical quantity that describes the ability of a medium to respond to an electric field. It is Boltzmann's constant, used to describe the relationship between the thermal energy of microscopic particles and temperature. It is the electron temperature, representing the average thermal energy of electrons in the plasma; By calculating the sheath thickness and local electric field distribution, the electric field environment inside the etching cavity is further simulated in greater detail, thus enhancing the accuracy of the simulation.

[0160] The Debye length model introduces an electric field enhancement correction term γ·Φ b :

[0161] in, E eff Indicates the effective electric field strength. It is the electric field enhancement correction factor. This represents the bias voltage, corrected by introducing an electric field enhancement term. This improves the accuracy of etching directionality simulation by correcting the bias voltage's effect on sheath thickness and electric field gradient.

[0162] Poisson equation potential distribution model:

[0163] It is electric potential The Laplace operator describes the second-order rate of change of electric potential in space, reflecting the curvature information of the electric potential field. This represents space charge density, which is the amount of charge per unit volume within the etching cavity. It reflects the density of charge distribution in space. It is the vacuum permittivity, which characterizes the vacuum's ability to "accommodate" an electric field and plays a key role as a medium in the relationship between electric field and charge. By using the Poisson equation potential distribution model, the corresponding potential distribution can be calculated based on the known space charge density distribution. Conversely, if the boundary conditions and some information of the potential are known, the charge distribution can also be inferred. This accurately depicts the charge distribution and spatial potential changes at various locations within the etching cavity, helps adjust and correct the plasma density distribution, improves the accuracy of simulation results, and helps optimize etching process parameters, thereby enhancing the quality and consistency of etched products.

[0164] Poisson equation model introduces a shielding correction term :

[0165] By introducing a shielding correction term, the shielding effect of the cavity material boundary on the spatial potential is reflected, thereby improving the accuracy of plasma modeling.

[0166] The plasma energy and diffusion modeling includes: Boltzmann energy Distribution model:

[0167] This indicates that the particle energy is At that time, the probability density function of particles within the plasma describes the distribution of particles in different energy states. It is a normalization constant used to ensure that the integral of the probability density is 1 over the entire energy range. Temperature is a physical quantity that represents the intensity of plasma thermal motion and is used to measure the degree of intensity of plasma thermal motion. By accurately calculating the energy distribution of particles within the plasma, which is closely related to temperature, a thermogram can visually display the temperature distribution of the plasma at different locations. The calculated particle energy distribution data provides the basic data for thermogram simulation, which can be further converted into the input information required for thermogram simulation. This enables the simulation to more realistically reflect the actual temperature environment, ensuring the accurate simulation and stability of the etching process temperature environment.

[0168] Heat conduction equation model:

[0169] Indicates temperature Regarding time The partial derivative describes the rate of change of plasma temperature with time. The thermal diffusivity is a comprehensive indicator of the thermal conductivity and heat capacity of a plasma. A higher thermal diffusivity indicates that heat travels faster within the plasma. It is temperature The Laplace operator is used to describe the second-order rate of change of temperature in space, that is, whether the temperature is uniformly distributed at each point in space and how it changes. The heat conduction equation model details the spatial and temporal changes in plasma temperature during the etching process, through... It allows for precise understanding of the plasma temperature variation trend at every moment during etching. For example, at the beginning of etching, the temperature may rise rapidly, but as etching progresses, the rate of temperature increase may gradually slow down, or even stabilize or decrease. This model can reflect temperature variations across different locations in space. Within the etching chamber, temperatures may differ at different locations. This model helps determine the shape of the temperature distribution, such as which areas have higher temperatures, which have lower temperatures, and the magnitude and direction of temperature gradients. These can all be quantitatively analyzed using this model. Furthermore, the model supports refined temperature control management and thermal field analysis. Based on a precise description of temperature changes over time and space, more sophisticated temperature control strategies can be developed. For example, the power of heating or cooling devices can be adjusted in real time according to the rate of temperature change to maintain the plasma temperature within a set range, ensuring the stability and consistency of the etching process. This allows for refined temperature control management. Moreover, the model facilitates in-depth analysis of the thermal field distribution during the etching process, thereby analyzing the impact of the thermal field on the etching effect. For instance, uneven temperature can lead to inconsistent etching rates and unsatisfactory etching profiles, providing a theoretical basis for optimizing the etching process.

[0170] The heat conduction equation model and the Boltzmann energy distribution model introduce correction terms. :

[0171] in, This indicates that the particle energy is At that time, the probability density function of particles within the plasma, The equivalent temperature is introduced to reflect the thermal effects of excited-state particles and collisions, thereby improving the realism of thermal field simulation.

