JPS6161421A - Control system in etching device - Google Patents

Abstract

PURPOSE:To enable to get the sectional configuration of a layer to be etched nearer the desired configuration by a method wherein the secular change pattern in the anisotropic degree of an etching rate is found out according to the desired sectional configuration and the secular change pattern in the anisotropic degree is converted into the secular change pattern of the setting conditions for the etching device. CONSTITUTION:The control program and the conditional table for on etching device are incorporated in a CPU 701. An experiment is conducted in advance on the device setting conditions at a gas flow rate, a high-frequency power and a stage temperature and the conditional table is made out. The operator inputs data on the desired sectional configuration of a layer to be etched and the position of the mask at the starting time of etching onto a tablet 703 and the CPU reads in the coordinate data. Moreover, the initial value of the ratio of selectivity and the absolute value of the etching rate are keyed in by a keyboard 702 and the CPU 701 finds out the secular change pattern PX of the setting conditions for the etching device and the evaluation value thereof and makes the etching device 706 operate while the CPU changes hourly the setting conditions according to the PX through a microcomputer 705.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a control system for an etching apparatus, and particularly to a method for determining conditions when setting conditions are changed over time.

[Background of the invention]

In order to approximate the cross-sectional shape of the etching layer to a desired shape, it is necessary to change the setting conditions of the etching device over time. However, since there is a large degree of freedom in how to change the setting conditions over time (hereinafter referred to as the "time change pattern"), it has traditionally been difficult to appropriately determine this according to the desired cross-sectional shape of the layer to be etched. It was embarrassing.

The method of setting conditions for etching equipment that has been proposed to date that involves temporal changes is to switch the etching gas in the middle of etching.An example is shown in Figure 1 (A, S, Bergendahl et al. ,,“
0ptiLIlization of Plasma
Processing for Silicon-Gat
e FET Manufacturing Application
cations”.

IBM J, RES, DEVELOP, ・VOL26
・Ha 5 (September 1983)].

In FIG. 1, 101 is a layer to be etched (polysilicon layer), 102 is a base (silicon oxide layer), and 103 is a mask (hole resist layer). FIG. 1(a) is a diagram of the desired cross-sectional shape. , (b) is a sectional view before the start of etching, (C) is a cross-sectional view at the time of switching the etching gas, and (d) is a cross-sectional view after etching is completed. By switching the gas, from (b) to (c
) is etching with a small degree of anisotropy, (c) to (d)
performs etching with a large degree of anisotropy.

Since this method of etching gas switching is a simple operation, for example, instead of the cross section of (a),
It has the disadvantage of low accuracy for fine control such as obtaining the cross section of ′).

[Purpose of the invention]

An object of the present invention is to provide a method for temporally changing etching conditions so that the cross-sectional shape of the layer to be etched approaches a desired shape after etching is completed.

[Summary of the invention]

In order to achieve the above object, one aspect of the present invention is to obtain a time-varying pattern of setting conditions of an etching device in two stages according to a desired cross-sectional shape. Namely.

In the first step, a time-varying pattern of the etching rate of the anisotropic liquid is determined depending on the desired cross-sectional shape of the layer to be etched. In the second step, the time change pattern of the anisotropic liquid is converted into a time change pattern of the setting conditions of the etching jig apparatus.

[Embodiments of the invention]

First, the principle of the present invention will be explained.

[Definition of anisotropic liquid time change pattern]

It can be expressed as a vector sum of the etching rate vt, the isotropic etching rate 2, and the anisotropic etching rate V (Fig. 2), where the isotropic etching rate vi is the etching rate in the normal direction of the etched cross section, The anisotropic etching rate V is an etching rate in a constant direction that does not depend on the orientation of the etched cross section.

Anisotropic liquid A is defined as follows.

The temporal change in the etched cross-sectional shape is determined by the absolute value IVtl of the etching rate vt and the temporal change pattern of the anisotropic liquid A.

When determining these time-varying patterns to obtain a desired cross section of the layer to be etched, changes in absolute values affect only expansion and contraction on the time axis, so the absolute values are assumed to be constant over time, and changes in It is only necessary to determine the time change pattern of the anisotropic liquid. Here, the time change pattern of the anisotropic liquid is expressed as in the following equation.

