JP2003035801A - Method for manufacturing solid-state image pickup device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for manufacturing a solid-state image pickup device by which a micro-lens having an excellent light condensing effect is obtained without causing any fusion between the micro-lenses even when the dimension of space between micro-lens patterns is set to <= about 0.3 μm in the case of forming the micro-lens by a thermal flow method. SOLUTION: A stage for forming the micro-lens includes a stage for forming the 1st micro-lens in checkered arrangement corresponding to half of light receiving parts, and a stage for forming the 2nd micro-lens in checkered arrangement in an area where the 1st micro-lens is not formed. The exposure is successively corrected concentrically from the center part to the periphery part of a semiconductor substrate in the case of exposure in both stages.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a solid-state image pickup device, and more particularly to a method for manufacturing a solid-state image pickup device having a narrow space between microlenses formed on the image pickup device.

[0002]

2. Description of the Related Art In recent years, solid-state image pickup devices have been miniaturized and pixels have been miniaturized for use in digital cameras and the like, and solid-state image pickup devices having a pixel size of 3 × 3 μm have also been developed. However, as such miniaturization progresses, the intensity of light incident on each light receiving portion becomes weak, and sufficient sensitivity cannot be obtained. Therefore, a method is generally used in which a microlens is formed on each light receiving portion to collect light around the light receiving portion to increase sensitivity.

Due to the miniaturization of the pixel size, it is necessary to reduce the gap between the microlenses formed on each light receiving portion as much as possible to improve the aperture ratio.
As a manufacturing technology of such a microlens, for example,
A method disclosed in JP-A-4-226073 or a method of laminating a transparent film (inorganic film or transparent resin film) on the microlens for increasing the light-collecting efficiency as in JP-A-61-177005. Or JP-A-4-303801
There is a method of forming a lens shape that is not affected by the lens aperture of a video camera as in Japanese Patent Publication No.

Usually, the microlens used for the solid-state image pickup device is formed by a heat flow method (a method of forming a lens using a photosensitive resin which is fluidized by heat treatment). In the heat flow method, the flow control during heat treatment is relatively difficult, and there is a tendency for fluctuations in lens outer dimensions and lens fusion to occur.
As a technique for avoiding this, Japanese Unexamined Patent Publication No. 5-326900 and Japanese Unexamined Patent Publication No. 6-13 have grooves formed between microlenses.
Techniques such as Japanese Patent No. 2505 are known.

The above-mentioned JP-A-4-226073 and JP-A-61
The technique of laminating an organic film on a microlens by a coating method disclosed in JP-A-0777005 and Japanese Patent Laid-Open No. 4-303801 is relatively difficult, and the microlens cannot be well coated at the time of coating, and the applied resin liquid is a microlens. There is a problem that it flows into the concave portion between and deteriorates the shape of the microlens.

Further, a method of forming an inorganic film by a PVD method such as vapor deposition or sputtering, or a CVD method is a good method for forming a uniform transparent film on a microlens.
However, in forming the transparent inorganic film by such a method, 1) sufficient countermeasures against particles such as foreign matters cannot be taken. 2)
The thermal expansion cannot be balanced with the organic resin film such as the underlying microlens or color filter, resulting in low reliability. 3) PVD equipment is expensive. 4) When the inorganic film is formed, it is difficult to deal with it because a post-process after the microlens formation is required, for example, skinning / exposing of an aluminum pad for electrical connection. There are many problems such as these, and practical application has not progressed.

Further, the technique of forming the groove portion disclosed in the above-mentioned JP-A-5-326900, JP-A-6-132505, etc., uses a high-resolution semiconductor resist before forming a microlens, or an etching technique. Is used together to form the groove portion in advance, and then the microlens is formed. Therefore, during heat flow,
Since the lens material does not flow into the groove, fusion of the microlenses is eliminated and a narrow space size is possible.

However, even if a high-resolution semiconductor resist and a stepper of a tip exposure apparatus are used, the resolution of the semiconductor resist and the tip exposure apparatus and the flow rate of the microlens material affect the narrow space size. However, the space between microlens patterns is substantially limited to a space of 0.3 to 0.4 μm.

