JP6035744B2 - Solid-state image sensor - Google Patents

Description

  The present invention relates to a microlens used for condensing a solid-state imaging device.

  In recent years, imaging devices have been widely used with the expansion of the contents of image recording, communication, and broadcasting. Various types of image pickup devices have been proposed. An image pickup device incorporating a solid-state image pickup device that has been stably manufactured with a small size, light weight, and high performance can be used as a digital camera or digital video. It has become widespread.

The solid-state imaging device has a plurality of photoelectric conversion elements that receive an optical image from a subject and convert incident light into an electrical signal. The types of photoelectric conversion elements are roughly classified into CCD (charge coupled device) type and CMOS (complementary metal oxide semiconductor) type. Also, there are two types of arrangements of photoelectric conversion elements: linear sensors (line sensors) in which photoelectric conversion elements are arranged in a row, and area sensors (surface sensors) in which photoelectric conversion elements are two-dimensionally arranged vertically and horizontally. It is divided roughly into. In any of the sensors, as the number of photoelectric conversion elements (number of pixels) increases, the captured image becomes more precise. In recent years, in particular, a method for manufacturing a solid-state imaging element having a large number of pixels at low cost has been studied. In order to increase the amount of information of an image photographed with a miniaturized solid-state imaging device, it is necessary to miniaturize and highly integrate a photoelectric conversion device serving as a light receiving unit.
When the photoelectric conversion element is miniaturized, the area of each photoelectric conversion element is reduced, and the area ratio that can be used as a light receiving portion is also reduced. And the effective sensitivity decreases. For this reason, one of the important issues in the performance required for the solid-state imaging device is to improve the sensitivity to incident light.

As a means to prevent the sensitivity of the miniaturized solid-state imaging device from being lowered, in order to efficiently capture light into the light receiving portion of the photoelectric conversion device, the light incident from the object is condensed and received by the photoelectric conversion device. A technique for forming a microlens leading to a portion on a photoelectric conversion element has been proposed. By condensing the light with the microlens and guiding it to the light receiving portion of the photoelectric conversion element, the apparent aperture ratio of the light receiving portion can be increased, and the sensitivity of the solid-state imaging device can be improved.
FIG. 3 shows a colorization method in which one colorless and transparent microlens is provided for each pixel on a color filter pixel made of a colored transparent resin layer to collect light, and the color-separated light is guided to a light receiving portion of a photoelectric conversion element. It is a schematic cross section for demonstrating the structural example of a solid-state image sensor.
The colored solid-state imaging device 1 is disposed on a light-receiving surface side surface 4 of a solid-state imaging device pixel portion in which a plurality of photoelectric conversion elements 3 regularly provided on a semiconductor substrate 2 are arranged in a plane via a transparent flattening layer 5. A plurality of colored transparent pixel patterns 8 in which a plurality of colors are repeatedly arranged are provided so as to correspond to each of the photoelectric conversion elements 3, and flattening on a plane in which the colored transparent pixel patterns 8 are arranged by the second transparent flattening layer 51 is performed. After performing, the above-described microlens 6 is provided in correspondence with the colored transparent pixel pattern 8 and the photoelectric conversion element 3, whereby the sensitivity can be improved. In addition, a structure has been proposed that reduces the sensitivity variation during imaging by optimizing the positional relationship between the colored transparent pixel pattern and the microlens (see Patent Document 1).

Various proposals have been made as a method for forming the above-described microlens, and a method utilizing deformation accompanying thermal melting of a photosensitive resin is mainly used.
For example, a flow lens type that applies photosensitivity to a transparent resin that is the material of a microlens, uniformly coats it, selectively forms a pattern by photolithography, and then uses the thermal reflow properties of the material to create a lens shape (See Patent Document 2) and a lens matrix by a photolithographic method and thermal reflow using a resist material having alkali solubility, photosensitivity, and thermal reflow on a flat layer of a transparent resin as a microlens material. There is a transfer type that forms and transfers the shape of the lens matrix to a transparent resin layer by dry etching (see Patent Document 3).

