JPS59155923A - Manufacture of semiconductor device - Google Patents

Abstract

PURPOSE:To obtain a minute pattern regardless of the kinds and presence of a stepped difference of a semiconductor substrate by coating the upper section of the substrate with a first resin layer, cooling the surface by using a solution, which does not dissolve the resin and the boiling point thereof is lower than a solvent for the resin, and laminating a second resin layer and patterning these laminated resins. CONSTITUTION:SiO2 17 and 18 of different thickness on a semiconductor substrate 11 are connected and formed, and a positive type photosensitive resin layer 12 thicker than a stepped difference between the SiO2 is applied onto the SiO2. Photosensitizing wavelength beams are irradiated to the whole surface to make the layer 12 alkali-soluble, and the surface of the layer 12 is cooled at 15 deg.C or less by using 1, 1, 2, trichlorotrifluoroethane, etc., which do not dissolve the resin in the layer 12 and do not dissolve cellosolve acetate as a solvent for the resin. Accordingly, a degenerated layer 13, solubility thereof reduces, is generated previously on the surface of the layer 12, a thin photosensitive resin film 14 not photosensitized is applied on the layer 13, and beams are irradiated to one parts 14a of the film 14 to make them alkali-soluble. openings 15 are bored to the laminated resins through etching.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a method for manufacturing a semiconductor device, and particularly to a method for manufacturing a semiconductor device that includes a step of forming a fine pattern using a photosensitive resin film.

Conventional Structure and Problems In the process of forming a photosensitive resin film pattern on a semiconductor substrate, a positive photosensitive resin film is usually used for microfabrication in which the pattern width is 2 to 3 μm or less. As an exposure method for this positive-type photosensitive resin, a step-and-repeat projection exposure method is used because it improves yield and allows correction of anti-9 of the semiconductor substrate.

As shown in FIG. 1, this projection exposure method is such that a light beam X passing through a mask 1 is imaged onto a resist (photosensitive resin) 4 on a semiconductor substrate 3 via a lens 2. . However, since this method uses a lens 2, using a normal light beam would cause chromatic aberration and hinder pattern miniaturization, so the light beam X must be monochromatic. However, since monochromatic light is used, a standing wave is generated by the single incident light X and its reflected light from the substrate 3.

This will be explained using FIG. 2. As shown in the figure, a silicon dioxide film 6 having different film thicknesses is formed on one semiconductor substrate 6.
．． After forming γ-this silicon dioxide film 6. A photosensitive resin film 8 is applied on the finished layer. At this time, since the refractive index of the silicon dioxide films 6, 7 and the photosensitive resin film 8 are almost equal, even if light is irradiated from above the photosensitive resin film 8, the interface between the silicon dioxide films 6, 7 and the photosensitive resin film 8 No reflection occurs. Instead, the incident light interferes with the light reflected by the semiconductor substrate 5, and the silicon dioxide film 6
, 7 and the resist film 8, a standing wave 9 is generated. If the wavelength of the incident light is 4358 people, the antinodes and nodes of the standing wave are 1743.
It can be done according to the human cycle. Therefore, when the difference (d+-d2) between the film thicknesses d1 and d2 of the silicon dioxide films 6 and 7 is an odd number multiple of 743 people, the difference in light intensity at the interface with the photosensitive resin film 8 becomes large.

For this reason, for example, nodes of the standing wave 9 are formed on the surface of the silicon dioxide film 6, while antinodes of the standing wave 9 are formed on the surface of the silicon dioxide film 70. In such a case, the irradiation intensity of the incident light becomes weak on the surface of the silicon dioxide film 7, and the photosensitive resin film 8 cannot be removed sufficiently after one development and rinsing. It's 1.
When developed to form a pattern of resin film 8, silicon dioxide film 6. There also arises a problem that the pattern width of the photosensitive resin film 8 on Y is different.

In addition, when a photosensitive resin film is formed on a metal film with high reflectance such as aluminum, standing wave nodes are formed near the interface between the metal film and the photosensitive resin film, resulting in a thin film remaining. It was difficult to form a resin film pattern with high precision.