Similarly, in a general method for manufacturing a microlens by the heat flow method, the space size between the microlens patterns is about 0.3 to 0.4 μm. If the space dimension between the microlens patterns is formed to be smaller than about 0.3 μm, fusion between the microlenses occurs. When the fusion of the microlenses occurs, the microlenses become large due to the sticking of the microlenses, and the microlenses lose their shape, resulting in lens defects. The collapse of the shape of the microlenses not only lowers the light-collecting effect, but also becomes a heavy defect leading to image unevenness and image defects.

Further, in the manufacture of the solid-state image pickup device, a method is employed in which a plurality of solid-state image pickup devices are formed by facing a single semiconductor substrate and finally diced into individual devices. ing. Therefore, in forming the microlens, it is necessary to collectively form the microlens in each element in the imposition state. When forming microlenses collectively in each element, the dimensional variation of the microlenses for each chip occurred.

[0011]

SUMMARY OF THE INVENTION The present invention has been made in view of such a problem, and when forming a microlens in a solid-state image pickup device by a heat flow method, the space between the microlens patterns is reduced. In the case where even if the thickness is about 0.3 μm or less, fusion between the microlenses does not occur, a microlens having a good light-collecting effect is obtained, and the microlenses are collectively formed on each chip in an imposition state. However, it is an object of the present invention to provide a method for manufacturing a solid-state image sensor in which variations in lens-to-lens dimensions do not occur between chips.

[0012]

According to the present invention, a photosensitive resin is applied, exposed, and developed on a semiconductor substrate formed by facing a solid-state image pickup device having a plurality of light receiving portions and charge transfer portions. When manufacturing a solid-state imaging device in which microlenses located on the light receiving portions are formed by a heat flow method consisting of heat treatment, the step of forming the microlenses is arranged in a checkered arrangement so that half of the light receiving portions are arranged. And a step of forming the second microlenses in a checkered arrangement in a region where the first microlenses are not formed so as to correspond to the remaining half of the light receiving portions. A method of manufacturing a solid-state image pickup device, which comprises sequentially performing exposure amount correction in a concentric manner from a central portion to a peripheral portion of a semiconductor substrate in exposure in a process.

Further, the present invention is the method of manufacturing a solid-state image pickup device according to the above invention, wherein in the exposure in both of the above steps,
It is a method for manufacturing a solid-state imaging device, which comprises performing exposure amount correction so that mutual space dimensions of all microlens patterns before heat treatment are 0.05 to 0.21 μm.

According to the present invention, in the method for manufacturing a solid-state image pickup device according to the present invention, either the first microlens or the second microlens is aligned with a checkered arrangement of green color filters. This is a method for manufacturing a solid-state image sensor.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION A method for manufacturing a solid-state image pickup device according to the present invention will be described below based on its embodiments. As a result of diligent studies, the present inventors have obtained a microlens having a good light-collecting effect without causing fusion between the microlenses even if the space dimension between the microlens patterns is approximately 0.3 μm or less. The manufacturing method of the solid-state image pickup device is found.

That is, a microlens located on the light receiving portion is formed on a semiconductor substrate on which a solid-state image sensor having a plurality of light receiving portions and a charge transfer portion is formed by facing, applying a photosensitive resin, exposing the microlens. When forming by a so-called heat flow method consisting of development and heat treatment, the step of forming a microlens,
A step of forming the first microlenses in a checkered arrangement so as to correspond to half of the light receiving portions, and a step of forming the first microlenses in a checkered arrangement in an area where the first microlenses are not formed so as to correspond to the remaining half of the light receiving portions 2 in the step of forming a microlens, and in the exposure in both of the above steps, the exposure amount correction is performed concentrically from the central portion of the semiconductor substrate to the peripheral portion to sequentially form the microlenses formed in the central portion of the semiconductor substrate. It was found that the dimensional variation between the lens and the microlenses formed in the peripheral portion can be formed to 0.08 μm or less.

The inventors of the present invention have made a rigorous analysis in advance of the conventional conditions when the exposure amount correction is not performed. The first reason why the fusion between the microlenses occurs at about 0.3 μm of the space dimension between the microlenses is that the variation in the space dimension between the microlens patterns in the semiconductor substrate (first variation) is can give. In the manufacturing method in which the first microlens and the second microlens are simply formed by the heat flow method, the microlens pattern formed in the central portion of the semiconductor substrate due to coating and development variations in the steps of forming each of them. And the microlens pattern formed in the peripheral portion, the space size of the microlens pattern varies by about 0.16 μm (3σ). The tendency of this variation is such that the microlens pattern in the central portion is small, and it gradually increases toward the peripheral portion. Therefore, in this method, since the above-mentioned coating and exposing steps are repeated twice, the variation in the space dimension between the microlens patterns in the central portion and the peripheral portion in the semiconductor substrate (first variation) is maximum. It becomes 0.32 μm.