Here, the absolute value of the etching rate I V,, 1 The above equation shows that the etching of the anisotropic liquid A1 is performed for t8 seconds, then the etching rate of the anisotropic liquid A is
, etching is performed for t2 seconds...Finally, etching of anisotropic liquid A is performed for t1 seconds.

Since a change in absolute value can be replaced by a change in time, in the above-mentioned time change pattern of the anisotropic liquid, by changing t without changing A, a time change pattern of arbitrary absolute values can be obtained. That is, the time change pattern (2) of the anisotropic liquid can be converted to (3).

l vtt l s Absolute value of etching rate at i-th time [Mizumoto method of anisotropic liquid time change pattern when mask etching has no effect] Obtain the desired cross-sectional shape when it can be assumed that the mask is not etched The method of determining the anisotropic liquid time change pattern for this purpose will be explained with reference to FIG.

FIG. 3(a) is a sectional view before starting etching.

301 is a mask, 302 is a layer to be etched, and 303 is a base.

FIG. 5B is an explanatory diagram of the etching speed V during the etching process. Here, it is assumed that the anisotropic etching rate v1 is perpendicular to the layer 302 to be etched. On the surface not covered by the mask (exposed surface) 304, the anisotropic etching speed V and the isotropic etching speed V point in the same direction, and the magnitude of the etching speed V is the sum of the magnitudes of both. It is.

In the area (protective surface) 305 that is in the shadow of the mask, anisotropic etching does not proceed, but only isotropic etching proceeds. Therefore, the etching rate V is the isotropic etching rate v
Equal to 1.

Figure (c) is a sectional view at the time of just etching. At this point, the layer to be etched has been completely etched on the exposed surface 304 and the underlying layer 303 is exposed. A line segment ST connecting the end point S of the mask and the thickness Tt of the layer to be etched is perpendicular to the underlying layer.

FIG. 4(d) is an explanatory diagram of the etching speed after the just-etch time. In this case, isotropic etching progresses on the protective surface 306.

FIG. 5(e) shows the cross-sectional shape of the desired layer to be etched.

The method for determining the time-varying pattern of anisotropy of etching rate according to the desired cross-sectional shape shown in FIG. 3(e) is shown below. Here, from the state shown in FIG. It is assumed that the shape is returned to the state shown in the opposite direction of the etching progress (reverse simulation).

(2) Inverse simulation from the state (a) to the state (C) From the state (e) to the state (C), the surface of the layer to be etched is returned to the normal direction of the surface. During this period, the anisotropic etching rate is irrelevant), so the degree of anisotropy may be any value less than 1. The moving distance in the normal direction at this time is 21. ,
((f) in the same figure).

■ Divide the inverse simulation line segment ST (length h) from state (C) to state (a) into n equal parts, and define each point as s,, s
□,...S, 1. s (FIG. 3(g)) - First, find the length a of the perpendicular line drawn from point S, -4 to the surface of the layer to be etched. The exposed surface of the layer to be etched is S,
The protective surface of the layer to be etched is returned to the height of -1 by Q in the normal direction (FIG. 3(h)), and the surface of the layer to be etched at one point before is determined by the removal. The degree of anisotropy at this time is A.

A,=1-Q*.-(4) When this operation is repeated for points 5a-19...p st+s, the state M (a) at the start of etching is returned. The anisotropy obtained at this time is Am-1g...g A41
Let it be A1.

By the processing of ■ and ■ above, the time change pattern of anisotropy P,
is the following formula.

Here, the anisotropy A obtained from A1~A, 20, -□
: Real number between 0 and less than 1 IV, I: Absolute value of etching rate [Method for solving anisotropy time change pattern when mask etching is affected] When it cannot be assumed that the mask is not etched, the desired cross-sectional shape can be determined. The method of determining the anisotropy time change pattern to obtain the above will be explained with reference to FIG.

FIG. 4(a) is a cross-sectional view of the etching before the start of etching, FIG. 4(b) is a cross-sectional view of the etching at the time of just etching, and FIG. 4('c) is a view showing the desired cross-sectional shape after the etching is completed. 401 is a mask, and 402 is a layer to be etched.

403 is a base.

Here, we will explain the inverse simulation in Section 2 of the principle explanation.
Transform ■ into ■′ and ■′ 6 ■ Reverse simulation from state (C) to state (b) Return the surface of the layer to be etched to the normal direction of the surface; during this time, move the mask in the lateral direction Speed V.

