JP4449472B2 - A simulation system or method using a hybrid reaction model. - Google Patents

Description

  The present invention relates to a simulation system or a simulation method for a phenomenon occurring at a gas phase-solid phase interface accompanied by a chemical reaction, such as thin film formation from a gas phase to a solid phase interface by plasma CVD or the like, or dry etching on a solid surface. In particular, the present invention relates to a simulation system or the like that can be applied when there is no or insufficient elementary reaction data relating to a raw material gas for forming a thin film or an etchant removed from a solid surface.

As a method of forming a thin film having a predetermined chemical composition different from that of the source gas through a chemical reaction using a source gas and a reactive gas, thermal CVD (Chemical Vapor Deposition) or plasma is used. Chemical vapor deposition methods such as CVD are known. According to this method, for example, when silane gas (SiH ₄ ) is used as the source gas and oxygen gas is used as the reactive gas, the SiO ₂ thin film can be formed on the silicon surface. Therefore, it plays an important role in creating various devices such as semiconductors and liquid crystal display devices.

  Such a reaction process for forming a thin film is roughly divided into a reaction process in which various chemical species are generated from a raw material gas in the gas phase and a reaction process by chemical bonding or the like on the surface of the solid phase. Each reaction process is said to be very complicated. For example, a commonly used silane gas is used as a source gas, an oxygen gas is used as a reactive gas, an argon gas is used as a carrier gas, and plasma CVD is used. When thin film formation is performed, it is said that 167 types of reactions occur in the gas phase and 96 types of reactions occur on the solid surface.

  With the progress of technological development, there is an increasing demand for using gases other than silane, such as tetraethoxysilane (hereinafter referred to as TEOS), hexamethyldisiloxane (hereinafter referred to as HMDSO), or the like, as a raw material gas. Further, when designing a new CVD apparatus or the like corresponding to such various source gases, there is a great demand for simulation of thin film formation in advance with respect to design elements such as electrode shape and arrangement. However, in order to perform the simulation, a reaction formula generated on the gas phase and the solid surface and a rate coefficient (or reaction rate constant or collision cross section) representing the probability of occurrence thereof are essential. However, the reaction process using source gases other than the above-mentioned silane is hardly known, and there is almost no data available for simulation.

The thin film formation was simulated by using TEOS as the source gas, oxygen as the reactive gas, and one reaction formula with oxygen atoms (O ( ³ P)) as the reaction involving TEOS. There are examples (for example, see Non-Patent Documents 1 and 2).

The same situation occurs in the case of performing dry etching or the like that cuts the solid surface using a corrosive gas, as opposed to the thin film formation. In this case, little is known about the specific reaction process in which the etchant desorbs from the solid surface into the gas phase. Therefore, it has been difficult to perform an accurate simulation of etching.
GBRaupp et al .: J. Vac. Sci. Technol. B (10), pp. 37-45 1992 K. Ishimaru et al .: Thermal Sci. Eng. Vol. 7 No. 5, p11, 1999

  In the present invention, when forming a thin film by CVD using various source gases, or when performing dry etching or the like that scrapes the solid surface using a corrosive gas, the source gas or etchant is involved in a gas phase or a solid phase. It is an object of the present invention to provide a system or method that enables simulation even when there is no or insufficient data regarding the reaction process in the phase.

The present invention has a gas phase, a solid phase, and a solid phase surface that is an interface between the two phases, and the gas phase is desorbed from the solid phase surface by a chemical reaction or the solid phase surface by a chemical reaction. A material gas capable of forming a gas, an intermediate gas between the material gas and a substance constituting the solid phase surface, and a reactive gas capable of reacting with the material gas or the solid phase surface, A simulation system for a system in which mass transfer accompanied by a chemical reaction occurs between the solid phase surface, an elementary reaction database storing known reaction data relating to elementary reactions of the reactive gas, the material gas and the intermediate and a simple model database storing response data two to four reaction process and a gas, a simulation is performed using the above said elementary reaction database simple model database Wherein with respect to the material gas containing reaction involving a simulation system, which comprises using the two to four kinds of simple model assuming only reaction process to.