The second reason is the alignment accuracy in the step of forming the first microlenses and the second microlenses. The alignment accuracy of the current stepper exposure machine is 0.01 μm to 0.05 μ at 3σ.
There is a variation of about m (second variation). The magnitude of this variation varies depending on the model of the exposure apparatus and the degree of maintenance. The space dimension between the microlens patterns is such that the central portion and the peripheral portion of the semiconductor substrate have a variation that is a total of the values of the first variation and the second variation.

That is, the space between the microlens patterns is formed to be narrower by about 0.37 μm in the peripheral portion than in the central portion of the semiconductor substrate. Therefore,
If the space between the microlens patterns is formed to a space size of 0.37 μm or less under the conventional conditions, the microlenses are fused to each other in the peripheral portion of the semiconductor substrate as shown in FIG. The lens shape at the edge portion of is deformed, and a good light condensing effect cannot be obtained. For these reasons, in the conventional microlens manufacturing method, if the space dimension between the microlens patterns is 0.37 μm or less, it is not possible to form a microlens having a good light-collecting effect, The space size between the microlens patterns has a limit of about 0.3 μm to 0.4 μm.

The inventors of the present invention sequentially increase the exposure amount in the step of forming the first microlens and the second microlens from the central portion to the peripheral portion within the semiconductor substrate, and determine the size of the first microlens and the second microlens. The microlenses could be formed to have a size of 0.08 μm or less with almost no variation within the semiconductor substrate. However, since the stepper alignment accuracy (second variation) in the step of forming the first microlens and the second microlens is 0.01 to 0.05 μm, the space dimension between the microlens patterns is 0.
If they are formed without gaps in μm, the microlenses are partially fused to each other due to the influence of this alignment accuracy, and thus they cannot be formed well. Therefore, the alignment accuracy (second variation) of the stepper exposure apparatus is 0.01 to 0.
It is necessary to provide a micro lens pattern gap of 05 μm.

The present inventors have also found that the variation in the size of the microlens pattern is related to the film thickness (lens height) of the microlens pattern. The effective lens height range of the microlens is approximately 0.8 μm to 2 μm, depending on the specifications and application of the solid-state image sensor. Corresponding to this lens height in Table 1,
The results are shown for the case where the exposure amount is corrected using the semiconductor substrate coated with the microlens material and the case where the exposure amount is not corrected by the conventional method.

[0022]

[Table 1]

As described above, the variation in the space dimension between the microlens patterns corresponding to the film thickness (lens height) of the microlens patterns (third variation).
However, it is shown that the variation becomes 1/2 or less when the exposure amount is corrected. The third variation is 0.04 to 0.16 μm, which is twice the variation in the above-described exposure amount correction, because the first microlens and the second microlens are formed in two steps. Therefore, the second variation is 0.01 to 0.05 μm.
By setting a value obtained by adding the third variation 0.04 to 0.16 μm to the space dimension between the microlens patterns, a solid-state image sensor without microlens fusion can be obtained. That is, as described in claim 2, the mutual space dimension of all the microlens patterns before heat treatment is 0.05 to 0.21 μm,
Correct the exposure amount.

In the primary color filters, the green (green) color filter having high visibility is twice as many as the number of pixels of red (red) and blue (blue) color filters (green 2: red 1: blue 1). From the viewpoint of color reproduction, it is preferable to dispose. In the present invention, as described in claim 2, it is proposed that the green color filter is arranged in a checkered pattern and either the first microlens or the second microlens is arranged so as to correspond thereto. It is a thing.

[0025]

EXAMPLES The method for manufacturing a solid-state image pickup device according to the present invention will be described in detail below with reference to examples. <Embodiment 1> The semiconductor substrate (1) used in Embodiment 1 has a chip arrangement with an imposition, as shown in FIG.
In Example 1, an 8-inch wafer was used. First, in order to flatten the surface step of the light receiving portion on the semiconductor substrate, a flattening layer was formed using a transparent resin and the surface was flattened.
Subsequently, a color filter composed of R, G, and B was formed on each chip by photolithography using a pigment dispersion resist. Furthermore, a flattening film was formed on the color filter using a transparent resin to flatten the color filter surface.