JP-A-8-316445 JP 2008-34509 A JP 2009-152315 A

  As described above, in the miniaturized solid-state imaging device, the sensitivity is improved by condensing the microlens, while the pixel size is 5 to 10 μm like the solid-state imaging device used in recent digital single-lens reflex cameras. The sensitivity can be increased by making it relatively large, such as □, so that it can cope with a fast shutter speed. In the latter case, since the purpose is to further increase the sensitivity, even a solid-state image pickup device having a large pixel is provided with a microlens in the same manner as in the case of a miniaturized solid-state image pickup device. The improvement of sensitivity can be aimed at.

  FIG. 4 is a schematic cross-sectional view for explaining a desirable light collection difference by the microlens when the pixel sizes of the solid-state imaging devices are different. FIG. 4A is a miniaturized pixel, and FIG. The case of a large pixel is shown. In any case, the distances a1 and a2 from the plane M on which the microlenses 61 and 62 are provided to the light receiving surfaces J of the photoelectric conversion elements 31 and 32 are approximately the same. In order to collect as much light as possible on the light receiving surface, the lens height h2 of the microlens 62 provided on the large pixel needs to be higher than the lens height h1 of the microlens 61 provided on the miniaturized pixel. There is.

  In order to increase the lens height by creating a large convex shape with a wide lens area, the microlenses on the large pixels described above are optimized for processing properties such as material viscosity and thermal reflow properties, and manufacturing process conditions. Is required and cannot be easily achieved. Moreover, in the case of a microlens with a high lens height, it is difficult to maintain independence from adjacent lenses by fusing in thermal reflow, and it takes a long time to etch the shape of the lens matrix by dry etching. There is a problem. If the lens height is lowered due to the above problem, the rate of light condensing by the microlens is reduced, so that it does not contribute much to the sensitivity improvement. Also, if the refractive index of the microlens material can be increased, the light condensing performance is improved even with a flat shape with a small curvature and a low lens height. Is necessary and is accompanied by further difficulties. Furthermore, if the distance a2 from the plane M on which the microlens 62 is provided to the light receiving surface J of the photoelectric conversion element 32 is designed to be longer than a1, the condensing distance can be increased, so that the photoelectric material can be obtained without greatly changing the conventional materials and process conditions. Although the light can be condensed on the light receiving surface of the conversion element, there are problems in the manufacturing process for forming a thick intermediate layer between the microlens and the light receiving surface, and more light is absorbed in the intermediate layer. Inherently, there is a problem such as reversing the sensitivity improvement of the solid-state imaging device.

  The present invention is proposed in view of the above-described problems, and the problem to be solved by the present invention is that in a solid-state imaging device having a large pixel, when the sensitivity is improved by a microlens, An object is to provide a solid-state image sensor that is highly sensitive by condensing light at a condensing distance similar to that in the case of a solid-state image sensor having fine pixels without greatly changing materials and manufacturing process conditions.

As means for solving the above-mentioned problems, the invention according to claim 1 is directed to a semiconductor substrate in which the surface is partitioned into a large number of pixels, and one photoelectric conversion element is disposed in each of the pixels, and the semiconductor substrate. 1 pixel with respect to a pixel having a polygonal outline in which a plurality of regular hexagons are densely gathered in a solid-state imaging device including a microlens that is arranged on the photoelectric conversion element and collects incident light on each of the photoelectric conversion elements The solid-state imaging device is characterized in that the plurality of microlenses arranged on the top are regular hexagons and are arranged densely in two directions.

  The present invention has a large pixel including a semiconductor substrate in which a photoelectric conversion element is disposed in each pixel, and a microlens disposed on the semiconductor substrate to collect incident light on each of the photoelectric conversion elements. In a solid-state imaging device, a plurality of microlenses are densely provided in the same plane on one pixel, so that it is the same level as that of a solid-state imaging device having fine pixels without greatly changing the material and manufacturing process conditions of the microlens. A large amount of light can be condensed at a condensing distance, and a highly sensitive solid-state imaging device can be easily provided.