  Here, it is preferable that a determination unit for determining the validity of the simulation result is further provided, and when the result is not valid, the calculation unit corrects the reaction data and repeats the simulation. Moreover, it is preferable that the said intermediate body is 1 type-3 types. Further, it is preferable that the chemical reaction is performed by plasma and the mass transfer occurs in the direction of forming a thin film on the solid phase surface. The gas phase pressure is preferably substantially the same as atmospheric pressure. Further, the reactive gas may be corrosive to the solid surface, and the mass transfer may occur in the direction of scraping the solid surface.

A second aspect of the invention has a gas phase, a solid phase, and a solid phase surface that is an interface between the two phases, and the gas phase is desorbed from the solid phase surface by a chemical reaction or the solid phase by a chemical reaction. A material gas capable of forming a phase surface, an intermediate gas between the material gas and a substance constituting the solid phase surface, and a reactive gas capable of reacting with the material gas or the solid phase surface. A simulation method for a system in which mass transfer accompanied by a chemical reaction occurs between a phase and the solid phase surface, wherein an elementary reaction database storing known reaction data related to elementary reactions of the reactive gas, the material gas, and the it said material gas is the two not respect elementary reactions involved in performing a simulation using a simplified model database storing response data two to four reaction process and an intermediate gas 4 Is a simulation method which comprises using a simple model assuming only reaction process like.

  When thin film formation is performed by CVD using various source gases, even if there is no data about the reaction process in the gas phase or solid phase in which the source gas is involved, A certain simulation can be performed. Also, when applied to dry etching or the like instead of CVD, it is possible to perform a simulation with physical validity as well. That is, since it is possible to perform a relatively accurate simulation without spending a great deal of cost and calculation time, the design of the CVD apparatus and the etching apparatus becomes easy.

Hereinafter, embodiments of the present invention will be described with reference to the drawings, but the present invention is not limited to the specific modes described below. In the following description, an example of forming a SiO ₂ thin film on a silicon surface using plasma CVD that operates under atmospheric pressure will be described. However, the present invention uses a thin film formed by thermal CVD or a corrosive gas. It can also be used for etching simulation.

FIG. 1 is a schematic diagram showing an image of a reaction in plasma CVD to be simulated. In the gas phase, TEOS is introduced as a material gas for forming a thin film, nitrogen is introduced as a discharge gas for causing a discharge, and oxygen is introduced as a reactive gas for causing a chemical reaction in the material gas. ing. Electrons and various ions, atoms, and decomposition products are generated in the gas phase by the discharge from the electrodes. As an example of the reaction, when the electrons 3 collide with the oxygen molecules 10, the oxygen molecules 10 are decomposed into oxygen atoms 11. When this oxygen atom 11 collides with the raw material molecule 21, the raw material molecule 21 is decomposed into, for example, a decomposition molecule 22 and a film precursor 23. When this film precursor is deposited on the substrate surface, a desired SiO ₂ thin film is produced.

  These are intricately intertwined in the actual reaction. For example, even if only an elementary reaction involving only oxygen occurs, the reaction results in a difference in whether it becomes an oxygen atom, an oxygen molecule, or ozone. For example, even if the reaction results in an oxygen atom, its electronic state can take various excitation levels. The same applies to oxygen molecules, and the same applies to other discharge gases and material gases. In particular, since the material gas has a relatively complicated molecular structure, the reaction can also take various steps. In addition to the gas phase, various reactions and movements can occur on the surface of the solid phase as in the gas phase. For example, original substrate atoms exist on the solid phase surface, but they are attacked by electrons or the like to generate dangling bonds and the like, and nonuniformity such as lattice defects existing on the substrate is also affected. In particular, in CVD under atmospheric pressure, there is a great opportunity for collision of atoms and the like, and the resulting reaction is extremely complicated.

