JPS58219601A - Dry etching control system - Google Patents

Abstract

PURPOSE:To minimize size variance in the manufacture of semiconductor by obtaining an optimum control system from the relation between a manipulated variable such as pressure and a gas flow rate and a controlled variable such as an etching rate and performing control. CONSTITUTION:A dry etching manufacturing device 1 is brought under the feedback control of a minor controller 2 based upon PID (proportional plus integral plus differential) control. Manipulated variable set values are supplied to the controller 2 from a manipulated variable calculating device 3. The device 3 inputs development size check data 31 from a development size checking device 4 after development in a photoresist process, controlled variable calculated value data 21 from the device 1, and finished size check data 32 from a finished size checking device 5 respectively to perform optimum calculation for determining the manipulated variable of the device 1. Designed value data 40 on an LSI to be checked is supplied externally. Control command value data 10 as plural manipulated variables based upon the optimum calculation of the device 3 are supplied to the controller 2 to control the device 1 on the basis of manipulated variable data 11.

Description

DETAILED DESCRIPTION OF THE INVENTION Object of the Invention The present invention relates to a dry etching control method in semiconductor manufacturing, and in particular to a dry etching method that minimizes dimensional variations within and between lots and at the same time allows finished dimensions to approach target values. Regarding control method.

In the conventional photoresist process in semiconductor manufacturing, the SD dimension (source) after development is performed by exposing and developing nine circuits of resist coating.

However, the SD dimension after development has large discrepancies within lots and between υnts. Therefore, it is desirable to absorb this variation and control the dry etching process without reducing the throughput to obtain the finished SD dimension.

The operating variables (operating ft) in the dry etching process include the pressure inside the device, high frequency power, and gas flow rate.

It is possible to list gas concentration, Ueno A temperature, etc.

However, in the past, as mentioned above, there was no optimal operation amount determination method for manipulating a large number of variables at the same time.
The operator was determined through repeated trial and error.

Therefore, there is a problem in that not only a lot of effort and time is required, but also the determined operation amount is not necessarily the optimum operation amount.

OBJECT OF THE INVENTION The present invention has been made in view of the above circumstances, and its purpose is to solve the above-mentioned problems in the conventional dry etching control method, and to solve the above-mentioned problems in the conventional dry etching control method, and It is an object of the present invention to provide a dry etching control system which determines the optimum operation amount for the large number of variables, thereby minimizing dimensional variations in semiconductor manufacturing and at the same time making it possible to bring the finished dimensions closer to target values.

The above object of the present invention is to control operating variables such as pressure, high frequency power, gas flow kite, gas concentration, and wafer temperature, as well as etching rate, side etching amount, and side etching variation, in a control system for a dry etching process in semiconductor manufacturing. The method is characterized in that each of the manipulated variables is controlled so that the objective function is minimized by using a model equation near the design value and an objective function created based on the relationship with the control mu.

Embodiment of the Invention FIG. 1 is an overall configuration diagram of an optimum dry etching control apparatus showing an embodiment of the present invention. In the figure, l is a dry etching manufacturing device, and a feed 7 is controlled by a minor controller 2 using PID (proportional integral-differential) control.
・It is tightly controlled. The operation amount setting value for the minor hunt roller 2 is given from an operation amount calculation device Ft 3, which will be described later.

The operator calculation bag @3 contains developed dimension inspection data 31 from the developed dimension inspection device 4 after development in the photoresist process, control amount measurement value data 21 (20) from the dry etching manufacturing device 1, and completed dimension inspection data. Completed dimension inspection data 32 is inputted from the apparatus 5, and optimal calculations for determining the operating amount of the dry etching manufacturing apparatus 1 are performed. In addition to this, as input from the outside, the target LS
Design value data 40 of I is input manually or by a host computer. 10 pieces of control target value data for a plurality of manipulated variables based on optimal calculations by the operation and song calculation device 3;
The operation amount data 11 given to the minor controller 2
The dry etching manufacturing apparatus 1 is controlled based on this.