It is a schematic plan view for demonstrating the example of arrangement | positioning of the micro lens on one pixel in the solid-state image sensor of this invention, (a) is an example which consists of the same shape, (b) is another which consists of a different shape. It is an example. In the solid-state image sensor of this invention, it is a schematic cross section for demonstrating an example of the condensing state in one pixel. It is a schematic cross section for demonstrating the conventional structural example of the color-ized solid-state image sensor using a microlens. It is a schematic cross section for demonstrating the difference of the desirable condensing by a microlens in case the pixel size of a solid-state image sensor differs, Comprising: (a) is a miniaturized pixel, (b) is a large pixel. Show the case. It is a schematic plan view for demonstrating another example which changed the pixel shape of the solid-state image sensor of this invention. It is a schematic plan view for demonstrating the example of the photomask for microlenses in manufacture of the solid-state image sensor of this invention, Comprising: (a) is a normal binarization photomask, (b) is a density | concentration scale. The case of a photomask with a tone is shown. It is a schematic cross section for demonstrating the 1st manufacturing method of the solid-state image sensor of this invention, Comprising: It shows in order of the process of (a)-(d). It is a schematic cross section for demonstrating the 2nd manufacturing method of the solid-state image sensor of this invention, Comprising: It shows in order of the process of (a)-(e). It is a schematic cross section for demonstrating the 3rd manufacturing method of the solid-state image sensor of this invention, Comprising: It shows in order of the process of (a)-(c). It is a schematic cross section for demonstrating the 4th manufacturing method of the solid-state image sensor of this invention, Comprising: It shows in order of the process of (a)-(d).

Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.
FIG. 1 is a schematic plan view for explaining an arrangement example of microlenses on one pixel in a solid-state imaging device of the present invention, where (a) is an example having the same shape, and (b) is a different shape. It is another example which consists of. FIG. 1A shows an example in which four microlenses 63 having a convex shape are provided adjacent to each other in the same plane on one pixel, which is the minimum unit of image input. Each of the microlenses has a rounded corner shape, but has a substantially rectangular shape in plan view, and is arranged densely so as not to impair independence from adjacent microlenses. Although there are differences depending on the manufacturing process of the microlens, it is difficult to completely eliminate the gap 7 between the microlenses. However, compared with a solid-state imaging device in which one large microlens is provided on one pixel, a solid-state imaging device in which a plurality of small microlenses are densely provided on one pixel is more A large amount of light can be collected at the same light collection distance as in the case of a solid-state imaging device having fine pixels without greatly changing the manufacturing process conditions, and a highly sensitive solid-state imaging device can be easily provided. .

  In FIG. 1B, an example in which microlenses 64, 64 ′, 65 having a convex shape are arranged adjacent to each other in the same plane on one pixel, which is the minimum unit of image input, in combination with different shapes. Show. The microlens 64 ′ is an arrangement in which the direction of the microlens 64 is changed, and the microlens 65 is an arrangement that is smaller than the microlens 64 and fills a gap in a plane on the pixel as much as possible. Each microlens has a rounded corner shape and is densely arranged so as not to impair independence from adjacent microlenses. Although there are differences depending on the manufacturing process of the microlens, it is difficult to completely eliminate the gap 71 between the microlenses. However, as compared with the solid-state imaging device in which one large microlens is provided on one pixel, a highly sensitive solid-state imaging device can be easily provided as in the case of FIG.

