JPH07115071A - Designing method for semiconductor device - Google Patents

Abstract

PURPOSE:To provide simulation which easily calculates the diffusing conditions of impurities in the vertical direction and horizontal direction even when the conditions for manufacturing an impurity diffusion layer are changed. CONSTITUTION:Several combinations of ion implanting condition and annealing condition are decided prior to simulation, and based on the conditions, concentration distribution of the impurity in an impurity diffusion layer is calculated or actually measured by Monte Carlo method (a) or C-V method and the results are stored in the data base. The data base is used for process simulation A for forming an impurity diffusion layer. As a result, the easy and highly accurate process simulation A can be conducted by using the data base even when manufacturing conditions are changed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor manufacturing technique, and more particularly to a technique particularly effective when applied to a design method for determining the manufacturing conditions of an impurity diffusion layer according to a process simulation. Related to useful technology for designing.

[0002]

2. Description of the Related Art In a manufacturing process of a semiconductor device, a method is used in which an impurity is ion-implanted into a semiconductor substrate and annealed to form an impurity diffusion layer.
This ion implantation condition (for example, the type of impurities, the energy at the time of implantation, the implantation angle, the implantation amount, etc.),
Manufacturing conditions such as annealing conditions (annealing temperature and annealing time) are set so that the impurity diffusion layer has a desired concentration distribution.
Determined using a process simulation tool. The impurity concentration is such that impurities are actually introduced into the semiconductor substrate according to a constant impurity ion implantation condition and annealing condition, and the depth direction (vertical direction) of the introduced impurities
Is detected by secondary ion mass spectrometry (SIMS analysis) or the like, and is determined based on the detection result. SI
The measurement of the impurity concentration by the MS analysis can be carried out, for example, in J.
Electronic Materials vol 18 (1989) pp. 145-150.

[0003]

However, the present inventors have clarified that the above-mentioned technique has the following problems. That is, in the above SIMS analysis, one-dimensional data (impurity concentration distribution in the vertical direction of the semiconductor substrate)
Therefore, in the simulation constructed based on the result of the SIMS analysis, the two-dimensional data of the impurity concentration distribution (for example, the concentration distribution in the depth direction and the concentration distribution in the lateral direction) is obtained. It cannot be reflected.
Therefore, for example, when forming the source / drain regions of a MOS transistor, with respect to the impurity diffusion layers formed on both sides of the gate, to what extent the diffusion layers extend just below the gate (concentration distribution in the lateral direction). Cannot be predicted.

Further, in recent semiconductor manufacturing technology, submicronization is progressing in the manufacturing process, and in the manufacturing process of such a scale, the influence of changes in manufacturing conditions on device characteristics becomes peculiar, and conventional manufacturing It was revealed that the device characteristics cannot be estimated by using the conventional simulation based on the relationship between the conditions and the device characteristics.

In a submicron-sized device, a high concentration and shallow impurity diffusion layer may be formed. In this case, at the time of initial annealing, the diffusion speed of the ion-implanted impurities becomes abnormally high, so-called “accelerated diffusion”.
Is known to occur. Regarding this enhanced diffusion as well, the concentration distribution in the one-dimensional direction (vertical direction) is
Although it can be measured by analysis etc., data considering the lateral spread (diffusion distance) is not required, and therefore,
The simulation for forming such a shallow impurity diffusion layer could not be performed accurately.

The present invention has been made in view of the above circumstances, and even when the manufacturing conditions of the impurity diffusion layer are changed, the impurities in the vertical and horizontal directions in the substrate under the manufacturing conditions are changed. It is a main object of the present invention to provide a method for designing a semiconductor device that can easily predict a diffusion state. The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

[0007]

The typical ones of the inventions disclosed in the present application will be outlined below. That is, in the first invention, the impurity ion implantation conditions and the annealing conditions for constructing the database are determined in advance, and the impurity concentration distribution is calculated by the Monte Carlo method based on the conditions thus determined, With the implantation conditions and the annealing conditions as parameters, the calculation results are made into a database, and based on this database, a simulation concerning the implantation conditions and the annealing conditions when actually forming the impurity diffusion layer is performed. In the second invention, the impurity ion implantation conditions and the annealing conditions for constructing the database are determined in advance, and the ion implantation and / or annealing is performed based on the conditions thus determined. The concentration distribution of the impurity diffusion layer is measured by an experiment, and the measurement result is made into a database using the above-mentioned impurity ion implantation condition and annealing condition as parameters, and the implantation condition when actually forming the impurity diffusion layer is based on this database. Also, the simulation regarding the annealing conditions is performed.