[0172] Plasma diffusion model:

[0173] Represents plasma particle density Regarding time The partial derivatives describe the plasma particle density. Over time rate of change, The diffusion coefficient reflects the strength of plasma particle diffusion ability in the etching cavity medium. The larger the diffusion coefficient, the faster the particles diffuse in the medium. It is particle density The Laplace operator describes the second-order rate of change of particle density in space, reflecting the curvature information of the particle density field in space, that is, whether the distribution of particle density at each point in space is uniform and how it changes. This model allows for a detailed understanding of the diffusion of plasma particles at different locations and times within the etching chamber. For example, it can analyze the diffusion process of particles from high-density areas to low-density areas, as well as the changes in particle density distribution during the diffusion process. It provides in-depth analysis of the diffusion behavior and density change trends of plasma particles within the etching chamber. By adjusting parameters such as gas flow rate, pressure, and temperature within the etching chamber, the diffusion and density distribution of particles can be controlled, thereby improving the uniformity, selectivity, and etching rate of the etching process, and ultimately enhancing the quality of the etched products.

[0174] The etching behavior modeling includes: Sputtering output Model:

[0175] This is the sputtering yield, representing the number of target atoms sputtered per incident ion. It reflects the material removal rate; the higher the value, the higher the sputtering efficiency. This refers to the incident particle energy. When the ion energy is low, increasing the energy can significantly improve the sputtering yield. However, once the ion energy exceeds a certain threshold, further increasing the energy will lead to a decrease in sputtering yield. It refers to the incident angle. As the incident angle increases, the ion collision cascades closer to the surface, increasing the probability of atom release. When the angle is too large, the residence time of ions on the surface decreases, leading to a drop in yield. The energy index reflects the degree to which ion energy affects sputtering yield. The value is typically 0.5-2, depending on the target material and the type of ions. The angle index reflects the degree to which the incident angle affects sputtering yield. The value is typically 1-3, depending on the target material and the type of ion. , The results were obtained by fitting experimental data under different material and process conditions. , The values ​​may differ; The sputtering yield model can calculate the sputtering yield under different ion energies and incident angles by correcting the model parameters α and β. This can make the simulation results closer to the actual process conditions, providing a reliable predictive basis for the simulation and optimization of etching processes.

[0176] Sputtering output model introduces correction term angle output function :

[0177] in, Indicates that at the angle of incidence is Sputtering output at that time When the incident light is perpendicular, that is, the angle of incidence Sputtering output at that time The angle between the incident particle and the normal to the target surface. It is an empirical index used to adjust the degree of influence of the angle on sputtering yield. The value of n is determined by experiment or theoretical analysis based on the specific sputtering system in order to accurately describe the variation of sputtering yield with incident angle.

[0178] Since the incident angle of the etching particles affects the yield, the ability of the model to simulate and reproduce structures such as sidewall etching can be enhanced by introducing a correction term angle yield function.

[0179] Etching rate Model:

[0180] This indicates the etching rate, which is the thickness of wafer material etched per unit time. This indicates the etching depth, which is the actual thickness of material removed during the etching process. Indicates the duration of the etching operation; In the semiconductor manufacturing field, because wafers contain various tiny structures such as transistors and wires, their dimensional accuracy directly affects the performance and reliability of the chip. By monitoring various parameters during the etching process, the etching rate can be adjusted in real time. and time To ensure etching accuracy, for example, if the etching rate is too fast, the etching depth may exceed the expected depth. In this case, the etching rate can be reduced by lowering the radio frequency power or reducing the flow rate of the reaction gas.

[0181] The etching rate model introduces a correction term for the material response coefficient. :

[0182] in, This represents the actual etching rate after considering the material response. This represents the baseline etching rate without considering material differences. By introducing this correction term, the etching response rates of different materials under plasma are corrected, ensuring accurate differentiation of multi-material etching processes.

[0183] Material selection ratio Control Model:

[0184] The selectivity ratio is an indicator used to measure the difference in etching rates between two different materials during the etching process. A higher selectivity ratio indicates a greater difference in etching rates between the target material and the protective layer material. This means the etching of the target material is more precise, and the protective layer material is better protected, resulting in effective etching of the target material. It is the etching rate of the target material. This indicates the etching rate of the protective layer material; for example, if the target etching depth is... The thickness of the protective layer is When the material selection ratio When the size is large enough, even if the etching time is long, the protective layer will not be completely etched away, thus ensuring the accuracy of the final etched structure. In semiconductor manufacturing etching processes, the target materials for etching are typically silicon, silicon oxide, and silicon nitride, while the protective layer materials are usually photoresist masks and adjacent circuit structures. Different etching gases exhibit different etching characteristics for different materials. By adjusting the type, proportion, and flow rate of the etching gases and optimizing the gas combination, the etching properties can be altered. and Thus optimizing the selection ratio For example, when etching silicon, using a mixture of fluorine-containing gas and oxygen may alter the etching rate of silicon and photoresist compared to using a fluorine-containing gas alone. Through experiments and simulations, suitable gas combinations can be found to achieve the desired results. The ideal value has been achieved.