The degree of anisotropy can be any value as long as it satisfies the condition that the speed of movement of the layer to be etched is smaller than the movement speed of the layer to be etched in the lateral direction.In this case, the movement distance in the normal direction is set to 2... ))
.

■ Inverse simulation from the state of (b) to the state of (a) The end point of the mask 401 in FIG. 4 (a) is S.

Let C be the foot of the perpendicular line drawn from S to the base 403.

Let T be the end point of the etching target M402 in FIG. 4(b). As shown in the same figure (e), 1sc+=h, ITC
1=b, divide the line segment ST into n equal parts, and divide each point into s, sl...
・Let it be S.

In the same figure (f), first, the length of the perpendicular line drawn from the point 51-2 to the protective surface 405 of the layer to be etched or its extension is 2. . Find 1. The exposed surface 404 of the layer to be etched is returned to a height of S, -1. The protective surface 405 of the layer to be etched is returned by α in the normal direction. In this case, the anisotropic bottom is A.

It is.

A,=1-1. - When the operation in (5) is repeated for the points 5s-zt...l s1+ SQ, the state returns to the state (a) at the start of etching.

The anisotropic base found at this time is A * -11..., A,
, shall be A-engineer.

From ■' and ■', the anisotropic base time change pattern PA is given by the following equation.

Here, the anisotropic base A, , 1 found by A□~A,:■
: Real number greater than or equal to 0 and less than 1 (however, the conditions shown in ■' are satisfied) t1~t, : n・Iv%I Q7.1 """(1-A, *t) lvtl lv, I : IV, , I=h: c lVtI: Absolute value of etching speed 1V=1: Absolute value of lateral movement speed of the mask [Expansion of solution method to other cases] In the above principle explanation, Sections 2 and 3 explain what happens when etching progresses. It is assumed that the absolute value of the etching speed, 1 v,+, and the lateral movement speed, V, of the tip of the mask are constant.

Even when these change, the present solution method can be applied by changing the line segment ST in FIG. 4(f) to a broken line.

Furthermore, it is easy to extend and apply the present solution method even when the etching speed vt changes depending on the inclination, curvature of the layer to be etched, the distance from the end point of the mask, etc.

[How to control etching equipment]

FIG. 5 shows a block diagram of the control procedure of the etching apparatus incorporating the method of solving the anisotropic base time-varying pattern.

The meaning of each variable is shown below.

P: Coordinate data of a curve representing the desired cross-sectional shape after etching is completed SQ: Coordinate data of mask end points before etching starts 1Vtol: Absolute value of etching rate of the layer to be etched 1v, reference value of 1 where lV, l: Absolute value r of the lateral movement speed of the mask layer. : initial value PA of r, : anisotropic base time change pattern r,. : Calculated value of r, P, : Temporal change pattern of etching equipment setting conditions J : Evaluation value of P, E : Conversion convergence flag In this block diagram, manual data is PFISIIIVtO
If r*. and the output data are P and J.

Next, the functions of each block will be explained.

Block 501 is an anisotropic base time-varying pattern solving block.The zero blocks are P,,S,,lVt,l.

Input rl (r6° for the first time) and output PA.

The processing contents of this block are as shown in Principle Explanation 4.

Block 502 is a setting condition time change pattern solving block. The zero blocks are pA, Iv,. Enter 1 and press P
, , J, R, .

Block 503 is a conversion convergence determination block. The six blocks input r8° and rl (rea for the first time), and if the magnitude of the difference is greater than a certain value, outputs E (=1).

Block 504 is a setting condition time change pattern output block. Three blocks input P, J, and E, and E=1.
If so, output P8 and J.

The operation of the entire block is as follows.

(1) Blocks 501 and 502 convert P into pA and p, and calculate an evaluation value J.

(2) Block 503 is r, 6 and r, (first time is r-a
), and if it is small, block 504 outputs P8. If it is larger, update r and return to (1).

Next, the processing flow of block 502 will be explained with reference to FIG. The processing flow is shown in FIG.

601: Input P--IV-ol.

602: Refer to the condition table 621 and select Pl.

Create a time change pattern P□ of r.