  In the present invention, these complicated reactions are classified into at least two. One is a first elementary reaction process involving only a reactive gas and a discharge gas, and the other is a second elementary reaction process involving a material gas. As for the first elementary reaction process, data of rate coefficients indicating conventionally known reaction equations and their occurrence probabilities are taken into the simulation. Thereby, an accurate reaction process can be reproduced for a part of the elementary reaction. Since the rate coefficient of the first elementary reaction process generally varies depending on the reaction conditions, it is represented by a fixed number table or a fixed function form such as an Arrhenius type.

  On the other hand, regarding the second elementary reaction process involving the material gas, there is almost no known data available except for the case of silane gas. Therefore, for example, a simple model assuming only two to four types of reaction processes such as the two types of reaction processes shown in FIGS. Push into the rate coefficient of the reaction process of the kind. The reason why the number of the second elementary reaction process is changed from 2 to 4 is within this range, and within the practical time and cost range, it is possible to perform a relatively accurate simulation that takes into account the intermediate. Because it becomes. More preferably, there are 2 to 3 types, and 2 types are most preferable. This is because as the number of types increases, time and cost are required, and the degree of improvement in simulation accuracy is small. Hereinafter, the case where the second elementary reaction process is the two types shown in FIGS. 2-1 and 2-2 will be described as an example.

First, FIG. 2A shows a process in which the material gas moves to the surface 2 of the silicon substrate 1 as it is and reacts with the oxygen atoms 10 in that state. FIG. 2B shows that, following (a), the reacted material gas 20 is separated into SiO ₂ 40 and the residue (1) 50 on the substrate surface, and SiO ₂ 40 is deposited on the surface 2. The process of forming a desired thin film is shown. Thus, one reaction process in which the material gas forms a thin film on the substrate surface as it is is assumed and used.

Further, in FIG. 2-2 (a), the oxygen atoms 11 collide with the material gas 21 in the gas phase to generate an intermediate 30 that becomes a film precursor (the remainder is omitted). Shows a process of moving to the surface 2 of the silicon substrate 1 and reacting with other oxygen atoms 12 in this state. FIG. 2-2 (b) shows that, following (a), the reacted intermediate 30 is separated into SiO ₂ 41 and the residue (1) 51 on the substrate surface, and SiO ₂ 41 is deposited on the surface 2. The process of forming a desired thin film is shown. Thus, the material gas once becomes an intermediate and is used assuming a reaction process in which a thin film is formed on the substrate surface.

  In this way, assuming that only one kind of intermediate is generated, the intermediate is not involved as shown in FIGS. 2-1 and 2-2 as the second elementary reaction process involving the material gas. Consider only two types of reaction processes, ones and those involved. Then, the second elementary reaction process involving the material gas is converted into a rate coefficient of these two kinds of reaction processes. Note that the rate coefficient of the second reaction process is expressed by a single numerical value.

  That is, the simulation is performed after the second elementary reaction process involving the material gas is converted into two parameters corresponding to FIGS. Therefore, even if there is no knowledge of the detailed elementary reaction process regarding the material gas, it is possible to perform a somewhat accurate simulation considering the intermediate. In addition, regarding the first elementary reaction process involving a reactive gas or the like, as described above, accurate data measured from an actual reaction is included. In other words, depending on the content of the reaction process, a hybrid type combining a process using accurate reaction data and a reaction process including intermediates using a simple model reduces the time and cost required for the simulation and makes a comparison. It is possible to achieve both accurate and accurate results. In particular, CVD under atmospheric pressure is effective because complex reactions are likely to occur.

When more accurate simulation is performed assuming that the second elementary reaction process is three or four, the intermediate type is changed to two or three according to the increased reaction process. Can be increased. In this case, similar in FIG. 2-1 and 2-2, and the reaction process via two intermediates until the SiO ₂ material gas, the intermediate material gas 3 kinds until SiO ₂ It is better to assume a reaction process, such as a reaction process through the body. It goes without saying that other highly likely reaction processes may be included. Then, the types of rate coefficients increase corresponding to each reaction process.