[0185] Material selection ratio model introduces a correction term, selection ratio correction factor. :

[0186] in, Indicates the selection ratio. It is the selection ratio correction factor. Indicates the etching rate of the target material. This indicates the etching rate of the protective layer material. By introducing this correction term, the etching suppression performance of protective layers such as photoresist is corrected, making the prediction of protective layer residue closer to reality.

[0187] The etching module employs an scalable correction term system to compensate for key influencing factors such as material differences, angle effects, and electric field distribution. This design helps enhance the stability and accuracy of the standard model in real-world process scenarios, especially providing more relevant results for typical problems such as multi-material etching and deep trench structure modeling.

[0188] Figure 4 This is a schematic diagram of the process parameters of the etching module. The operator can set the types of process parameters through this interface.

[0189] like Figure 5 The image shows the initial morphology of the sidewalls before etching begins in CCP mode of the etching module. The units for both the horizontal and vertical axes are in nm. Figure 5 The intermediate structure is a typical initial morphology before etching, with silicon oxide and silicon nitride layers stacked on the substrate, providing a material basis for the subsequent formation of the gate sidewalls.

[0190] Figure 6 This is the morphological evolution result of the sidewall etching after etching in CCP mode of the etching module, corresponding to Figure 5 The evolution morphology of the structure after sidewall etching under CCP process conditions is shown in the figure. It can be seen that under the action of ion directional etching, silicon nitride and silicon oxide materials are removed anisotropically, forming sidewall structures with typical geometric features on both sides of the gate. This result shows that the constructed model can accurately reflect the material selectivity, directionality and morphological evolution behavior during CCP etching.

[0191] Figure 7 This is the initial structural morphology of the gate before etching in the ICP mode of the etching module. It shows the initial morphology of the gate structure before the start of the ICP mode etching process. The structure includes a photomask, gate material, such as polysilicon, and the underlying gate dielectric layer, which provides the basis for the subsequent formation of the gate morphology.

[0192] like Figure 8 The figure shows the morphological evolution of the gate after etching in the ICP mode of the etching module. Figure 8 This demonstrates the effect under ICP etching conditions. Figure 7 As the initial morphology, the morphology of the gate material after anisotropic etching is shown. During the etching process, the polysilicon is directionally removed, while the mask layer remains intact, ultimately forming a vertical gate structure with clear boundaries. This verifies the model's ability to simulate etching behavior under ICP high-density plasma conditions.

[0193] Figure 9 This is a three-dimensional distribution map of argon ion density in ICP mode of the etching module. This map shows the three-dimensional density cloud distribution of argon ions inside the ICP chamber. Colors represent the relative density of argon ions in space, with high density in the central region and low density at the edges. This map helps to intuitively understand the relationship between the discharge coupling region and the etching region.

[0194] Figure 10 This is a two-dimensional distribution diagram of argon ion density in ICP mode of the etching module. The figure shows the spatial distribution of argon ions in the cross-section of the ICP high-density plasma reactor. The red area is the region with the highest ion density, which is mainly concentrated below the induction coil and corresponds to the region with the largest high-frequency electric field. This demonstrates the excellent plasma uniformity and deep trench etching adaptability of the ICP system.

[0195] like Figure 11The figure shown is a three-dimensional distribution diagram of argon ion density in CCP mode of the etching module, which shows the three-dimensional structure of argon ion distribution in the entire cavity space. It can be observed that the density decreases from the discharge center to the edge. This result is used to analyze the influence of ion source directionality and sidewall etching flux.

[0196] Figure 12 This is a two-dimensional distribution diagram of argon ion density in CCP mode of the etching module. This figure shows the Ar⁺ ion density distribution in the cross section of the etching reaction chamber under CCP discharge conditions. The red area represents the ion density concentration area, and the blue-purple area represents the low density area at the edge, reflecting the non-uniform distribution characteristics of argon plasma between electrodes in the CCP process.