[0008]

In the first aspect of the present invention, the impurity concentration distribution is obtained with high accuracy by the Monte Carlo method based on the predetermined ion implantation and annealing conditions, and the calculation results are stored in a database for simulation. Therefore, it is possible to predict the impurity concentration distribution in a two-dimensional manner, and moreover, when the manufacturing conditions are changed, it is suitable for various changed manufacturing conditions only by setting a predetermined number of conditions for constructing the database. The simulation can be easily performed. In the second invention, the impurity diffusion layer is actually formed according to the predetermined ion implantation and annealing conditions, the concentration distribution of the impurities is measured, and the measurement result is made into a database and reflected in the simulation. Therefore, the impurity concentration distribution can be predicted two-dimensionally, and when the manufacturing conditions are changed, a simulation suitable for various changed manufacturing conditions can be performed only by setting a predetermined number of conditions for constructing the database. It can be done easily.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing the configuration of a diffusion layer forming process simulation tool of this embodiment. As shown in this figure, in the simulation of this embodiment, desired data necessary for forming the impurity diffusion layer is stored in the database in step (a) in advance. And
When the process simulation is actually performed in step (A), the optimum simulation is performed based on this database.

By the way, in the database creation in the step (a), a plurality of sets of manufacturing conditions (ion implantation conditions, furnace annealing conditions) for the impurity diffusion layer are determined in order to construct the database, and based on these manufacturing conditions. Then, the calculation by the Monte Carlo method is performed, and the impurity concentration distribution when the impurity diffusion layer is formed under the determined conditions is calculated with high accuracy. Then, a process simulation of step (A) is performed based on the above database, and then a device simulation of step (B) is performed to predict the device characteristics of the LSI, and a trial manufacture is performed in step (C).

When the database is created in the step (a), the type of impurities (phosphorus, boron, arsenic, etc.), implantation amount (dose amount), implantation energy, implantation angle, furnace annealing temperature, furnace annealing time are set to 1 The manufacturing conditions for the set are determined. That is, when the above six conditions (1) to (5) for creating the database are respectively determined, the concentration distribution state of the impurities in the semiconductor substrate is calculated based on these values by the known Monte Carlo method. In this case, a two-dimensional density distribution (vertical diffusion distance and horizontal diffusion distance) is obtained. In this embodiment, the concentration distribution of impurities immediately after ion implantation and the concentration distribution of impurities after being diffused by annealing are obtained. The data obtained at this time is made into a database, and graph (b), table (c), functional model (d)
Etc. (FIG. 1). When actually forming the impurity diffusion layer on the semiconductor substrate, a desired impurity concentration distribution should be obtained according to the graph (b), the table (c), or the functional model (d) based on the above database. Then, a process simulation is performed to determine the actual manufacturing conditions (-) for obtaining desired device characteristics.

FIG. 2 is a cross-sectional view showing the device structure of the n-type MOS transistor 10 whose manufacturing conditions are determined by the above diffusion layer formation process simulation. In FIG. 2, 1 is a p-type semiconductor substrate and 2 is a gate oxide. Membrane, 3 is source
N-type impurity diffusion layers 4 forming a drain region indicate gate electrodes. Immediately after the ion implantation, the pn junction surface is formed along the isoconcentration line S1 shown in the figure, and after annealing, the pn junction surface is formed along the isoconcentration line S2 shown in the figure. The inside of the line S2 becomes the source / drain region 3.

The above manufacturing conditions (1) to (4) are set to arbitrary values, the impurity concentration distribution after phosphorus (P) ion implantation is performed under the manufacturing conditions is calculated by the Monte Carlo method, and further, after annealing is performed. The concentration distribution of the impurities in is calculated by the Monte Carlo method. In this case, the depth direction distance V _{1 of} the pn junction surface (S1) immediately after ion implantation and the depth direction diffusion distance V ₂ after annealing can be obtained by the Monte Carlo method. Further, the diffusion length (lateral diffusion length) L ₁ immediately after ion implantation and the lateral diffusion length L ₂ after annealing, which extends below the gate oxide film in the transistor 10, can be obtained by the calculation by the Monte Carlo method. As described above, the ion implantation conditions to the annealing conditions are set to arbitrary values, and the calculation by the Monte Carlo method is performed based on the conditions thus set, whereby the impurity concentration distribution of the impurity diffusion layer formed under the predetermined manufacturing conditions is determined. Data can be obtained to predict

Next, a method of calculating the impurity concentration distribution by the Monte Carlo method based on the sampled manufacturing conditions 1 to 3, and making it a database will be described. First, immediately after ion implantation (before annealing),
The vertical distance (V ₁ ) and the horizontal distance (L ₁ ) of the impurities are calculated based on the Monte Carlo method, and the results are expressed with the energy at the time of ion implantation (acceleration voltage E) as a parameter (FIGS. 3 and 4). ). These relationships can be approximated by the following equations (1) and (2) as a linear function of the acceleration voltage E. V ₁ = a ₁ E (1) L ₁ = a ₂ E (2) Here, a ₁ and a ₂ are coefficients determined based on the calculation data by the Monte Carlo method.