  By the way, in order to perform the simulation, it is necessary to determine in advance the rate coefficient of the second elementary reaction process. In order to determine these rate coefficients, for example, in the actual reaction, only the reactions corresponding to FIGS. 2-1 and 2-2 are specified, the rate coefficients are experimentally obtained, and these are used for simulation as temporary reaction data. it can. However, the above-mentioned rate coefficients relating to the second elementary reaction process are rounded up to two or four kinds including all of the various complex elementary reaction processes that actually occur. Therefore, even if the reaction is specified experimentally and the rate coefficient is obtained, the simulation result using such a rate coefficient does not always match the result of the thin film formation experiment.

  Therefore, as the rate coefficient of the second elementary reaction process used in the simulation, first, simulation is executed using experimental values or analogy values from other experiments as temporary reaction data. Next, the results obtained from the simulation are compared with the actual experimental results corresponding to the simulation. In this example, the film formation rate obtained by simulation is compared with the film formation rate obtained from an actual thin film formation experiment. If both the film forming speeds match within a certain range, it is determined that the initially determined rate coefficient is correct, and the process is terminated.

  On the other hand, when the film formation rate obtained by the simulation is smaller than the film formation rate obtained from the actual thin film formation experiment, the rate coefficient is adjusted to increase and the simulation is performed again. At this time, which of the two rate coefficients is increased is arbitrary, but in this example, the two are increased at the same ratio. On the contrary, when the film formation rate obtained by the simulation is larger than the film formation rate obtained from the actual thin film formation experiment, the rate coefficient is adjusted to decrease and the simulation is performed again. Again, in this example, the two rate coefficients are reduced by the same ratio.

  In this way, the simulation is repeated until the film formation rate obtained by the simulation substantially matches the film formation rate obtained from the actual thin film formation experiment, and the simulation is repeated until the film formation rate substantially matches. At this stage, an improved rate factor consistent with the experiment can be obtained. The rate coefficient thus obtained can be used for simulation under different conditions, for example, when the electrode shape of the apparatus is different.

  In this example, the validity of the two rate coefficients was determined by comparing only the film formation speeds. However, the simulation was performed by varying each one individually to obtain a more appropriate rate coefficient. May be. In addition, when other physical property values such as electrical resistance and film thickness distribution are obtained by simulation, the validity of a plurality of rate coefficients can be individually determined and adjusted based on them.

  Next, a system for performing such a simulation will be described. FIG. 3 is a block diagram showing an outline of the simulation system 100 as viewed from the control side. The system 100 includes a storage unit 200 having an input selection database 210, an elementary reaction database 220, a simple model database 230, an arithmetic expression database 240, and a control program and data read from the storage unit 200 to a ROM. And a processing unit 110 constituted by a ROM, an input device 160 such as a keyboard and a mouse, and a display 170 such as a liquid crystal display or a CRT for displaying an operation screen are connected to each other via a bus via a necessary interface. .

  The input selection database 210 includes types of discharge gases such as nitrogen and argon, types of reactive gases such as oxygen and hydrogen, types of material gases such as silane, TEOS, and HMDSO, and gas composition ratios introduced into the gas phase. Design CVD equipment etc., such as gas flow rate, reactor size and shape, area of substrate to be accommodated, electrode size and structure, voltage or power applied to electrode and drive frequency, dielectric constant and permeability of structure Various types of data required for operation and operation are stored in a selectable manner. Moreover, whether the gas phase of the simulation model is a particle model or a fluid model, and data for specifying the reaction process of the second elementary reaction process are selectably stored.

  The elementary reaction database 220 includes a reaction equation of an elementary reaction process in which a discharge gas such as nitrogen and argon and a reactive gas such as oxygen and hydrogen are involved, and a material gas and a decomposition product thereof are not involved, and generation of each reaction equation. Data of rate coefficient (or reaction rate constant) indicating probability is stored. These data have already been stored with accurate elementary reaction data that has been obtained by some experimentation and is available.

  The simple model database 230 includes reaction formulas of two to four elementary reaction processes assumed as described above, and rate coefficients (or reaction rates) of the respective reaction formulas for the reaction processes involving the material gas. Constant) data is stored. This rate coefficient is reaction data when a complicated and diverse reaction process involving a material gas is rounded by two to four elementary reaction processes. When these rate coefficients are actually measured data or analogy data from another experiment, these rate coefficients are provisional reaction data. These temporary reaction data are improved so as to be more consistent with the experiment by repeatedly performing simulation and correction of the rate coefficient. The provisional reaction data in the simplified model database 230 is sequentially replaced with the improved rate coefficient.