The above results in the disadvantage that, especially on a substrate having a step, the thickness of the photosensitive resin film varies greatly at the step, making it impossible to form a fine pattern with high precision.

Purpose of the Invention The present invention addresses the above-mentioned drawbacks as cancer, and provides a method for manufacturing semiconductor devices that require fine patterns, especially semiconductor devices in which patterns are formed using a projection exposure method. It is intended to provide all methods that can form the structure regardless of the difference in level.

DESCRIPTION OF THE INVENTION The present invention aims to reduce differences in pattern width on one substrate and variations in pattern width depending on the type, and to obtain a pattern with a thick resin film (
After forming a first resin layer on a semiconductor substrate, the first resin layer is treated with a solution that does not dissolve the first resin layer and has a lower boiling point than the solvent for the first resin, and then cooled. After that, a second resin layer is formed. As the first resin layer, for example, a non-photosensitive alkali-soluble resin or a photosensitive resin is applied, and then the entire surface is irradiated with light of a photosensitive wavelength to make it alkali-soluble. As the second resin layer, for example, a photosensitive resin is used and is made alkali-soluble after being selectively irradiated with a photosensitive wavelength. After this, the second resin layer and the first resin layer are treated with the same alkaline solution to form a predetermined pattern.

Description of examples An embodiment of the present invention will be described using FIG. 3.

For example, a positive photosensitive resin layer 12 is formed as a first resin on the silicon dioxide films 17 and 18 having different thicknesses formed on the semiconductor substrate 11 to be thicker than the steps of the silicon dioxide 11g and 17 and 18 (A). Next, the entire surface is irradiated with light at a photosensitive wavelength to make the resin layer 12 soluble in alkali, and then a solution that does not dissolve the first resin and has a low boiling point such as a solvent for the first resin, such as selonlube acetate, is applied. , for example, 1.1, 2. with a boiling point of 100°C or less. ) Lichlorotrifluoroethane 1 The substrate is immersed in petroleum ether or normal pentane, or this liquid is applied onto the substrate, and at least the surface of the photosensitive resin layer 12 is cooled to, for example, 15°C or less. By bringing the low boiling point solution into contact with the photosensitive resin layer 12 in this way, the temperature of the resin decreases, the solubility of the resin in the solvent decreases, and the altered layer 13 is formed (B
). Further, the solvent in the resin and the low boiling point solution are dissolved, and the solvent evaporates together with the low boiling point solution, so that the content of the solvent in the first resin layer 12 is reduced.

Next, a second resin, such as an unexposed photosensitive resin film 14, is applied thinner than the first resin layer 12.

At this time, the content of the solvent in the first resin layer 12 is reduced in the step, so that the content of the solvent in the first resin layer 14 and the first resin layer 12 are reduced.
When the resin layers 12 are mixed, it is possible to obtain a uniform resin film with less reduction in the overall thickness of the resin layers 12 and 14. That is, the first. If the second resin layers 12 and 14 are mixed a lot, the total film thickness will be smaller than the sum of the individual resin layers, but if the altered layer 13 is shaped, mixing will be less likely to occur. Decreased film thickness is less likely to occur.

Next, a photomask (not shown) is applied to the second resin layer 14.
A portion 14a is selectively irradiated with light of a photosensitive wavelength using a
After making it alkali-soluble, (C) - the alkali-soluble region 14 & of the second resin and the first resin layer 12 below it are developed and removed using an alkaline developer to form a predetermined pattern of openings 16 ( D) 0th said pattern opening part 1
From step 6, the semiconductor substrate is processed and a semiconductor device is manufactured. Note that the opening 1 is formed by the lift-off method using the pattern made of the remaining resin layers 12, 13, and 14 as a lit-off material.
A metal wiring layer or the like may be formed on 6.

Alternatively, unsensitized photosensitive resin may be used for both the first and second resin layers, and a low boiling point solution treatment may be performed immediately after coating the first resin layer, and then the second resin may be coated.

