JPH05114569A - Simulation of impurity in semiconductor element - Google Patents

Abstract

PURPOSE:To provide the simulation method of an impurity in semiconductor element capable of precise calculation in a short time of calculation. CONSTITUTION:Within the impurity simulation method wherein a semiconductor element structure is to be discretely mesh-divided S2, the meshes are to be redivided corresponding to the impurity concentration gradient (S6-S8) per time step DELTAt repeating the calculation step in the diffusion process.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for simulating impurities in a semiconductor device, and more particularly to a method for simulating impurities in which a semiconductor device structure is discretized by mesh division.

[0002]

2. Description of the Related Art As one of computer simulation techniques, there is a process simulation in which a semiconductor element is manufactured based on a physicochemical model in a computer. Further, as a process simulator for a silicon semiconductor, oxidation, ion implantation, diffusion, CVD, etching, lithography,
Functions such as vapor phase growth are required. Then, by combining these, an element region, an element isolation region, a well region, a diffusion layer wiring region, etc. are formed.

Here, the processing procedure of the conventional process simulation will be described with reference to the flowchart of FIG. In FIG. 4, first, data such as process data, structure, and calculation conditions are input (step S4).
1) Then, the semiconductor device structure is discretized by mesh division (step S42). This mesh division is performed according to the impurity concentration gradient after the calculation.

Next, the calculation in the oxidation process, the calculation in the ion implantation process, the calculation in the diffusion process, and the calculation in other processes are sequentially executed (step S43).
_{1 to} S43 _n ), and when the calculation in all processes is completed (step S44), a series of process simulation is ended.

[0005]

As described above, in the conventional process simulation, once the semiconductor element structure is discretized by mesh division, the calculation is performed until the end. Therefore, in the case of impurity simulation, the diffusion process requires the most calculation time, and in order to perform accurate calculation, the mesh needs to be made finer. Conversely, if the mesh is finely divided, the calculation time becomes long. There was a problem. Even if the mesh is redivided in the middle of the calculation, it is only when the process is changed, such as when the shape is changed in the process such as deposition.

The present invention has been made in view of the above problems, and an object of the present invention is to provide an impurity simulation method for a semiconductor device, which requires a short calculation time and enables accurate calculation.

[0007]

According to the impurity simulation method of the present invention, in the impurity simulation performed by discretizing the semiconductor device structure by mesh division, in the heat treatment step, at least the impurity concentration gradient is determined at each time step where the calculation is repeated. To subdivide the mesh.

[0008]

In the impurity simulation, for example, in the diffusion process, the mesh is first subdivided according to the impurity concentration gradient, diffusion calculation for a time step Δt time is repeated for the mesh, and diffusion for Δt time is performed. The mesh is subdivided according to the impurity concentration gradient, and diffusion calculation for the next Δt time is performed on the mesh.

[0009]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 is a flowchart showing a processing procedure of a process simulation according to the present invention. In the figure, first, the data of all steps of the actual process, the shape of the simulation, the calculation conditions in the simulation calculation, and the like are input (step S1), and then the semiconductor structure element structure is discretized by mesh division (step S1). S2).

In this mesh division, when the mesh is re-divided, which will be described later, a process of interpolating and changing the data on the old mesh is performed. At that time, an error should be made as little as possible. To do this, a fine basic mesh that does not depend on the mesh division in each process is created. Then, the mesh subdivision in each process is performed via this basic mesh.

Next, the calculation process is performed in each process. In the ion implantation process, since there is no shape change or time dependency process, first, the calculation time and impurity distribution required for the process are taken into consideration. Then, the mesh is subdivided (step S3). In the case of this mesh subdivision, the mesh is usually rougher than the basic mesh. Then, ion implantation calculation is performed (step S
4) Then, the original fine mesh is re-divided (step S5).

In addition, since the diffusion process, which is one of the heat treatment processes, is a time-dependent process, the mesh is first subdivided according to the impurity concentration gradient (step S6), and the mesh time is used. Diffusion calculation for step Δt time is performed (step S7). Subsequently, it is determined whether or not the diffusion calculation for the time t required for the diffusion process is completed (step S8), and until it is completed, the mesh is subdivided according to the impurity concentration gradient after the Δt time diffusion, The process of performing diffusion calculation for the next Δt time on the mesh is repeated (steps S6 and S7). Then, when the diffusion calculation for the time t is completed, the process proceeds to step S5 and the original fine mesh is re-divided.

Further, after performing calculations in the oxidation step and other processes and subdividing into the original fine mesh, it is judged whether or not the calculations in all processes are completed (step S9), and when all are completed, Finish a series of process simulations. Note that in the case of an oxidation step, which is one of the heat treatment steps, it is assumed that the mesh is subdivided in consideration of both the shape and the impurity concentration gradient, assuming that the diffusion and the shape change are combined. Further, in the case of a deposition process or an epitaxy process involving a shape change, the mesh is re-divided according to the shape change after time step Δt.

FIG. 2 is a diagram showing an example of how the mesh changes as the shape changes due to the deposition and oxidation steps. In the figure, the thin line a shows the mesh and the thick line b shows the material boundary. There is. In the figure, (A) shows the initial state, (B) shows the state after the shape change by the deposition process, and (C) shows the state after the shape change by the oxidation process. The changes in the mesh shown in FIGS. 2A to 2C are due to the conventional technique.

FIG. 3 is a diagram showing an example of how the mesh is changed due to the change of impurities due to the ion implantation and diffusion steps (including thermal diffusion by oxidation). In the figure, a thin line a represents the mesh and a thick line. b indicates the material boundaries. In the figure, (A) shows the initial state, (B) shows the state of the mesh at the time of ion implantation, (C) shows the state of the mesh after Δt _i time diffusion, and (D) shows the state after Δt _j time diffusion. Shows the state of each mesh. In order to spread the time t, t = Δt ₁ + Δt ₂ + ... + Δt _i + ...
If + t _n , the mesh is reorganized every minute time Δt.

[0016]

As described above, according to the present invention,
In the impurity simulation performed by discretizing the semiconductor element structure by mesh division, in the heat treatment step of the diffusion step and the oxidation step, the mesh is re-divided according to at least the impurity concentration gradient at each time step where the calculation is repeated. The calculation time is short and the calculation can be performed accurately. As a result, many impurity simulations can be used, and the semiconductor development period can be shortened.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing procedure of a process simulation according to the present invention.

FIG. 2 is a diagram showing an example of how a mesh changes with a shape change due to a deposition and oxidation process.

FIG. 3 is a diagram showing an example of how the mesh changes due to changes in impurities due to ion implantation and diffusion steps.

FIG. 4 is a flowchart showing a processing procedure of a conventional process simulation.

[Explanation of symbols]

 a mesh b material boundary

Claims (1)

[Claims] 1. In an impurity simulation performed by discretizing a semiconductor device structure by mesh division, in a heat treatment step, each time step for repeating calculation,
A method for simulating impurities in a semiconductor device, characterized in that the mesh is re-divided according to at least an impurity concentration gradient.