  The operational formula database 240 is necessary for models used for executing simulations such as various equations such as Poisson equation and continuity formula, programs and data for Monte Carlo collision calculation, calculation formulas and data for surface reaction calculation, etc. Programs and data are stored. As a simulation model, a fluid model that treats the entire gas phase as a fluid and can be easily adapted to CVD under atmospheric pressure, or a particle model that treats individual atoms as particles and is easily adapted to CVD under reduced pressure can be selected. . In addition, screen information and the like necessary for system operation are also stored.

  The processing unit 110 of the simulation system 100 includes a condition input unit 120, a calculation unit 130, a determination unit 140, and a result display unit 150. The processing of the processing unit 110 will be described with reference to the flowchart of FIG.

  When the process starts, the condition input unit 120 detects the CVD operation condition and the apparatus condition, the simulation model condition to be used (whether the gas phase is a fluid model or a particle model, and the second elementary reaction process). As the reaction process, and what kind of reaction process is assumed) and the like are read out from the input selection database 210, and screen information for selecting and inputting conditions is read out from the arithmetic expression database 240 and the display 170 is read out. To display. In addition, screen information that can input physical property values such as film thickness obtained in actual thin film formation experiments is similarly read out and displayed on the screen. When data necessary for the simulation is input, the data is sent to the calculation unit 130 (step S10).

  Next, based on the data selected and input in step S10, the arithmetic unit 130 obtains reaction formulas that can occur in the thin film formation process as elementary reaction data relating to the selected reactive gas and discharge gas from the elementary reaction database 220. The known data of each rate coefficient is read (step S20). This makes it possible to reproduce an accurate reaction process for the first elementary reaction process that does not involve the material gas.

  Subsequently, based on the data input in step S10, the arithmetic unit 130 determines the number of types of the elementary reaction processes selected previously and the respective elementary reactions for the second elementary reaction process involving the material gas. The process is specified, and the corresponding reaction equation and the rate coefficient (or reaction rate constant) data, which is provisional reaction data, are read (step S30). This makes it possible to execute a simulation based on a simple model without specific data relating to the second elementary reaction process involving the material gas, and by combining with the data read in step S20, A so-called hybrid simulation model can be constructed. This completes the data required for the simulation.

  Subsequently, a simulation is executed using these data (step S40). In this example, a rectangular plane substantially corresponding to the substrate position and the reactor size is considered in a direction perpendicular to the substrate surface as shown in FIG. 1 or FIG. 2, and boundary conditions corresponding to the boundary properties are set. . In addition, regarding the vertical direction of this plane, a periodic boundary condition corresponding to the length of the reactor is set, and simulation is performed in two dimensions. Needless to say, other boundary conditions and coordinate shapes may be used. For example, a cylindrical coordinate system may be used and a plane including the Z axis may be considered. In this case, the periodic boundary condition is not necessary. Further, the simulation is performed under the condition that a material gas is introduced into the reaction furnace at a constant flow rate, a film is formed at a constant speed, and the gas phase constitutes a steady state. That is, the state at an arbitrary time t during the thin film formation is started as an initial state, and convergence is determined based on whether or not a steady state is realized.

  Based on these assumptions, a detailed flow of step S40 when the simulation is performed based on the fluid model will be described with reference to the flowchart of FIG. First, the Poisson equation is calculated, and the potential distribution is obtained assuming that the charge distribution is neutral (step S41). Next, a continuous equation of charged particles such as ions and electrons is calculated (step S42). Thereby, the charge density at an arbitrary position in the gas phase is obtained. Next, the momentum conservation formula is calculated (step S43). Thereby, the velocity distribution of the particles in the gas phase is obtained, and the charge density is determined together with the continuous equation. Next, an average electron energy equation is calculated from the potential distribution and the electron charge density (step S44). Thereby, the average energy of the plasma electrons generated through the first elementary reaction process is obtained under the apparatus conditions and the operating conditions to be simulated. This equation is for reflecting, in the simulation, a non-linear rate coefficient change that reflects the characteristics of the apparatus in the first elementary reaction process. Next, from the particle density and particle flux obtained by these calculations, physical property values such as a film forming speed for forming a thin film on the substrate surface are obtained (step S45).