FIG. 2 is a schematic cross-sectional view for explaining an example of a condensing state in one pixel in the solid-state imaging device of the present invention. The figure shows a cross section of a range of one pixel with each of the photoelectric conversion elements 3 arranged in a matrix on the plane of the semiconductor substrate 2 as a pixel. On the pixel, one colored transparent pixel pattern 8 composed of a plurality of colors regularly arranged in one-to-one correspondence with one photoelectric conversion element 3 and the plurality of microlenses 63 provided thereon are configured. The whole assembly in a planar matrix form with this configuration as a unit is a solid-state imaging device. The transparent flattening layer 5 can be provided on the lower surface of the colored transparent pixel pattern 8 and the second transparent flattening layer 51 can be provided on the upper surface of the colored transparent pixel pattern 8 as in the conventional case.
In FIG. 2, incident light 91 indicated by block arrows entering from above a plurality of microlenses 63 provided in one pixel is converted into a refracted light 92 as indicated by solid arrows when passing through each microlens. Each of the three light receiving surfaces is condensed at a predetermined position. The light receiving surface of the photoelectric conversion element 3 is not effectively provided over the entire length of the pitch width of one pixel shown in the figure, but it is provided in a wider area than the conventional configuration using one microlens for one pixel. However, since it is easy to cover a plurality of condensing points, the effect of the present invention can be increased.

  FIG. 5 is a schematic plan view for explaining another example in which the pixel shape of the solid-state imaging device of the present invention is changed. For example, if one pixel of the solid-state imaging device has a pixel shape 35 having a contour in which a plurality of regular hexagons are densely gathered, in this example, four substantially regular hexagonal microlenses 60 are arranged on one pixel in plan view. Can be arranged densely. As described above with reference to FIG. 1, the microlens having a generally rectangular shape in plan view has a rounded corner, and a certain gap is generated between the microlenses. On the other hand, as shown in this example, when the regular hexagonal microlenses are arranged in plan view, the angle between the vertices is obtuse, so the gap between the microlenses due to the roundness of the corners is considerably larger. As a result, the gap 72 between the microlenses can be reduced. Further, as shown in the figure, the pixel shape 35 having a regular hexagonal gathered outline can be repeatedly arranged at a constant pitch in the vertical direction and the horizontal direction in the plane, and the corresponding colored transparent pixel patterns are also provided with the same shape. It can be formed corresponding to the position.

  In the microlens for the solid-state imaging device described in the various examples described above, a plurality of microlenses provided on one pixel are made of the same transparent resin, and each microlens is a convex surface having a high light collecting property on the same plane. It is desirable to have a shape. For example, an acrylic transparent resin material can be used as the transparent resin layer used as a microlens material. As will be described later, whether the transparent resin material used for the raw material is given photosensitivity or non-photosensitive is selected according to the microlens manufacturing process.

Hereinafter, the manufacturing process of the microlens in the manufacturing method of the solid-state image sensor of this invention is demonstrated.
FIG. 7 is a schematic cross-sectional view for explaining a first manufacturing method of the solid-state imaging device of the present invention, that is, an example in which the microlens process includes a flow lens type manufacturing process. ).
FIG. 7A is a schematic cross-sectional view before providing a microlens for a color solid-state imaging device, and is a solid-state imaging in which a plurality of photoelectric conversion elements 3 regularly provided on a semiconductor substrate 2 such as silicon are arranged in a plane. A plurality of colored transparent pixel patterns 8 in which a plurality of colors are repeatedly arranged corresponding to the photoelectric conversion element 3 are provided on the light receiving surface side surface of the element pixel portion via the transparent flattening layer 5, and the second transparent flattening layer After the flattening on the plane on which the colored transparent pixel patterns 8 are arranged by 51, the transparent resin layer 40 that should constitute the microlens is uniformly applied and formed. As the transparent resin layer 40, alkali-soluble, photosensitive and thermal reflow properties are imparted to an acrylic transparent resin material, and the film can be uniformly applied to a film thickness of 0.5 to 0.8 μm when dried using a spin coater. The drying process can be performed uniformly in a short time using a hot plate.