FIG. 2 is a block diagram showing details of the manipulated variable calculation device 3. The developed dimension inspection data 31 is sent from the developed dimension inspection device 4.
Production equipment operation data from dry etching production equipment (
(including experimental data) 21 is also a completed dimension inspection device δ
The completed dimension inspection data 32 are respectively input to the pre-calculation section 3A in the manipulated variable calculation device 3. The pre-calculation section 3A is
The upper and lower limits of input data are checked to exclude abnormal data, and multiple data are averaged to create wafer-based or quadrilateral data. By inputting the post-test data 3a that has undergone such processing into the regression analysis calculation section 3B, a function indicating the relationship between each manipulated variable and the controlled variable can be created. This function data 3b is stored in the function storage section 3C.

l! , an example of data stored in the Il number storage section 3C is shown in FIG. Figure 3 shows the etching rate and side etching amount (SD after current expansion) as control variables (fI monovalent items).
dimension - finished SD dimension) and side etching variation, and pressure as the manipulated variable.

The graph shows the relationship between high-frequency power, gas flow fIt, gas concentration, and wafer temperature in combinations of variables, with the horizontal axis representing the manipulated variable and the vertical axis representing the control variable.

Here, Yl knee rate (mmAnin)
Y: Side etching force (μm) Y3: Side etching variation (μm) X1' Pressure (I"a) X2' High frequency electric power (W) X3: Gas flow Ik (OC/n1n) (%)
-F(X,,) (1-1~3.j-1~5)...
... is stored as (1). Note that each data shown in $, (1) and FIG. 3 shows the situation when each control variable is a function of one manipulated variable and the other manipulated variables are fixed at design values.

Based on the function data 30 sent from the function notation section 3C, the differential coefficient of the function near the designed value is determined by inputting the design value data 3d into the differential coefficient calculation section 3D. On the other hand, the design value data 3d and the post-development 8D dimension inspection data 3
From e, the target side etching amount is determined in the side etching meter n section 3E. The above target side etching data 3f and differential coefficient data 3g are combined with etching rate and side etching Panski design value data 3h.
By creating a constraint condition and inputting ff+(3F), a relationship matrix between the manipulated variable and the controlled variable can be created. In other words, from formula α), the relationship between all variables can be expressed as in the following formula. .

Y −f (X) ・・・・
・■Here, Y − (Y, , Y2. Y3) TX − (
Xl, X2. X3. X4. X, )' Equation (2) is set to the design value (
Around Xo, Y 0) Here, when XvV is expanded with blur, it becomes 7-λ ma... (4)
In other words, equation (4) is now 11 + "! 2 + 71 ' deviation from the design value X1 + x2 +... It can be written as follows.

f'■-y-λx-Q...
・(6) Within the constraints of the above formula (υ), the objective function is g (==)″2 (xl +xl″+c, ″+x4″
−1-c, ”) , , , , , ■ are given by the objective function creation unit 3G.

The constraint matrix data 31 is sent from the constraint condition creation section 3F, the objective function data 3j is sent from the objective function creation section 3G, and the optimal calculation section 3
By inputting data into H, the manipulated variables X1 to X, which set the side etching variation as a target value and set the etching rate and side etching amount as design values, are determined. For this reason, in equation (0) and 2■, Lagrange? i%
When λ is introduced, the Lagrange function becomes L(ma+T) ”g+TTt'(x)
...(8). That is, L −, (y−÷:c:+ x3” + x4” −
) ::,')+λ1("1 'flyl-a12"
2 '+3"3 'L4"4 '11S"5)+
λ2(y, '21xL 't2"2 'n"l
-'24"4-'25"6)+λ3('3'31"l
'311x2-a33X3 '34X4'&l
”ri...0).In equation (9), if we set, then we can obtain. Therefore, the value obtained by the above-mentioned differential coefficient calculation section) 3D is given to aij in equation (2). Input the design value y□ to yi.