[0197] Example 3: Reference Figures 13-19 The deposition module includes: a plasma-enhanced chemical vapor deposition (PECVD) module, a low-pressure chemical vapor deposition (LPCVD) module, and a physical vapor deposition (PVD) module; The input parameters for the plasma-enhanced chemical vapor deposition (PECVD) module are: material selection, deposition gas flow rate setting, time, temperature, and pressure. The default parameters for this module are: a list of materials to choose from and the corresponding gas types for each material. The module outputs: two-dimensional or three-dimensional morphological evolution diagrams of the wafer deposition process and plasma thermograms under the corresponding process parameters.

[0198] The input parameters for the low-pressure chemical vapor deposition (LPCVD) module are: material selection, deposition gas flow rate setting, time, temperature, and pressure. The default parameters for this module are: a list of materials to choose from and the corresponding gas types for each material. The module outputs: two-dimensional or three-dimensional morphology evolution diagrams during wafer deposition and flow field diagrams under corresponding process parameters; The input parameters for the physical vapor deposition (PVD) module are: target selection, time, power, and pressure. The default parameter for this module is the target material list. The module outputs: two-dimensional or three-dimensional morphological evolution diagrams of the wafer deposition process and plasma thermograms under process conditions.

[0199] like Figure 13 As shown, the specific steps of the plasma-enhanced chemical vapor deposition (PECVD) module include: S201. After receiving the input parameters and default parameters, the PECVD module performs plasma behavior modeling. S202. Based on the plasma behavior model, model the particle flux and transport of plasma to determine the particle flux and particle transport characteristics reaching the material surface. S203. Based on the modeling results of particle flux and particle transport characteristics of the flux and transport model, establish a surface reaction and deposition behavior model. S204, the surface reaction and deposition behavior model simulation outputs deposition morphology graphics, showing the morphology and distribution of the deposited material on the wafer surface, and also outputs multiphysics field thermograms, showing the distribution of multiphysics fields during the deposition process.

[0200] Plasma-enhanced chemical vapor deposition (PECVD) simulates the process and combines theories such as plasma physics models, statistical distributions, and transport mechanisms to digitally model and dynamically simulate deposition behavior in real time. The final output is a two-dimensional / three-dimensional wafer deposition morphology and thermal maps of various physical field distributions. In actual PECVD process simulations, relying solely on idealized basic models such as Maxwell's distribution, Langevin's equations, and Boltzmann relations is insufficient to fully cover the variable process conditions and complex material-plasma interactions. Therefore, to improve the physical realism and simulation accuracy of the model, this module introduces a series of correction terms to compensate for various non-ideal factors, including but not limited to material reactivity, structural geometry, local field changes, and particle transport characteristics.

[0201] Figure 14 This is a schematic diagram of the process parameters of the PECVD module. Operators can use this interface to set the types of process parameters.

[0202] like Figure 15 As shown, the plasma-enhanced chemical vapor deposition (PECVD) module specifically includes: Electric drift model:

[0203] in, Represents ion mass. Represents ion velocity Over time The formula describes the process of ions being accelerated by the electric field within the plasma sheath, which helps predict the ion trajectory and the directionality of the deposited film. After considering collisions, the Langevin equation model is as follows:

[0204] Where q is the amount of charge carried by the charged particle. It describes the damping effect of gas molecule collisions on charged particles. It is the collision frequency, which reflects the number of collisions between charged particles and gas molecules per unit time. It refers to the velocity of charged particles. The magnitude of the velocity affects the momentum exchange during collisions and other factors. It describes the damping effect of gas molecule collisions on charged particles as they are accelerated in an electric field, and can more realistically predict the motion behavior of particles in actual process environments.

[0205] Maxwell's distribution model:

[0206] in, The velocity of the particle is represented by e≈2.71828. Using this formula to calculate the energy distribution of particles in the plasma helps to predict the probability of particles participating in the reaction, which directly affects thin film deposition.

[0207] Maxwell's distribution model introduces an energy distribution correction term. :

[0208] in, This represents the corrected energy distribution function. This represents the energy distribution function based on the Maxwell distribution. This represents the correction factor. ΔE represents the particle energy, and ΔE represents the energy scale parameter. The term represents the exponential term, describing the distribution of electron energy in the non-thermal tail as a function of energy. The variation of the form, by introducing the influence of non-thermal tail electrons, improves the modeling accuracy of high-energy electrons in excitation and ionization reactions.

[0209] Debye shielding model:

[0210] The effects of sheath thickness and local electric field are calculated to clarify the potential relationship between plasma and substrate, and to accurately simulate the spatial distribution of electric field during deposition.

[0211] Debye shielding model introduces electric field correction term :

[0212] in, This indicates the corrected Debye length. Indicates the original Debye length. This represents the correction factor. Sheath potential, by introducing an electric field correction term Considering the influence of sheath potential on the spatial electric field, especially under high power and fast modulation, it is used to correct the sheath width and field gradient.