  Next, as shown in FIG. 7B, the photosensitive transparent resin layer 40 is patterned by a photolithography method in the exposure process. Hereinafter, an example using a positive photosensitive material will be described. The positive-type photosensitive transparent resin layer 40 is provided with a pattern of the light-shielding portion 11 corresponding to the region to be left, and the other is preliminarily patterned with the light-transmitting portion 12 as a photomask 10. For example, exposure can be performed by irradiating parallel irradiation light 9 from an i-line stepper using light having a wavelength of 365 nm from a mercury lamp light source from the back surface of the photomask 10. The exposure conditions are determined by the film thickness and sensitivity of the photosensitive transparent resin layer 40 to be used, the conditions of the developer, and the like.

  Next, a developing process is performed to form a transparent resin pattern 41 as shown in FIG. For development, TMAH (tetramethylammonium hydroxide) or an amine-based organic alkali developer can be used. The development process of the transparent resin layer 40 can be uniformly advanced by a spin development or paddle development method, and the other parts are dissolved and removed, leaving the transparent resin pattern 41 corresponding to the light shielding part 11 of the photomask 10.

  Next, as shown in FIG. 7D, the microlens 66 is formed by molding the transparent resin pattern 41 by thermal melting using the thermal reflow property of the material. As described above, the flow lens type microlens 66 can be formed in a relatively short process by using the photosensitivity and thermal reflow property of the transparent resin that is the material of the microlens. Moreover, although the microlens of the present invention is a microlens used for a solid-state imaging device having a large pixel, a plurality of small microlenses are densely provided on one pixel, so there is no need to increase the lens height. According to the microlens manufacturing method of this example, it is possible to easily provide a high-sensitivity solid-state imaging device using appropriate manufacturing conditions.

FIG. 8 is a schematic cross-sectional view for explaining a second manufacturing method of the solid-state imaging device of the present invention, that is, an example in which the microlens process has an etching transfer type manufacturing process. ).
FIG. 8A is a schematic cross-sectional view before providing a micro lens for a color solid-state image sensor, and has a configuration similar to that of FIG. 7A described above, but imparts photosensitivity as the uppermost layer in the figure. Instead of the transparent resin layer 40, a non-photosensitive transparent resin layer 46 is laminated. The transparent resin layer 46 can uniformly apply an acrylic transparent resin material to a film thickness of 0.5 to 0.8 μm when dried using a spin coater. The drying process can be performed uniformly in a short time using a hot plate. In addition, since the transparent resin material used for this example assumes the transcription | transfer by the dry etching mentioned later, the material excellent in dry etching aptitude is desirable.
Furthermore, another transparent resin layer having a different etching rate can be provided on the transparent resin layer 46. By stacking other transparent resin layers having different etching rates, the height of the microlens can be adjusted. As the material, an acrylic transparent resin material can be used.

Next, a photoresist layer 42 is formed by uniformly applying a photoresist material, which is a resin having alkali solubility, photosensitivity, and thermal reflow property, onto the transparent resin layer 46 by a spin coater. As the photoresist layer 42, an acrylic or novolac positive resist material can be uniformly applied to a dry film thickness of 0.5 to 0.8 μm using a spin coater. Drying is performed at a low temperature of 100 ° C. or lower.
Next, the substrate on which the photoresist layer 42 is formed is patterned by a photolithography method. In the exposure step, as shown in FIG. 8B, a pattern of the light-shielding portion 11 is provided corresponding to the region where the positive photoresist is to be left, and the other is used as a light-transmitting portion 12 to form a photomask 10 previously patterned. The exposure can be performed by accurately aligning with the target substrate and irradiating from the back surface of the photomask 10 with parallel irradiation light 9 from an i-line stepper that uses light of wavelength 365 nm from a mercury lamp light source, for example. The exposure conditions are determined by the film thickness and sensitivity of the photoresist material to be used, the developer conditions, and the like.

  Next, a development process is performed to form a photoresist pattern 43 as shown in FIG. For development, TMAH (tetramethylammonium hydroxide) or an amine-based organic alkali developer can be used. By the spin development or paddle development method, the development process of the photoresist layer 42 can be progressed uniformly. The photoresist pattern 43 corresponding to the light-shielding portion 11 of the photomask 10 is left, and the others are dissolved and removed.

