JPS59201418A - Exposing apparatus - Google Patents

Abstract

PURPOSE:To improve total transfer accuracy by eliminating the effect of luminous intensity, developer of mask alignment and dispersion of resist film thickness on the mask transfer accuracy by optimization of exposing time. CONSTITUTION:The finished mask size WMASK and resist size WDEV after developing are input to a calculator 30 and the process parameters A, B and luminous intensity I of super high pressure mercury lamp 22 are also input. The calculator 30 calculates the optimum exposing time (t) from WMASK, WDEV and obtained process parameters A, B, I and outputs such time to the circuit 281 through a microcomputer 291. The shutter 25 is driven with the exposing time (t). As a result, the resist on the SiO2 film of wafer 27 is exposed only for the time (t). With such structure, effect of luminous intensity of mask aligners 211, 212 and resist film thickness which affects the mask transfer accuracy can be eliminated by adjusting the exposing time and thereby the transfer accuracy of exposing apparatus can be improved.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] This invention controls the amount of deviation of mask dimensions from design values by adjusting the exposure time so that the dimensions of the photoresist after development become equal to the design dimensions. related to exposure equipment.

[Technical field of invention]

Conventionally, negative type photoresists have been mainly used in photolithography processes.

However, with the recent trend towards ultra-high integration, the resolution size has increased to 2.
(~3 μm photoresist) IJ lithography technology is now required, and negative-type photoresists are being replaced by bottom-type photoresists. In addition, as resolution dimensions have become finer, photomasks have also changed from the conventional emulsion mask to a mask made of metals and their oxides, which is called a 100%-metal mask. The reason why the photomask is changing to a node mask is that the conventional Emaru Noyo 7 type mask. This is due to the fact that the resolution dimension accuracy is not good. As described above, as patterns become finer, higher resolution and dimensional accuracy are required. Currently, the mask accuracy is ±0.2~ for those used in the photolithography process.
It is about ±0.5 μm (value of 3σ). In addition, as the patterns become finer and finer, it is natural that the transfer accuracy, and the accuracy when transferring from the edge shift mask to the photoresist coated on the semiconductor substrate, must be as accurate as the accuracy of the single mask itself. has been done.

[Problems with background technology]

However, in the photolithography process, especially in the process of transferring a pattern onto the rennost using a completed mask, the precision of the mask is transferred onto the photoresist as is, so no matter how accurate the pattern is transferred, it will not be as accurate as the mask. I can't do it. In reality, it's a little worse than this. Here, the overall transfer accuracy when transferring the 7 mother turns of the photomask to the photoresist is σtot
If al, σtotal = ('mask'+cItra2)
"Here, σmask is the mask accuracy, and σ5 is the transfer accuracy. For example, when transferring a 30 μm photomask with a dimensional accuracy of ±02 μm (value of 3σ) to a photoresist, the transfer accuracy is ±05 μm. (value of 3σ), the overall transfer accuracy is ±057 μm (value of 3σ).

Although the overall transfer accuracy does not seem to be affected by the mask, in reality, when looking at each mask, the dimensions vary between 28 and 3.2 μm. There is a gap of 0.4 μm in average mask size between the largest and smallest mask dimensions.

In this way, the overall transfer accuracy is determined by the mask accuracy and the transfer accuracy, but in the past, even if the mask accuracy was improved, if the transfer accuracy was poor, the overall transfer accuracy would deteriorate. Causes of such deterioration of transfer accuracy include variations in the illuminance of the mask aligner, variations in the developer, and variations in the resist film thickness.

[Purpose of the invention]

This invention was made in view of the above points, and its purpose is to control the exposure time by controlling the effects of changes in mask aligner illumination, developer, resist film thickness, etc. that affect mask transfer accuracy. In particular, it is an object of the present invention to provide an exposure apparatus that can improve transfer accuracy without any problems and improve overall transfer accuracy.

[Summary of the invention]

By adjusting the exposure time to the optimum time, the effects of changes in mask aligner illuminance, developer changes, resist film thickness changes, etc. that affect mask transfer accuracy are eliminated, and overall transfer accuracy is improved. There is.

[Embodiments of the invention]

An exposure apparatus according to an embodiment of this invention will be described below with reference to the drawings. As shown in FIG. 1, when exposing and developing a bottom-type photoresist 11 using a mask 12, if the pattern size of the mask 12 is W and the irradiation energy is E, then the size of the pattern transferred to the photoresist 11 is The difference ΔW between the pattern size of the mask 12 and the pattern size of the mask 12 is expressed as follows. This situation is shown in Figure 2.

ΔW = A (tnF24-B)
・-==-JJ) Here, the constants A and B are nozzle parameters that change depending on the fluctuation of the developer in the photolithography process and the fluctuation of the coating film thickness. Here, as mentioned above, Δ~
V"'WDE'V-WMASK. Since we want to obtain the same value as the design value of the mask as the dimension after development, Wng
The design value of the mask is substituted as v, and the average value of the actually completed mask dimensions is substituted as WMASK. Here, if we transform the first equation, E, = EXP(B+
(WDEv-W, AsK)/A) ,,,,,・
-<2) Here, since the exposure energy E is the product of the illuminance I and the exposure time t, the second equation is modified to calculate the exposure time t, t=gXP(B-tnI+(WD, - WMA8K)/A
:) (3) As shown in the third equation, the deviation ΔW between the dimension after development and the design value of the mask can be corrected by appropriately adjusting the exposure time. By determining the process parameters A and B of the photolithography process experimentally, it is possible to eliminate the effects of variations in the developer and coating film thickness during the silisogel coating process.

Next, a case in which process parameters A and B are calculated will be specifically described. SiO2 film process number 9 parameter A,
The case of calculating B will be explained.