[0213] Boltzmann relation model:

[0214] in, This indicates that at an electric potential of electron density at that location, This is the electron density at a reference point, which is typically taken as a location where the potential is 0. Represents the elementary charge. Representing electric potential, describing the potential level at a point in space, this model helps analyze the influence of the plasma spatial potential gradient on deposition uniformity by describing the distribution law of electron density as a function of electric potential.

[0215] Particle flux model:

[0216] nu represents particle flux. Indicates the particle number density at The gradient in the direction is used to accurately assess the thickness, growth rate, and uniformity of the deposited film by calculating the net flux rate of the deposited material on the wafer surface.

[0217] The particle flux model introduces a correction term for the material reactivity coefficient η:

[0218] in, This represents the corrected particle flux. The initial particle flux is represented by the material reaction coefficient η, which is introduced as a correction term to characterize the actual reactivity of different materials in plasma deposition reactions. For example, Si and SiO2 have different reaction efficiencies in PECVD, and this is used for material difference modeling.

[0219] The particle flux model also introduces an angle-dependent function. : Commonly used

[0220] in, This represents the particle flux after taking into account the particle incident angle. The particle flux at perpendicular incidence is represented by an angle-dependent function. To simulate the effect of the angle between the particles and the substrate surface on the deposition efficiency, the shading attenuation is considered when the angle is large, in order to enhance directional control.

[0221] Finally, the PECVD model also introduces an edge deposition correction term. :

[0222] in, This indicates the corrected edge deposition rate. This indicates the deposition rate inside the wafer or in a normal structural region. Indicates the edge deposition correction factor. Indicates edge location, suppresses deposition rate in wafer boundaries or structural edge regions, prevents non-physical sharp corners or accumulation, and improves output image quality.

[0223] Compared to traditional simulation methods that rely solely on static models, the PECVD module introduces a correction term mechanism to express the dynamic changes in the deposition process caused by factors such as temperature, electric field, and material response in a parametric manner. This method enhances the model's adaptability to non-ideal process environments, enabling it to have better generalization performance under complex conditions such as multi-material and multi-gas combinations. It has certain engineering practical value and model expansion potential.

[0224] Figure 16 This is an initial structural morphology diagram of the PECVD module before nitride layer deposition, showing the initial morphology of the sidewall region of the gate structure before the PECVD process. The silicon nitride protective layer has not yet formed in the structure, serving as a simulation of the initial state of nitride layer deposition evolution.

[0225] Figure 17 This is a diagram showing the morphological evolution of the nitride layer after deposition in the PECVD module. Figure 17 Demonstrates PECVD mode, Figure 16 As the initial morphology, the structural morphology after silicon nitride deposition shows that the nitride layer achieves a continuous and uniform coverage effect in the sidewall region, reflecting the good adaptability of this process to complex microstructure surfaces.

[0226] Figure 18 This is a three-dimensional density distribution map of argon ions in the PECVD module, which shows the three-dimensional distribution characteristics of argon ion density in the reaction space of the PECVD module. The spatial distribution law of ions in the cavity is presented in the form of isosurface or volume rendering, which helps to evaluate the deposition flux intensity in different regions and analyze the deposition uniformity and structural adaptability.

[0227] Figure 19 This is a two-dimensional density distribution map of argon ions under the PECVD module, showing the two-dimensional argon ion density distribution in the reaction region during the PECVD process. The color changes in the figure represent the ion density intensity at different locations. The red area is the ion density concentration area, and the blue-purple area is the low density area. This figure is used to analyze the non-uniformity of flux distribution in the deposition region and the influence of cavity structure on plasma behavior.

[0228] Example 4: Reference Figures 20-26 like Figure 20As shown, the specific steps of the physical vapor deposition (PVD) module include: S301. Adjust the dynamic magnetic field according to the input parameters and default parameters of the PVD module; S302. After completing the dynamic magnetic field control, conduct plasma sputtering simulation to obtain the spatial distribution of particles; S303. Calculate the thin film deposition rate and dynamically update the morphology; S304 Output deposition morphology diagrams to visually display the surface morphology of the thin film.

[0229] The Physical Vapor Deposition (PVD) module vaporizes solid targets through heating or sputtering, causing them to condense and form thin films on substrate surfaces. Combining dynamic magnetic field control, plasma simulation, and particle deposition modeling, it achieves high-precision digital simulation, outputting two-dimensional / three-dimensional wafer deposition morphology and plasma physical field thermograms. Furthermore, to further enhance the engineering adaptability and accuracy controllability of PVD process simulation, a series of correction terms are introduced on top of the fundamental physical models, such as magnetic field modeling, sputtering kinetics, and deposition rate formulas. These correction terms express physical effects observed in experiments but not fully covered by standard formulas, such as edge attenuation of the actual magnetic field distribution, the influence of the target bombardment angle on sputtering efficiency, the shielding and loss of deposited particles in micro / nano structures, and flux shifts caused by temperature field or energy attenuation.