  Next, as shown in FIG. 8D, a convex transfer microlens pattern 44 as a lens matrix is formed by forming a photoresist pattern 43 by thermal melting using the thermal reflow property of the material. Is formed on the transparent resin layer 46. For the thermal melting in this process, a hemispherical shape regulated by the surface tension at the time of material melting is good by a step baking method in which a high temperature heating at 200 ° C. is performed after heating at 160 ° C. using a hot plate device. Can be formed. It goes without saying that the heating temperature is effective for molding from the photoresist pattern 43 to the transfer microlens pattern 44 and should be determined under conditions that do not adversely affect the lower layer material including the transparent resin layer 46. There is no.

Next, the transfer microlens pattern 44 formed on the transparent resin layer 46 is etched and transferred to the transparent resin layer 46 as an etching resist to form a microlens 67 as shown in FIG. 8 (e). As an etching method, dry etching using a fluorine-based gas such as C ₃ F ₈ can be performed.
The cross-sectional shape of the microlens 67 transferred by dry etching is basically faithful to the transfer microlens pattern 44 used as the lens matrix, but the photoresist layer 42 that is the material of the transfer microlens pattern 44 and The condition varies depending on the individual case because it varies with the etching rate ratio under the actual dry etching conditions and the etching depth with the transparent resin layer 46 which is the material of the microlens 67.

  The above-described etching transfer type microlens 67 can be formed with high accuracy by using a lens matrix obtained from the photoresist layer 42 having relatively high exposure sensitivity and capable of forming a fine pattern with high accuracy. Moreover, although the microlens of the present invention is a microlens used for a solid-state imaging device having a large pixel, a plurality of small microlenses are densely provided on one pixel, so there is no need to increase the lens height. By using the manufacturing method of this example, the dry etching time can be shortened as compared with the conventional example in which the lens height is increased. Therefore, it is possible to easily provide a highly sensitive solid-state imaging device without causing an increase in process load or a reduction in quality.

FIG. 9 is a schematic cross-sectional view for explaining a third manufacturing method of the solid-state imaging device of the present invention, that is, an example in which the microlens process is not a flow lens type but includes a direct formation process using a photomask with density gradation. It is a figure and is shown in the order of steps (a) to (c).
FIG. 9A is a schematic cross-sectional view before providing a microlens for a color solid-state image sensor,
It has a configuration similar to that shown in FIG. 7A, and has a transparent resin layer 40 imparted with photosensitivity as the uppermost layer.

  Next, as shown in FIG. 9B, the photosensitive transparent resin layer 40 is patterned by the photolithography method in the exposure process. Here, with respect to the positive type photosensitive transparent resin layer 40, the density of the light-shielding portion 21 with density gradation corresponding to the region to be left in the microlens shape is provided, and the other is formed as a pattern with the light transmission portion 22 formed. A photomask 20 with gradation is prepared in advance. FIG. 6 is a schematic plan view for explaining an example of a microlens photomask in the production of the solid-state imaging device of the present invention. FIG. 6A shows a case of a normal binarized photomask. Indicates the case of a photomask provided with a density gradation. Of the two methods for manufacturing a solid-state imaging device of the present invention, the above-described two examples, that is, the first manufacturing method and the second manufacturing method, use a normal binarized photomask shown in FIG. The area of the portion 11 was a photomask having a uniform light shielding density. On the other hand, in the third manufacturing method of this example, a photomask provided with the density gradation shown in FIG. 6B is used, and, for example, the central part of each light shielding part pattern is maximum in the area of the light shielding part 21. The light-shielding part 21 with a density gradation that has a light-shielding density and the light-shielding property decreases toward the periphery. By appropriately controlling the density gradation, it is possible to realize a gradual change of the exposure amount in the photolithography method and obtain a convex curved surface after development. A photomask provided with the same density gradation can also be used in a fourth manufacturing method described later.