Subsequently, first microlenses were formed in a checkered pattern on the flattening layer as follows. JSR Corporation
Using a photosensitive resin (trade name: MFR380H), spin coating was performed at a coating speed of 2100 rpm, and then 90 at 100 ° C. using a hot plate.
A pre-baking of sec was performed. Further, thereafter, pattern exposure was performed on each chip by a stepper equipped with a pattern exposure photomask. When exposing
From the central chip to the peripheral chip of the semiconductor substrate shown in FIG. 1, exposure was performed concentrically in a chip-by-chip manner while sequentially increasing the exposure amount.

The exposure amount to each chip in the semiconductor substrate will be described with reference to FIG. 1. Based on the central chip X5Y5,
X5Y5 = 400msec, X6Y5 = 430msec, X7Y
5 = 460 msec, X8Y5 = 490 msec, and so on, and the exposure was performed by sequentially increasing the exposure amount of 30 msec / chip from the central portion to the peripheral portion of the semiconductor substrate. After that, development was performed for 40 seconds using an alkaline developer (TMAH: 1.2%), and then rinsed with pure water. Then, using a hot plate, 180 ℃
Heat treatment for 6 min at
Heat treatment was performed at 0 ° C. for 6 minutes to form the first microlenses in a checkered arrangement.

Subsequently, the second microlenses were formed in a checkered arrangement on the semiconductor substrate having the first microlenses formed as follows. JSR photosensitive resin (trade name: M
FR380H) was applied by spin coating at an application speed of 3500 rpm, and then prebaked at 100 ° C. for 90 seconds using a hot plate. Further, thereafter, pattern exposure was performed on each chip by a stepper equipped with a photomask for pattern exposure. At the time of exposure, as in the case of forming the first microlenses, the exposure amount was corrected in chip units from the central portion to the peripheral portion in the semiconductor substrate. Referring to FIG. 1, with reference to the central chip X5Y5, X5Y5
Y5 = 450msec, X6Y5 = 480msec, X7Y5 =
510 msec, X8Y5 = 540 msec, and so on, from the central portion of the semiconductor substrate to the peripheral portion, 3 sequentially.
The exposure was performed by increasing the exposure amount of 0 msec / chip.

After that, an alkali developing solution (TMAH: 1.
(2%) was used for development for 40 seconds, and then rinsed with pure water. Then, using a hot plate,
Heat treatment was performed at 180 ° C. for 6 minutes, and as additional heat treatment, heat treatment was performed at 200 ° C. for 6 minutes to form the second microlenses in a checkered arrangement, and finally the microlenses as shown in FIG. 3 were formed. . FIG. 3A shows the central portion of the semiconductor substrate, and FIG. 3B shows the peripheral portion.

At this time, in the formed microlenses, the space size between the adjacent microlenses is 0.1 μm in the central chip in the semiconductor substrate and 0.
The space dimension variation between the microlenses in the lot including in-plane and between devices was 0.045 μm in 3σ. In the microlenses formed by this manufacturing method, the microlenses were not fused to each other even in the central portion and the peripheral portion of the semiconductor substrate, and the microlenses having a good condensing effect could be formed.

<Comparative Example 1> The semiconductor substrate used in Comparative Example 1 has a chip arrangement as shown in FIG. In Comparative Example 1, an 8-inch wafer was used as the semiconductor substrate.
First, in order to flatten the surface step of the light receiving portion on the semiconductor substrate, a flattening layer was formed using a transparent resin and the surface was flattened. Then, using a pigment dispersion resist, R,
A color filter consisting of G and B was formed on each chip by photolithography. Furthermore, a flattening film was formed on the color filter using a transparent resin to flatten the color filter surface.

Subsequently, first microlenses were formed in a checkered pattern on the flattening layer as follows. JSR Corporation
Using a photosensitive resin (trade name: MFR380H), spin coating was performed at a coating speed of 2100 rpm, and then 90 at 100 ° C. using a hot plate.
Pre-baked sec. Further, thereafter, pattern exposure was performed on each chip by a stepper equipped with a pattern exposure photomask. The exposure was performed with a uniform exposure amount of 400 msec in the semiconductor substrate. After that, using an alkaline developer (TMAH: 1.2%),
After developing for 40 seconds, rinse with pure water was performed. Then, using a hot plate, 6 m at 180 ℃
In heat treatment was performed, and as additional heat treatment, heat treatment was performed at 200 ° C. for 6 minutes to form the first microlenses in a checkered arrangement.