[0230] Figure 21 This is a schematic diagram of the process parameters of the etching module. The operator can set the types of process parameters through this interface.

[0231] like Figure 22 As shown, the Physical Vapor Deposition (PVD) module specifically includes: Dynamic magnetic field control model:

[0232] in, Representing three-dimensional coordinates The magnetic field strength at different locations in the target sputtering region affects the trajectory of charged particles and sputtering behavior. This can be achieved by studying the magnetic field strength... Precise control can optimize the distribution of ion sputtering regions. This is the baseline magnetic field strength, representing the magnetic field strength value without the influence of dynamic control factors. This represents the time-varying characteristics of the magnetic field, where, Indicates frequency, By introducing a time-varying term in the form of a sine function to represent time, the magnetic field can be dynamically changed over time, thereby achieving dynamic control of the magnetic field in the sputtering region of the target material. It adjusts the contribution of the space attenuation term to the magnetic field strength. The value of affects the distribution pattern of the magnetic field in space. Is In the middle, the magnetic field strength is determined by... and The decay rate of directional spatial position change, The larger the magnetic field, the faster it decays in regions far from the center.

[0233] This model enables precise control of the magnetic field in the sputtering region of the target material, dynamically optimizes the distribution of particles in the sputtering region, and guides the movement of charged particles through dynamic magnetic field changes. This makes the distribution of particles in the sputtering region on the target surface more reasonable, reduces uneven sputtering of particles in the edge region, significantly reduces edge effects, and effectively improves the uniformity and quality of thin film deposition. Furthermore, by using finite element simulation tools to simulate the magnetic field distribution, the model predicts the magnetic field strength and spatial distribution, provides an intuitive understanding of the magnetic field characteristics, and offers high-precision dynamic magnetic field visualization analysis to guide the optimization of actual process parameters.

[0234] The dynamic magnetic field control model introduces a marginal magnetic field correction coefficient. :

[0235] Among these improvements, the edge magnetic flux density attenuation is corrected to enhance magnetic field uniformity.

[0236] Plasma sputtering model:

[0237] in, It is the curl operator. It is the magnetic field strength vector. Indicates magnetic flux density The curl of a magnetic field reflects its circulation characteristics in space and describes one aspect of the spatial variation of the magnetic field. The permeability of vacuum is used to characterize magnetic properties in a vacuum. Representing the current density vector, it describes the amount of charge passing through a unit area per unit time, reflecting the flow of charge in the plasma. Indicates electric field strength Regarding time The partial derivative of represents the rate of change of the electric field with time; The fluid dynamics model is based on the fundamental assumption of a continuous medium, treating plasma as a continuous fluid. By solving the partial differential equations of continuity, the spatial distribution of plasma particles is obtained, which is suitable for simulating high-density plasma. By simulating the generation and distribution evolution of high-density plasma, the model analyzes the trajectory of particles under the action of electric and magnetic fields, energy distribution, and collision sputtering process with the target material, accurately reproducing the physical scene of real sputtering deposition.

[0238] The plasma sputtering model introduces an angle-dependent sputtering yield function correction term. :

[0239] in, This indicates the revised sputtering output. It represents the sputtering yield under vertical incidence, expresses the sputtering rate as a function of the incident angle, and improves the accuracy of angle modeling.

[0240] The plasma sputtering model also incorporates a target fatigue correction factor. :

[0241] in, This indicates the corrected effective sputtering yield. Indicates the initial sputtering yield. This indicates that the sputtering efficiency has decreased due to prolonged bombardment of the target material.

[0242] Thin film deposition rate Model:

[0243] Indicates the thin film deposition rate. This represents the flux of incident particles, specifically the number of ions incident on the target surface per unit area per unit time. Indicates the target area. This indicates deposition efficiency, taking into account various factors during sputtering and deposition that cause atomic loss. For example, some sputtered atoms may be redeposited onto the target or fail to reach the substrate due to collisions with surrounding gas molecules. The base area represents the size of the region that can receive sputtered atoms. This indicates the density of the thin film material.