  Note that a normal binary photomask is formed by separating a light-shielding part pattern made of a light-shielding film such as metal chromium on the surface of a substrate such as quartz or glass having good transparency with respect to exposure light and distinguishing it from the light transmitting part. By doing so, it is produced by a generally known method. In order to form the light-shielding portion 21 with density gradation, a normal method for manufacturing a photomask with density gradation can be applied to the substrate. In other words, the method of providing a concentration gradient in the region by gradually changing the film thickness of the light-shielding metal film, etc., and the aggregate state of the fine pattern of the light-shielding film such as dot (halftone dot) arrangement and line and space For example, a gray-tone type method in which the average light-shielding density of each region of the fine pattern is changed to be changed is possible.

  In the exposure process shown in FIG. 9B, the photomask 20 with density gradation is accurately aligned with the position where the transparent resin layer 40 is to be patterned, and for example, light having a wavelength of 365 nm from a mercury lamp light source is emitted. Exposure can be performed by irradiating parallel irradiation light 9 from the i-line stepper to be used from the back surface of the photomask 10. The exposure conditions are determined by the film thickness and sensitivity of the photosensitive transparent resin layer 40 to be used, the conditions of the developer, and the like.

  Next, a developing process is performed to form a transparent resin pattern as it is as a microlens 68 as shown in FIG. For development, TMAH (tetramethylammonium hydroxide) or an amine-based organic alkali developer can be used. The development process of the transparent resin layer 40 can proceed according to the local exposure amount of the previous process by the spin development or paddle development method, and the convex shape corresponding to the light shielding part 21 with the density gradation of the photomask 20 can be obtained. Other portions are dissolved and removed so as to leave the microlens 68 having the same.

  As described above, the microlens 68 by the direct forming process using the photomask with density gradation instead of the flow lens type is formed by a relatively short process using the photosensitivity of the transparent resin that is the material of the microlens. be able to. Moreover, although the microlens of the present invention is a microlens used for a solid-state imaging device having a large pixel, a plurality of small microlenses are densely provided on one pixel, so there is no need to increase the lens height. According to the microlens manufacturing method of this example, it is possible to easily provide a high-sensitivity solid-state imaging device using appropriate manufacturing conditions.

FIG. 10 is a schematic cross-sectional view for explaining a fourth manufacturing method of the solid-state imaging device of the present invention, that is, an example in which the microlens process includes an etching transfer type manufacturing process using a photomask with density gradation. These are shown in the order of steps (a) to (d).
FIG. 10A is a schematic cross-sectional view before providing the microlens of the color solid-state imaging device, and has a configuration similar to that of FIG. 8A described above. A resin layer 46 is laminated. The transparent resin layer 46 can uniformly apply an acrylic transparent resin material to a film thickness of 0.5 to 0.8 μm when dried using a spin coater. The drying process can be performed uniformly in a short time using a hot plate. In addition, since the transparent resin material used for this example assumes the transcription | transfer by the dry etching mentioned later, the material excellent in dry etching aptitude is desirable. Furthermore, another transparent resin layer having a different etching rate can be provided on the transparent resin layer 46. By stacking other transparent resin layers having different etching rates, the height of the microlens can be adjusted. As the material, an acrylic transparent resin material can be used.

Next, a photoresist material 42 which is a resin having alkali solubility and photosensitivity is uniformly applied on the transparent resin layer 46 by a spin coater to form the photoresist layer 42. As the photoresist layer 42, an acrylic or novolac positive resist material can be uniformly applied to a dry film thickness of 0.5 to 0.8 μm using a spin coater. Drying is performed at a low temperature of 100 ° C. or lower.
Next, as shown in FIG. 10B, the photoresist layer 42 is patterned by photolithography. Here, a density gradation in which the pattern of the light-shielding part 21 with density gradation is provided on the positive photoresist layer 42 corresponding to the region to be left in the microlens shape, and the others are patterned as the light transmission part 22. The attached photomask 20 is prepared in advance similarly to the one used in FIG. In the exposure step shown in FIG. 10B, the photomask 20 with density gradation is accurately aligned with the position where the photoresist layer 42 is to be patterned, and, for example, light having a wavelength of 365 nm from a mercury lamp light source is emitted. Exposure can be performed by irradiating parallel irradiation light 9 from the i-line stepper to be used from the back surface of the photomask 10. The exposure conditions are determined by the film thickness and sensitivity of the photoresist layer 42 used, the developer conditions, and the like.