Subsequently, the second microlenses were formed in a checkered arrangement on the semiconductor substrate having the first microlenses formed as follows. Coating speed 3500 rp using photosensitive resin (trade name: MFR380H) manufactured by JSR Corporation
The coating was performed by spin coating according to m, and then prebaking was performed at 100 ° C. for 90 seconds using a hot plate. Further, thereafter, exposure was carried out by a stepper using a photomask. The exposure was performed with a uniform exposure amount of 400 msec in the semiconductor substrate, as in the case of forming the first microlens.

After that, an alkaline developer (TMAH: 1.
(2%) was used for development for 40 seconds, and then rinsed with pure water. Then, using a hot plate,
The heat treatment was performed at 180 ° C. for 6 minutes, and the additional heat treatment was performed at 200 ° C. for 6 minutes to form the second microlenses in a checkered arrangement. Finally, a microlens as shown in FIG. 4 was formed.

At this time, in the formed microlenses, the space between adjacent microlenses was 0.1 μm in the central chip in the semiconductor substrate, but fusion of the microlenses occurred in the peripheral chips. did. In the microlenses formed by this manufacturing method, fusion of the microlenses occurs even in the peripheral portion of the semiconductor substrate, and
The edge shape of the microlens was deformed, and it was not possible to form a microlens having a good condensing effect. The microlens formed is shown in FIG. Figure 4 (a)
Is the central part of the semiconductor substrate, and (b) is the peripheral part.

[0036]

According to the present invention, the step of forming the microlenses corresponds to the step of forming the first microlenses in a checkered arrangement so as to correspond to half of the light receiving portions and the remaining half of the light receiving portions. , Including the step of forming the second microlenses in a checkered arrangement in the region where the first microlenses are not formed, and in the exposure in both of the above steps, the semiconductor substrate is concentrically formed from the central portion to the peripheral portion. Since the exposure amount is sequentially corrected, when the microlenses are formed by the heat flow method, even if the space dimension between the microlens patterns is about 0.3 μm or less, fusion does not occur between the microlenses, which is excellent. A method for manufacturing a solid-state image pickup device, in which a microlens having a light collecting effect is obtained.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a chip arrangement on a semiconductor substrate according to a first embodiment.

FIG. 2 is an SEM photograph of a microlens in a conventional method.

FIG. 3A is an SEM photograph of a microlens in the central portion of the semiconductor substrate in Example 1. (B) is an SEM photograph of the microlenses around the semiconductor substrate in Example 1.

FIG. 4A is an SEM photograph of a microlens in a central portion of a semiconductor substrate in Comparative Example 1. (B) is an SEM photograph of a microlens around the semiconductor substrate in Comparative Example 1.

[Explanation of symbols]

1. Semiconductor substrate

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 4M118 AA10 AB01 BA10 CA40 FA06                       GC08 GD02 GD04 GD06

Claims (3)

[Claims] 1. A semiconductor substrate on which a solid-state imaging device having a plurality of light receiving portions and a charge transfer portion is formed by facing, by a heat flow method comprising coating, exposing, developing, and heat treating a photosensitive resin. When manufacturing a solid-state imaging device having microlenses formed on the light receiving portions, the step of forming the microlenses includes the step of forming the first microlenses in a checkered arrangement so as to correspond to half of the light receiving portions. A step of forming second microlenses in a checkered arrangement in a region where the first microlenses are not formed so as to correspond to the remaining half of the light receiving portions, and in the exposure in both steps, the semiconductor substrate A method for manufacturing a solid-state image pickup device, comprising: performing exposure amount correction in a concentric manner sequentially from a central portion to a peripheral portion of the solid state image pickup device. 2. In the exposure in both the steps, the exposure amount is corrected so that the mutual space dimension of all the microlens patterns before heat treatment is 0.05 to 0.21 μm. Item 2. A method for manufacturing a solid-state image sensor according to Item 1. 3. The method according to claim 1, wherein either the first microlens or the second microlens is formed by aligning with a checkered arrangement of green color filters. Of manufacturing the solid-state image sensor of.