That is, according to the graph showing the relationship between the acceleration voltage (E) and the vertical distance (V ₁ ) shown in FIG. 3, it can be seen that these are in a substantially proportional relationship. Similarly, according to the graph of FIG. 4 showing the relationship between the acceleration voltage (E) and the lateral distance (L ₁ ), it can be seen that these are in a substantially proportional relationship.
For example, at the time of implanting phosphorus, the ratio between the coefficient a ₁ representing the relationship between the acceleration voltage E and the vertical distance V ₁ and the coefficient a ₂ representing the relationship with the vertical distance L ₁ is 3.8: 1. It will be about 0.

As shown in the above graph, if a predetermined number of ion implantation conditions (accelerating voltage) and the like are sampled and the data is stored in a database, even if other manufacturing conditions change, calculation is performed each time. Even if the Monte Carlo method, which requires a long time to perform, or concentration measurement by experiment, is not performed, a simulation incorporating the database is performed,
The vertical distance V ₁ and the horizontal distance L ₁ can be easily and accurately predicted from the acceleration voltage E. Incidentally, as shown in FIGS. 3 and 4, even when other impurities (As, B, BF ₂ ) are introduced into the substrate, the acceleration voltage E and the vertical distance V, or the horizontal distance L are as follows. It becomes a substantially proportional relationship, and the above (1),
It can be approximated by the equation (2).

The lateral diffusion distance (L
It was found from the calculation result by the Monte Carlo method that the relationship between ₂ ) and the vertical diffusion distance (V ₂ ) has the relationship shown in FIG. 5 regardless of the ion implantation conditions. Further, these relationships can be expressed by the following equation (3). _{L 2 = a 3 V 2 +} b ... (3) where a ₃ is a coefficient which is determined based on the calculated data by the Monte Carlo method, b is variable.

Further, according to the graph showing the relationship between the vertical diffusion distance V ₂ and the horizontal diffusion distance L ₂ shown in FIG. 5, the coefficient of the equation (3) representing the relationship between V ₂ and L ₂ is shown. a ₃ is about 0.
It becomes 9. The lateral diffusion length L ₂ after annealing can be predicted based on the vertical diffusion distance V ₂ . This result is also made into a database and incorporated in the process simulation. Incidentally, as shown in the figure, even when impurities (As, B, BF ₂ ) other than phosphorus are ion-implanted into the substrate and then annealed, the vertical diffusion distance V ₂ and the horizontal diffusion are obtained. It has been found that the distance L ₂ has a proportional relationship of 0.9. FIG. 6 shows the calculation results obtained by the calculation by the Monte Carlo method as described above under the manufacturing conditions.
It is a table based on.

(1) to (3) obtained as described above
The formula, the graphs shown in FIGS. 3 to 5, and the operation result stored in the database in any of the tables shown in FIG. 6 are used when the impurity diffusion layer is formed on the semiconductor substrate as described above. Is incorporated into the process simulation, and then a device simulation based on the simulation is performed to prototype the device (FIG. 1).

Instead of calculating the impurity concentration distribution by the Monte Carlo method described above, an impurity diffusion layer is actually formed under the sampled conditions 1 to 3, and the two-dimensional impurity concentration distribution of this diffusion layer is measured ( C-V measurement method, etc.), the vertical distance V ₁ , the horizontal distance L ₁ , the vertical diffusion distance V ₂ , and the horizontal diffusion distance L ₂ are obtained from the measurement results,
Similar to the above, these measurement results can be stored in a database. As described above, by measuring the two-dimensional data of the impurity diffusion layer and storing the measurement result in the database, the spread of the impurity diffusion layer under desired manufacturing conditions can be predicted easily and accurately.