[0244] This formula accurately predicts the rate of film thickness growth per unit time. By precisely setting relevant parameters, the control of the film deposition process can be optimized, improving the quality and performance of the film. Furthermore, molecular dynamics simulations are used to simulate the collision, diffusion, nucleation, and growth of deposited particles on the substrate surface at the atomic scale. This comprehensively reproduces the microscopic deposition morphology and dynamic evolution characteristics of the film, providing a deeper understanding of the microscopic mechanisms of film growth. Ultimately, precise two-dimensional and three-dimensional film deposition morphology data are output, helping researchers to better design and control the structure and performance of films.

[0245] The thin film deposition rate model introduces a correction term, the temperature enhancement factor. :

[0246] in, This indicates the corrected film deposition plastic sheet. Indicates the initial thin film deposition rate. This represents the temperature enhancement factor, which simulates the effect of temperature on chemical reactions and diffusion behavior.

[0247] The thin film deposition rate model also introduces a correction term, the geometric shading function. :

[0248] in, This represents the corrected particle flux distribution function. This represents the uncorrected particle flux distribution function. Represents the geometric occlusion function. The coordinates of the substrate surface indicate the degree to which deposits in the microstructure are obscured by the structure.

[0249] The thin film deposition rate model also introduces a correction term, the flux attenuation coefficient Φ. loss :

[0250] in, This represents the corrected effective particle flux. This represents the initial particle flux. Indicates the flux attenuation coefficient. Indicates the distance the particle travels. The attenuation factor represents the loss of deposited particles due to scattering, energy decay, etc.

[0251] This module introduces parameterized correction terms, enabling the system to flexibly adapt to deposition behavior under different materials, geometries, and process conditions without altering the original main model architecture. This significantly enhances the versatility, engineering fit, and future scalability of the simulation model.

[0252] Figure 23 This is an initial structural morphology diagram before cobalt deposition in the PVD module. It shows the initial morphology of the structure to be deposited before the PVD process, with no metal coverage on the surface, used to simulate the initial state of the cobalt deposition process.

[0253] Figure 24 This is a diagram showing the morphological evolution of cobalt deposition under the PVD module, illustrating the morphological evolution of cobalt under the PVD mode. Figure 23 The image shows the initial morphology of the structure after cobalt deposition. As can be seen, the cobalt film has been continuously deposited across the entire surface of the structure, forming a uniformly covered metal layer, indicating that the process possesses strong overall coating capability under directional control conditions.

[0254] Figure 25 This is a magnetic field distribution diagram of the magnetron sputtering module, showing the magnetic field distribution above the magnetron sputtering target area in the PVD module. The colors or vector directions in the diagram reflect the magnetic field strength and direction, which are used to constrain the electron motion path and enhance the discharge efficiency. This diagram reflects the influence of the magnetic field configuration on the plasma density distribution and sputtering uniformity.

[0255] Figure 26 This is a cavity electron density distribution diagram of the PVD module, which shows the spatial distribution of electron density during the magnetron sputtering process. It shows the concentration state of plasma near the target area. Regions with high electron density correspond to regions with high sputtering activity and strong deposition flux. This diagram helps to analyze the discharge uniformity and film thickness distribution.

Claims

1. A method for semiconductor chip virtual production line simulation software, characterized by, The method comprises the following steps of: The etching module comprises the following steps of: S101, input parameters and default parameters are obtained, wherein the input parameters comprise etching gas group selection, etching gas flow setting, power, bias, pressure and time, and the default parameters of the module are gas components corresponding to the etching gas group; S102, after the etching module receives the input parameters and the default parameters, a plasma environment model is established; S103, based on the plasma environment modeling, plasma energy and diffusion modeling of the diffusion behavior of the plasma in the environment is performed; S104, according to the plasma environment model and the plasma energy and diffusion model, etching behavior modeling is performed; 2. The semiconductor chip virtual production line emulation software method of claim 1, wherein, S105, according to the etching behavior model, an etching topography of the wafer and a multi-physical field thermodynamic diagram are outputted; S106, the output results comprise a two-dimensional or three-dimensional topography evolution diagram of the wafer etching process and a plasma thermodynamic diagram under corresponding process parameters. The deposition module comprises a plasma enhanced chemical vapor deposition (PECVD) module, a low pressure chemical vapor deposition (LPCVD) module and a physical vapor deposition (PVD) module; The input parameters of the plasma enhanced chemical vapor deposition (PECVD) module comprise material selection, deposition gas flow setting, time, temperature and pressure; The default parameters of the module comprise a material selection list and gas types corresponding to the material; The output results of the module comprise a two-dimensional or three-dimensional topography evolution diagram of the wafer deposition process and a plasma thermodynamic diagram under corresponding process parameters. The input parameters of the low pressure chemical vapor deposition (LPCVD) module comprise material selection, deposition gas flow setting, time, temperature and pressure; 3. The method of claim 1, wherein the method further comprises: The default parameters of the module comprise a material selection list and gas types corresponding to the material; The output results of the module comprise a two-dimensional or three-dimensional topography evolution diagram of the wafer deposition process and a flow field diagram under corresponding process parameters; The input parameters of the physical vapor deposition (PVD) module comprise target material selection, time, power and pressure; The default parameters of the module comprise a target material list; The output results of the module comprise a two-dimensional or three-dimensional topography evolution diagram of the wafer deposition process and a plasma thermodynamic diagram under process conditions. The specific steps of the plasma enhanced chemical vapor deposition (PECVD) module comprise the following steps of: S201, after the PECVD module receives the input parameters and the default parameters, plasma behavior modeling is performed; S202, according to the plasma behavior model, particle flux and transmission of the plasma are modeled to determine particle flux and particle transmission characteristics reaching the material surface; ​ ​ 4. The method of claim 3, wherein the software is configured to: ​ ​ ​ S203, according to the particle flux and particle transport characteristics of the flux and transport model, a surface reaction and deposition behavior model is established; S204, the surface reaction and deposition behavior model simulates and outputs a deposition morphology diagram, and displays the morphology and distribution of the deposited material on the wafer surface, and outputs a multi-physical field thermal diagram to show the distribution of the multi-physical field in the deposition process.