  Next, a developing process is performed to form a transfer microlens pattern 45 using a photoresist, as shown in FIG. For development, TMAH (tetramethylammonium hydroxide) or an amine-based organic alkali developer can be used. By the spin development or paddle development method, the development process of the photoresist layer 42 can be proceeded according to the local exposure amount of the previous process, and the convex shape corresponding to the light shielding part 21 with the density gradation of the photomask 20 can be obtained. Other portions are dissolved and removed so that the transfer microlens pattern 45 is left.

Next, the transfer microlens pattern 45 formed on the transparent resin layer 46 is etched and transferred to the transparent resin layer 46 as an etching resist to form a microlens 69 as shown in FIG. As an etching method, dry etching using a fluorine-based gas such as C ₃ F ₈ can be performed.
The cross-sectional shape of the microlens 69 transferred by dry etching is basically faithful to the transfer microlens pattern 45 used as the lens matrix, but the photoresist layer 42 that is the material of the transfer microlens pattern 45 and The condition varies depending on the individual case because it changes due to various influences on the etching rate ratio in the actual dry etching conditions and the etching depth with the transparent resin layer 46 that is the material of the microlens 69.

The transfer-type microlens 69 described above can be formed with high accuracy using a lens matrix obtained from the photoresist layer 42 that has relatively high exposure sensitivity and can form a fine pattern with high accuracy. Moreover, although the microlens of the present invention is a microlens used for a solid-state imaging device having a large pixel, a plurality of small microlenses are densely provided on one pixel, so there is no need to increase the lens height. By using the manufacturing method of this example, the dry etching time can be shortened as compared with the conventional example in which the lens height is increased. Therefore, it is possible to easily provide a highly sensitive solid-state imaging device without causing an increase in process load or a reduction in quality.

DESCRIPTION OF SYMBOLS 1 ... Solid-state image sensor 2 ... Semiconductor substrate 3 ... Photoelectric conversion element 4 ... Light-receiving surface side surface 5 of a solid-state image sensor pixel part ... Transparent flattening layer 6 ... Micro lens 7- .... Clearance between microlenses 8 ... Colored transparent pixel pattern 9 ... Irradiation light 10 ... Photomask 11 ... Light shielding part 12 ... Light transmission part 20 ... Photomask (density gradation) With)
21 ... Light shielding portion with density gradation 22 ... Light transmitting portions 31, 32 ... Photoelectric conversion element 35 ... Pixel shape (contour line)
40 ... Transparent resin layer (photosensitive)
41 ... Transparent resin pattern 42 ... Photoresist layer 43 ... Photoresist pattern 44, 45 ... Microlens pattern for transfer (lens matrix)
46 ... Transparent resin layer (non-photosensitive)
51 ... Second transparent flattening layer 60, 61, 62, 63, 64, 64 ', 65, 66, 67, 68, 69 ... Micro lens 71, 72 ... Gap 91 between micro lenses ... Incident light 92 ... Refracted light M ... Plane J on which microlenses are provided ... Light-receiving surfaces a1, a2 ... Distances h1, h2 between M-J ... Lens height

Claims (1)

A semiconductor substrate in which the surface is partitioned into a large number of pixels and one photoelectric conversion element is disposed in each of the pixels, and a microlens that is disposed on the semiconductor substrate and collects incident light on each of the photoelectric conversion elements. In a solid-state imaging device comprising:
A plurality of microlenses arranged on one pixel are regular hexagonal and densely arranged in two directions with respect to a pixel having a polygonal outline in which a plurality of regular hexagons are densely assembled. A solid-state imaging device.