As described above in detail, according to the simulation of this embodiment, several ion implantation conditions and annealing condition groups are determined in order to construct a database, and based on these determined conditions, Monte Carlo method,
Since the concentration distribution of impurities in the impurity diffusion layer is calculated by the CV method, the measurement is actually performed, the results are stored in a database, and the process simulation is performed using the database. Even if is changed, a highly accurate process simulation can be performed with a simple method. In addition, when the impurity diffusion layer is formed by the manufacturing process of submicron in recent years, the database for the accelerated diffusion that affects the impurity concentration distribution can be easily obtained only by predetermining some manufacturing conditions. Is constructed, and the simulation considering the effect of enhanced diffusion can be performed.

Although the invention made by the present inventor has been specifically described based on the embodiments, the present invention is not limited to the above embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say. For example, in the above-described embodiments, the process of forming the impurity diffusion layer at the time of manufacturing the MOS transistor has been described as an example, but the simulation of the present invention may be applied to the formation of the diffusion layer of the bipolar transistor. Further, in the present embodiment, the example in which the furnace annealing is performed to diffuse the impurities has been described, but the simulation can also be applied to the impurity diffusion by the short-time annealing (RTA). In addition, this embodiment is 2
In simulating the three-dimensional concentration distribution,
Although the method of converting the dimensional data into a database has been described, it is also possible to combine the two-dimensional database into a database of data representing a three-dimensional diffusion state.

In the above description, the case where the invention made by the present inventor is mainly applied to the formation of the impurity diffusion layer constituting the transistor, which is the field of application of the background, has been described, but the present invention is not limited thereto. However, it can be generally used in the technology of forming an impurity diffusion layer on a semiconductor substrate.

[0024]

The effects obtained by the representative one of the inventions disclosed in the present application will be briefly described as follows. That is, the two-dimensional concentration distribution of impurities implanted in the semiconductor substrate can be easily predicted. In addition, a new process simulator adapted to changes in manufacturing conditions can be easily constructed, and the impurity concentration distribution and device performance can be easily predicted.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a diffusion layer forming process simulation tool of the present embodiment.

FIG. 2 is a cross-sectional view of an n-type MOS transistor 10 whose manufacturing conditions are determined by simulation of this embodiment.

FIG. 3 is a graph showing the relationship between the vertical distance V ₁ of impurities after ion implantation and the acceleration voltage E.

FIG. 4 is a graph showing the relationship between the lateral distance L ₁ of impurities after ion implantation and the acceleration voltage E.

FIG. 5 is a graph showing a relationship between a lateral diffusion distance L ₂ and a vertical diffusion distance V ₂ after annealing.

FIG. 6 Manufacturing conditions of impurity diffusion layer and vertical / horizontal diffusion distance V
₂ is a table showing the relationship between L ₂ and L ₂ .

[Explanation of symbols]

 1 p-type semiconductor substrate 2 gate oxide film 3 n-type impurity diffusion layer 4 gate electrode 10 MOSFET (n-type MOS transistor) S1 pn junction surface after ion implantation S2 pn junction surface after annealing

Front page continuation (72) Inventor Hiroo Masuda 2326 Imai, Ome-shi, Tokyo Hitachi, Ltd. Device Development Center

Claims (3)

[Claims] 1. When forming an impurity diffusion layer on a semiconductor substrate, a desired impurity concentration distribution is obtained according to a predetermined simulation of an impurity ion implantation condition and an annealing condition subsequently performed. In the method of designing a semiconductor device to be determined, impurity ion implantation conditions for constructing a database in advance, and annealing conditions are determined, based on the determined implantation conditions, or based on the determined implantation conditions and annealing conditions,
A semiconductor characterized in that a concentration distribution of impurities is calculated by a Monte Carlo method, a database of the calculation results is obtained using the above-mentioned impurity ion implantation conditions or annealing conditions as parameters, and a predetermined diffusion layer formation process simulation is performed using this database. Equipment design method. 2. A method of designing a semiconductor device, comprising: determining a condition for ion implantation of impurities into a semiconductor substrate and an annealing condition subsequently performed so as to obtain a desired impurity concentration distribution according to a predetermined simulation. Impurity ion implantation conditions for constructing a database in advance, or annealing conditions are determined, and ion implantation is performed on the substrate based on the implantation conditions, or
Ion implantation and annealing are performed based on the implantation conditions and annealing conditions, the impurity concentration distribution in the semiconductor substrate obtained as a result is measured, and the measurement results are databased using the above-mentioned impurity ion implantation conditions or annealing conditions as parameters. And a method for designing a semiconductor device, characterized by performing a predetermined diffusion layer formation process simulation using this database. 3. The method for designing a semiconductor device according to claim 1, wherein the two-dimensional concentration distribution of the impurities represents a two-dimensional diffusion distance with the acceleration voltage at the time of ion implantation as a parameter.