By inputting λ, (1-1 to 3) can be obtained. By substituting this λ□ into the formula αD, we get
Center value of operation amount during design x1゜+X20+Xwa, x
g, +Xso can be obtained, and from this the actual operation amount can be obtained as: Minor controller 2 as value chi operation amount data 10 of the above formula α■
The minor controller 2 controls the dry etching manufacturing device 1 based on this value.

FIG. 4 shows a method of connecting the minor controller 2 and the dry etching manufacturing apparatus 1. The minor controller 2 controls the pressure of the dry etching device 1, high frequency power,
It has the function of adjusting the SD dimension of the wafer to a target value by controlling the gas flow, gas concentration, and wafer temperature using PID. The pressure is obtained by sending vacuum pump control data 1b based on pressure measurement data la from the vacuum pressure gauge IA and controlling the vacuum pump IB. The high frequency power to the electrode IT is obtained by sending voltage control data 1a based on the voltage data 10 of the voltmeter IC and controlling the voltage of the power supply ID with the voltage control device IE. The gas flow rate is obtained by sending out pulp f[U data 1f based on the gas flow rate measurement data 1e of the gas flow meter IF, and controlling the gas flow rate from the gas cylinder IG with the valve IH.

Further, the gas concentration is obtained by sending gas concentration control data 1h based on gas concentration measurement data 1g from the gas concentration meter IJ and controlling it by the gas concentration control device IK. The wafer temperature is determined by sending voltage control data 1k based on wafer temperature measurement data 1j from the thermometer IL, and
is obtained by controlling the voltage with a voltage controller IN. Note that IP represents a wafer and IQ represents a heater. In addition, the minor controller 2 controls the Ueno 1 dispensing device IR and the wafer receiving device IS by sending Ueno 1 movement instruction data 1m based on the Ueno 1 state measurement data 1) in the cassette. Control transport. The control of the environment within the dry etching manufacturing apparatus 1 and the wafer transport by the minor controller 2 shown in the above embodiment is merely an example, and it goes without saying that various other control methods can be used.

Effects of the Invention As described above, according to the present invention, in a control system for a dry etching process in semiconductor manufacturing, manipulated variables such as pressure, high frequency power, gas flow rate, gas concentration, and wafer temperature, and etching rate can be controlled.

Model equations and objectives near the design values created based on the relationship between side etching amounts and controlled variables such as side etching variations! ! By using I@, each manipulated variable is controlled so that the target lpl number is minimized, so it is possible to determine the optimal manipulated variable for the large number of variables, and it is possible to reduce dimensional variations in semiconductor manufacturing. At the same time, it is possible to minimize the completed dimensions and bring them closer to the target values. Furthermore, it has the effect of making microfabrication possible in the dry etching process, which makes it possible to stably manufacture high-performance semiconductor capacitors/bodies, which has a significant economic effect. Note that the functional form of the objective function is not limited to the one shown in the above example.
I'll take a look.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram of a device showing an embodiment of the present invention, FIG. 2 is a configuration diagram showing its main parts, FIG. 3 is a diagram showing an example of stored data, and FIG. 4 is a specific example of control. FIG. 1: Dry etching device, 2: Minor controller, 3 = Operation counter 7 device, 4: Developed size inspection device, 5
: Completed dimension inspection equipment [1fL110: Control target value data,
11: manipulated variable data, 20: controlled variable measured value data. Special applicant: Hitachi, Ltd.

Claims (1)

[Claims] (1) Pressure, high frequency power, and gas flow rate in the control method of the dry etching process in semiconductor manufacturing. The objective function is calculated using a model equation and objective function near the design value created based on the relationship between manipulated variables such as gas concentration and wafer temperature, and controlled variables such as etching rate, side etching amount, and side etching variation. A dry etching control method characterized in that each of the operation amounts is controlled such that the amount of operation is minimized. (2> The dry etching control method according to claim 1, wherein the objective function is the sum of squares of deviations of each of the manipulated variables from a design value.