5. The method of claim 3, wherein the software is configured to: The physical vapor deposition (PVD) module includes the following steps: S301, according to the input parameters and default parameters of the PVD module, the dynamic magnetic field is regulated; S302, after the dynamic magnetic field regulation is completed, plasma sputtering simulation is carried out to obtain the spatial distribution of particles; S303, the film deposition rate is calculated, and the morphology is dynamically updated; S304, output the deposition morphology diagram to intuitively display the surface morphology of the film.

6. The method of claim 1, wherein the method further comprises: The input parameters of the glue coating module are: glue coating time; Default parameters: wafer rotation speed, glue coating temperature and glue dropping speed; Output result: two-dimensional or three-dimensional coating coverage morphology diagram of the wafer coated with photoresist, used to reflect the coverage range and thickness of the photoresist.

7. The method of claim 1, wherein the method further comprises: The input parameters of the photoetching module are: mask selection; Default parameters: mask and chip process association rules; Output result: two-dimensional or three-dimensional graph structure profile diagram formed by photoetching.

8. The method of claim 1, wherein the method further comprises: The input parameters of the cleaning module are: solution type and cleaning time; Default parameters: solution temperature, solution rate and applicable material; Output result: two-dimensional or three-dimensional change diagram of the wafer surface morphology after cleaning.

9. The method of claim 1, wherein the method further comprises: determining a number of dies on the semiconductor wafer; and determining a number of dies on the semiconductor wafer that are defective. Further comprising: A photoresist removal module for simulating the removal of residual photoresist on the wafer surface and outputting a wafer structure diagram after photoresist removal; An ion implantation module for simulating the implantation of ions into the wafer and controlling the depth and concentration of the implanted ions, and outputting a wafer structure diagram after implantation; A rapid thermal processing module for simulating the activation of implanted ions and forming an oxide layer on the wafer surface, and outputting a wafer structure diagram of the rapid thermal annealing process; A thermal oxidation module for simulating the generation of a dense insulating layer on the wafer surface by high-temperature oxidation, and outputting a wafer morphology diagram during thermal oxidation; A CMP module for simulating the planarization of the wafer surface, and outputting a wafer morphology diagram after polishing.

10. The method of claim 9, wherein the method further comprises: The input parameters of the photoresist removal module include: time; Default parameters: solution concentration, flow rate, wafer rotation speed and temperature; Output result: two-dimensional or three-dimensional wafer structure diagram after photoresist removal, used to display the exposed area and remaining pattern; The input parameters of the ion implantation module include: ion species; Default parameters: ion dose, ion energy; Output result: two-dimensional / three-dimensional evolution diagram of the wafer structure after implantation. The input parameters of the rapid thermal processing module include: mode selection, oxide film function setting RTO, sub-type diffusion RTA and time; Default parameters: RTA time and temperature, RTO temperature; Output result: two-dimensional / three-dimensional evolution diagram of the wafer structure in the rapid thermal annealing / oxidation process. The input parameters of the thermal oxidation module include: oxide film function setting and time; Default parameters: temperature; Output result: two-dimensional or three-dimensional evolution diagram of the wafer morphology during thermal oxidation; The input parameters of the CMP module include: time; Default parameters: pressure, polishing pad rotation speed and wafer rotation speed; Output result: two-dimensional or three-dimensional evolution diagram of the surface topography of the polished wafer.

Images