  Next, the time t is advanced by Δt (step S46), and the calculation from step S41 to step S45 is repeated again to obtain physical property values such as electron charge density and film formation speed at time t + Δt.

  Next, it is determined whether the simulation result realizes a steady state (step S48). Specifically, the periodic average value of the charge density of the electrons at the time t obtained in steps S41 to S45 is compared with the periodic average value of the charge density of the electrons at the time t + Δt determined in step S47. When the difference is 1% or less, it is determined that the steady state is reached. If it is determined in step S48 that the steady state is not established, the flow branches to the right and steps S46 to S48 are performed. That is, the time t is further advanced by Δt, and the calculation from step S41 to step S45 is repeated. If the steady state is realized in step S48, the flow branches downward and ends step S40. In the state where the steady state is realized, the thin film increases in thickness, but in the gas phase, the raw material gas supplied from the outside and the substance deposited on the substrate surface are balanced, and the gas phase is Thus, a steady state is realized.

Next, returning to FIG. 4, the flow proceeds from step S40 to step S50. In this step, it is determined whether or not the simulation result reaching the steady state matches the result of the thin film formation experiment performed separately and input in step S10. Specifically, it is determined whether or not the deposition rate of the SiO ₂ thin film substantially matches the experimental result within a certain error range.

  If the two do not match, the flow branches to the right and proceeds to step S70. If the film formation rate obtained by the simulation is smaller than the film formation rate obtained from the actual thin film formation experiment, the extent is determined. Depending on, the two rate factors are increased by the same ratio. On the other hand, when the film formation rate obtained by the simulation is higher than the film formation rate obtained from the actual thin film formation experiment, the two rate coefficients are decreased at the same ratio according to the degree. Subsequently, the flow returns to step S40, and the simulation calculation is performed again.

  On the other hand, if the two match, it can be determined that the simulation actually reproduces the experiment well, so the result display unit 150 displays the simulation result on the display 170 and is finally obtained. The improved rate coefficient of the second elementary reaction process is stored in the simplified model database 230, and all the processes are completed.

  In addition, when the rate coefficient of the second elementary reaction process necessary for the simulation has already been improved from the provisional reaction data by going through the flow of FIG. 4 once, when the shape of the apparatus has changed When performing the simulation, it is not necessary to go through the steps S50 and S70, the simulation of the step S40 is performed using the improved rate coefficient, and the result is immediately displayed in the step S60.

  Next, the flow when the particle model is used instead of the fluid model in step S40 of FIG. 4 will be described with reference to the flowchart of FIG. The particle model fits when the furnace pressure is relatively low, such as a vacuum. In this case, as in the case of using the fluid model, a rectangular plane is assumed in the reaction furnace, and the boundary condition corresponding to the boundary property is set in the same manner as the fluid model. The convergence of the simulation result is the same as in the fluid model.

  When the flow starts, first, assuming that the charge distribution is medium positive, a potential is obtained from the Poisson equation for each intersection of grids provided by dividing the plane into a plurality for calculation (step S81). Next, an electric field at an arbitrary position in the unit cell divided by the grid is interpolated from the previously obtained potential (step S82). Next, using this electric field, a motion equation of a predetermined number of particles selected to represent the behavior of the plasma is calculated, the position of each particle is obtained, and the particle density is obtained by statistical processing (S83). Step). The type of collision between the particles and the neutral gas such as the raw material gas or the discharge gas and the direction after the collision are obtained by Monte Carlo collision calculation using a pseudo random number, and the particle flux is obtained (step S84). Thereafter, similarly to the fluid model, the surface reaction calculation is performed (step S85), the time t is increased by Δt, the steps S81 to S85 are repeated (steps S86 and 87), and the convergence determination (step S88) is performed. Then, a steady-state result is obtained and the process ends. Hereinafter, returning to step S50 in FIG. 4 is the same as in the fluid model.