First, a silicon wafer is heat treated in an acidic atmosphere at 1100° C., and a silicon oxide film is formed at about 500° C. on top. Next, a bottling oxide resin was applied to the silicon oxide film to a thickness of about 1.5 μm using a spin coater.
Thereafter, heat treatment is performed on a hot plate at 105° C. for 90 seconds. Next, this sea urchin... is exposed to light using an exposure device using a nohed mask with a 30 μm ci cut-out turn and varying the exposure time. (The illuminance at this time is 502 mW/cm2.) Thereafter, after developing with a spray type photoresist developer, the developed dimensions are measured using a micrometer. The experimental results obtained at this time are shown in FIG. In this case, by changing the exposure time, the cutan dimensions at different −)fC exposure energies are obtained. The parameters A and Be are calculated below using the least squares method. As a result, not only SiO2 but also 5I
Norameter can be calculated in the same way for sN4ty1'resilicon, AL, etc. The situation is shown in the table below.

Next, an exposure apparatus according to the present invention will be explained with reference to FIG. In FIG. 3, 211 is a mask aligner. This mask aligner uses ultra-high pressure mercury run F22.
Concave mirror 23. Convex lens 24. Sha, Ri25. As a result of the mirror 26, the light emitted from the ultra-high pressure mercury lamp 22 is emitted onto the wafer 27 via the shutter 25.

Here, for example, a SiO2 film is formed on the way 27, and in order to pattern this SiO2 film, photorenost is applied onto the 8102 film, and the photoresist is exposed using a mask. The exposure time for the wafer 27 is adjusted by a shutter drive circuit 28.
It is controlled by many people.

There are multiple mask aligners and shutter drive circuits, numbered 212...

It is shown as 282, . Also, figure number 291゜29
2, . . . are microcomputers provided corresponding to the mask aligners 211, 212°, . . . for controlling the overall operation of the mask aligners.

The above microcomputers 291, 292. ... are largely controlled by the host computer 30. In fact, the computer 30 calculates the mask dimension WM A S K output from the mask dimension measuring device 31, and the dimension WDEv of the photoresist after development outputted from the developing dimension measuring device 32.
The process parameters "A" p(t B # and illuminance 6E") obtained by the process parameter measuring method are input. The computer 30 includes a mask dimension measuring device 3), a developing dimension measuring device 32, and input. mask size WMASK j development size”ogv (-r square
The optimum exposure time shown in the third equation is calculated based on the process/nine parameters ``A#, ``B'' and the illuminance ``I''.

Next, the operation of the exposure apparatus shown in FIG. 3 will be explained. For example, a case will be described in which a S s O2 film formed on the wafer 27 is patterned.

First, the dimension of the actually completed mask is measured by the mask dimension measuring device 31. This value WM A B K is input into the computer 30 . In addition, the developed size measuring device 3
7 The dimensions of the photoresist after development according to 2 are measured.

This value WDEv is input to the computer 30. Furthermore, the process parameters "A" and "B" shown in the table obtained by the process parameter measuring method and the illuminance I of the ultra-high pressure mercury lamp 22 are input to the computer 30. Here, as shown in the table, the process parameter “A#
is '0.64'', process parameter "tBs is '9.63"
”. Next, the computer 30 calculates the mask dimension W input from the mask dimension measuring device 31 and the developing dimension measuring device 32.
MAsK and mask design dimensions WD, V, process parameters obtained by the above process parameter measurement method? The optimum exposure time tt- shown in the third equation is calculated based on the parameters "A#", "B#" and the illuminance I. This exposure time t is determined by the shutter drive circuit 2 via the microcomputer 291.
It is output to 8. And this shutter drive circuit 28
1 outputs a drive signal to the shutter 25 so that the exposure time becomes t. As a result, 5i formed on the wafer 27
The resist coated on the 02 film is exposed for an exposure time t.

By the way, in the conventional photolithography process, only for manufacturing processes that require high-density patterning, the exposure time is temporarily set for 2 to 3 out of 10 pieces consisting of 20 to 30 waes. The method used was to determine the exposure time for the remaining 10 p by performing exposure and development, measuring the dimensions of the photorenost (such processing is referred to as advance). However, in the conventional method, 10 bib
However, when using the exposure apparatus according to the present application, no preliminary processing is required, and therefore the work time for each 10 shots can be done in about 30 minutes.

〔Effect of the invention〕

As detailed above, this invention eliminates the effects of changes in mask aligner illuminance, Lennost film thickness, etc. that affect mask transfer accuracy by adjusting the exposure time, resulting in good transfer accuracy. An exposure device can be provided.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing the relationship between a photorenost and a mask, FIG. 2 is a diagram showing exposure energy and pattern dimensions, and FIG. 3 is a diagram showing an exposure apparatus according to embodiments of the present invention. 211.212...Mask aligner, 25...Shutter, 27...Wafer, 2811282...Shutter drive circuit, 3...Mask dimension measuring device, 32...
Developed size measuring device.

Claims (1)

[Claims] a Nyatta drive circuit that controls the opening and closing of a shutter that adjusts the exposure time; a mask dimension measuring means that measures the dimensions of the mask; a developed dimension measuring means that measures the dimensions of the photoresist after development; a process parameter A; B'i) - Calculating process parameters measuring means, dimensions of the mask measured by the mask dimension measuring means, dimensions of the photoresist measured by the developed dimension measuring means, and the process parameters A, B and an exposure time calculation means for calculating an exposure time such that the dimensions of the photoresist after development are equal to the mask design dimensions based on the illuminance, and driving the shutter drive circuit based on the exposure time calculated by the exposure time calculation means. An exposure apparatus characterized by comprising means for