The condition table 621 has 0 tables whose conceptual diagram is shown in FIG.
It has a three-dimensional configuration with the axes of degree A and selectivity r, and each lattice point stores the setting conditions of the etching apparatus that realizes the values of each axis and its index Zz.6For example, a parallel plate plasma etching apparatus is used. In this case, items of setting conditions include high frequency power, pressure, etching gasification ratio, DC bias voltage, distance between electrodes, stage temperature, etc. Let these items be Zyu...Zk. The index Zz has a numerical value representing the desirability of these setting conditions, and the closer it is to 0, the higher the degree of desirability.

The following formula is used.

Elements of P (Z, □...Zl,)...(Za+tt
. . zl, □, ) are the 1st . Each element of P is determined to satisfy the following conditions.

Here, A 1 ': Anisotropy degree A at the first setting Vtt t at the Si-th setting ZZZz: Z at the first setting a1Na4 = constant The left side of formula (8) is the evaluation for determining the setting conditions The value is J. The first term is the magnitude of the error between the calculated value of the degree of anisotropy, the second term is the error between the absolute value of the etching rate and the reference value, the third term is the degree of desirability of the setting conditions, and the fJS4 term is the setting The smaller J, which indicates the magnitude of the change in conditions over time, is the time change pattern P for which the setting condition is ro, which can be expressed by the following equation.Here, e
l'g1m Selection ratio at 1st setting' a t P F
IL is determined at the same time as P8 is determined.

603: Find r, . r, is defined as the following equation.

604: Output Pm #J*”@11.

An embodiment of the present invention will be described below with reference to FIG. Figure (a) is a hardware block diagram of an etching equipment control system used in the semiconductor device manufacturing process. )7
05, parallel plate type plasma etching equipment ff1706.

The gas flow rate, high frequency power, and stage temperature of the etching device 706 are variable and can be automatically adjusted by the microcomputer 705.

The CPU 701 has a built-in etching device control program and condition table for the control procedure explained in section 5 of the principle explanation. 1. The device setting conditions stored in the condition table are gas flow rate, high frequency power, and stage temperature. 0. Conduct experiments in advance and create a condition table.

Figure (b) is a cross-sectional view of the element processed by this system.

Shown in (c), (b) is the cross-sectional shape before the start of etching,
(c) represents the desired cross-sectional shape at the end of etching.
721 is a mask (photoresist layer), 722 is an etching target M (polysilicon layer), and 723 is a base (silicon oxide layer).

The processing during manufacturing is as follows.

1) The operator inputs the desired etching cross-sectional shape and the mask position at the start of etching on the tablet 703, and
The PU 701 reads the coordinate data.

2) The operator inputs the initial value r of the selection ratio from the keyboard.

is keyed in using the keyboard 702.

3) The operator keys in the absolute value 1vtal of the etching speed using the keyboard 702.

4) The CPU 801 determines the time change pattern P of the setting conditions of the etching apparatus and its evaluation value J, and displays them on the CPU 701.

5) The worker looks at the results of 4) and returns to 3) or 2) if unsatisfied. If you are satisfied, proceed to 6).

6) The operator instructs the start of etching using the keyboard.

7) The CPU 701 uses the etching device (170
Operate 6.

〔Effect of the invention〕

As explained above, according to the present invention, by using the degree of anisotropy of the etching rate as an intermediate variable, when a desired etching cross-sectional shape is given, it is possible to control temporal changes in the setting conditions of the etching apparatus to obtain the desired etching cross-sectional shape. The method can be easily determined quantitatively, which enables precise operation and control of the etching apparatus.

[Brief explanation of drawings]

Fig. 1 is an etched cross-sectional view of an element in an example of the prior art, Fig. 2 is an etched cross-sectional view for explaining the degree of anisotropy in the principle explanation section 1, and Fig. 3.4 is an etching cross-sectional view for explaining the degree of anisotropy in the principle explanation section 2 and 3. An etching cross-sectional view for explaining the method of solving the direction time change pattern, Figure @5 is a block diagram of the control procedure of the etching device in section 5 of the gprt explanation, and Figure 6 is a ) is a flowchart of the process, (b) is a conceptual diagram of the condition table that it refers to, FIG.
'1 t tia... (shi) (c) ccl, ) (Cn
(cL) (e)
(Chino ¥5B Koto 4 eyes

Claims (1)

[Claims] By calculating the temporal change in the degree of anisotropy of the etching rate and changing the etching conditions over time accordingly, the cross-sectional shape of the etched layer after etching is brought closer to the desired shape. Control method for etching equipment.