  Here, the model used for the simulation is not limited to the above-described fluid model or particle model, and may be any model that is considered more appropriate. Also, the equations used for the calculation can be used in various combinations of various equations and approximate equations. In addition, while the simulation system is executing the process, it is optional to display the state during calculation on the display as needed. In addition, the determination of convergence to a steady state may use other physical property values instead of the fluctuation range of the electron density, and even when the electron density is used, a certain fluctuation greater than 1% is allowed. May be. Furthermore, in determining the validity of the simulation, other physical property values obtained from both actual experiments and simulations may be used, not limited to the film formation rate.

  When this simulation system is used for etching, the reactive gas in CVD becomes a corrosive gas for the substrate, and the material gas and intermediate gas in CVD become various etchants desorbed from the substrate surface. Therefore, the second elementary reaction process during etching may be considered by appropriately assuming one to three intermediates so as to reproduce the components contained in the exhaust gas when etching is performed.

  The simulation system can also be expressed in the form of a program that is executed by a computer. These programs may be stored in a computer-readable recording medium. Here, the computer-readable recording medium refers to a recording medium such as a flexible disk, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the program can be divided into arbitrary appropriate parts, and the divided parts can be stored in the storage medium.

It is the schematic diagram which showed the image of the gas phase reaction in CVD. It is the figure which showed typically one of the elementary reaction processes of a simple model. It is the figure which showed typically another one of the elementary reaction process of a simple model. It is the block diagram which showed the schematic structure which looked at the simulation system from the control surface. It is the flowchart which showed the outline of the processing flow of a process part. It is a flowchart explaining the detail of step S40 at the time of using a fluid model. It is a flowchart explaining the detail of step S40 at the time of using a particle model.

Claims (7)

A gas phase, a solid phase, and a solid phase surface that is an interface between the two phases, and the gas phase can be detached from the solid phase surface by a chemical reaction or formed by the chemical reaction Comprising: a material gas; an intermediate gas between the material gas and a substance constituting the solid phase surface; and a reactive gas capable of reacting with the material gas or the solid phase surface; A system that generates mass transfer with a chemical reaction between
An elementary reaction database storing known reaction data relating to elementary reactions of the reactive gas, and a simple model database storing reaction data of two to four kinds of reaction processes including the material gas and the intermediate gas. In the simulation using the elementary reaction database and the simple model database, a simple model assuming only the two or four kinds of reaction processes is used for the elementary reaction involving the material gas. Simulation system.   2. The determination unit according to claim 1, further comprising a determination unit configured to determine the validity of the simulation result, wherein when the result is not valid, the calculation unit corrects the reaction data and repeats the simulation. Simulation system. The simulation system according to claim 1 or 2, wherein the intermediate gas has one to three types.   The simulation system according to claim 1, wherein the chemical reaction is performed by plasma, and the mass transfer occurs in a thin film formation direction on the solid phase surface.   The simulation system according to claim 1, wherein the pressure of the gas phase is substantially the same as atmospheric pressure.   The simulation system according to claim 1, wherein the reactive gas is corrosive to the solid surface, and the mass transfer occurs in a direction in which the solid surface is scraped. A gas phase, a solid phase, and a solid phase surface that is an interface between the two phases, and the gas phase can be detached from the solid phase surface by a chemical reaction or formed by the chemical reaction Comprising: a material gas; an intermediate gas between the material gas and a substance constituting the solid phase surface; and a reactive gas capable of reacting with the material gas or the solid phase surface; A simulation method for a system that causes mass transfer with chemical reaction between
Using an elementary reaction database storing known reaction data relating to elementary reactions of the reactive gas, and a simple model database storing reaction data of two to four kinds of reaction processes including the material gas and the intermediate gas A simulation method characterized by using a simple model assuming only the two to four types of reaction processes for the elementary reactions involving the material gas when performing the simulation.

Images