The second resin layer may be thicker than the first resin layer, but when the second resin is applied, the first resin also dissolves to some extent to form a mixed layer, so in order to make the pattern edges vertical, it is necessary to It is better to make the second resin layer thinner than the first resin layer.

Further, if a non-photosensitive alkali-soluble resin is used as the first resin layer 12, the entire surface is not required to be exposed to light, and one step becomes simple.

In the method shown in Fig. 3 above, the second resin layer can be formed on the first resin layer with a flat surface, so the method shown in Fig. 3 (C
), the nodes and antinodes of the standing wave 19 generated on the substrate act uniformly on the second resin layer 1''4 regardless of the substrate level difference, and the phenomenon shown in FIG. 2 does not occur. A resin film pattern having a similar pattern can be formed.

Further, the present invention has a large effect when a resin pattern is formed on four parts of a metal layer 20 having a high reflection coefficient, such as aluminum, formed on a SIO 2 film 17 having a step as shown in FIG. 4(A). . In the conventional one-layer resist method, the reflected light 23 from the stepped portions sensitizes even the radiation-sensitive resin under the diagonal area 22 of the photomask 21, that is, in the case of a radiation-sensitive resin film thickness of 2.4 μm, as shown in FIG. (B
), the distance d from the center of the one-step difference is 0.5
A resin film in the range of ~1.5 μm is also exposed, and when the pattern width of the thin area 22 is 2 μm or less, no horizontal resin film remains after development, and the remaining pattern width becomes Who, resulting in no pattern formation. b” becomes Noh. However, in the method of the present invention,
Since the radiation-sensitive resin film of the first layer is formed thicker than the step portion, the reflected light 23 from the step portion when irradiating the radiation-sensitive resin film of the second layer for pattern formation is reflected from the step portion. The radiation-sensitive resin pattern does not reach the resin film of the layer, and can form a radiation-sensitive resin pattern having a constant pattern width regardless of the presence or absence of a stepped portion, as shown in (2) in FIG. In addition, the first
The characteristic (■) of the present invention in Figure 0 (B) is the first. This is an example when the total film thickness of the second layer resin is 3.4 μm. Effects of the Invention According to the method of the present invention, the second and first resin layers are cooled by a low boiling point solution, so that the first resin layer Even when the solvent of the first resin and the second resin are the same, there is little mixing with the first resin when forming the second resin layer, and a multilayer film with good uniformity can be formed. In addition, when applying a low boiling point solution from a nozzle, the application of the IrJ-first resin, the application of the low boiling point solution, and the application of the second resin can be sequentially performed using the same device, which simplifies the process and equipment. Multilayer films can be easily formed. Further, in the method of the present invention, the first resin layer is formed thicker than the substrate step, and the second resin layer is formed thicker than the substrate step.
A pattern is formed on the resin layer, and the first resin is selectively developed using the second resin pattern as a mask, so the first and second resin patterns have good dimensional accuracy in relation to the level difference and type of the substrate. Patterns can be formed.

[Brief explanation of drawings]

1 and 2 are structural sectional views for explaining conventional pattern formation, and FIGS. 3(A) to 3(D) are process and cross-sectional views for explaining a pattern forming method according to an embodiment of the present invention. 4(A) is a sectional view of the present invention in the presence of reflected light, and FIG. 4(B) is a characteristic diagram of the pattern width due to reflected light. 11... Semiconductor substrate, 12.14...
Resin layer, 15... Opening part, 17.18...
Silicon dioxide film. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure 2 Figure 3 Figure 4

Claims (1)

[Claims] (1) The surface of the first resin layer formed on the semiconductor substrate is treated with a solution that does not dissolve the first resin layer and has a lower boiling point than the solvent for the first resin layer. after that, forming a second resin layer on the first resin layer; and forming a predetermined pattern on the first and second resins. Production method. Q) The first resin layer is an alkali-soluble resin, the second resin layer is a photosensitive resin, and the second resin layer is exposed to light to selectively become alkali-soluble. 2. The method of manufacturing a semiconductor device according to claim 1, wherein a predetermined pattern is formed on the second and first resin layers using an alkaline solution.