WO2021009816A1 - Solid-state imaging device - Google Patents

Solid-state imaging device Download PDF

Info

Publication number
WO2021009816A1
WO2021009816A1 PCT/JP2019/027807 JP2019027807W WO2021009816A1 WO 2021009816 A1 WO2021009816 A1 WO 2021009816A1 JP 2019027807 W JP2019027807 W JP 2019027807W WO 2021009816 A1 WO2021009816 A1 WO 2021009816A1
Authority
WO
WIPO (PCT)
Prior art keywords
solid
image sensor
state image
uneven pattern
silicon substrate
Prior art date
Application number
PCT/JP2019/027807
Other languages
French (fr)
Japanese (ja)
Inventor
横山　敏史
郁夫 水野
嘉昭 西
Original Assignee
パナソニック・タワージャズセミコンダクター株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by パナソニック・タワージャズセミコンダクター株式会社 filed Critical パナソニック・タワージャズセミコンダクター株式会社
Priority to PCT/JP2019/027807 priority Critical patent/WO2021009816A1/en
Publication of WO2021009816A1 publication Critical patent/WO2021009816A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
    • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
    • H01L27/14—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components sensitive to infrared radiation, light, electromagnetic radiation of shorter wavelength or corpuscular radiation and specially adapted either for the conversion of the energy of such radiation into electrical energy or for the control of electrical energy by such radiation
    • H01L27/144—Devices controlled by radiation
    • H01L27/146—Imager structures
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/70—SSIS architectures; Circuits associated therewith

Definitions

  • This disclosure relates to a solid-state image sensor.
  • Patent Document 1 there is a conventional example in which light is refracted on the surface of a Si substrate by forming periodic irregularities on the surface of the Si substrate to lengthen the optical path length in the Si substrate.
  • the surface of the Si substrate which is a CMOS sensor, is etched to form periodic inverted pyramid-shaped irregularities. Since light is refracted on such a Si substrate surface and the optical path is bent, the optical path becomes longer than when the Si substrate surface is flat. The incident light is reflected by the surface of the trench structure in which the oxide film or the like is embedded and is confined in the Si substrate. As a result, more incident light is absorbed in the Si substrate and the quantum efficiency is improved. This is especially effective in near infrared light.
  • the object of the present disclosure is to realize a solid-state image sensor that can obtain a better image while increasing the quantum efficiency.
  • the solid-state image sensor of the present disclosure is formed on a silicon substrate in which a plurality of pixels having a photoelectric conversion region are arranged, an insulating film formed on the silicon substrate, and an insulating film. It includes an uneven pattern and a flattening film that covers the uneven pattern.
  • the uneven pattern is made of a high refractive index material having a higher refractive index than the flattening film.
  • the incident light is scattered by the uneven pattern made of the high refractive index material, so that the optical path length in the photoelectric conversion region becomes long and the quantum efficiency is improved.
  • FIG. 1 is a cross-sectional view schematically showing an exemplary solid-state image sensor according to the embodiment of the present disclosure.
  • FIG. 2 is a plan view showing a concave-convex pattern configuration of the solid-state image sensor of FIG.
  • FIG. 3 is a diagram showing the relationship between the thickness of the uneven pattern and the quantum efficiency of the solid-state image sensor of FIG.
  • FIG. 4 is a diagram showing the relationship between the area occupancy of the convex portion and the quantum efficiency of the solid-state image sensor of FIG.
  • FIG. 5 is a diagram showing incident angle characteristics of the solid-state image sensor of FIG.
  • FIG. 6 is a diagram showing another configuration example of the uneven pattern shown in FIG.
  • FIG. 7 is a diagram showing a solid-state image sensor of a first modification of the present embodiment.
  • FIG. 8 is a diagram showing a configuration of pixel and deep trench element separation for the solid-state image sensor of the present embodiment.
  • FIG. 9 is a diagram showing another configuration of pixel and deep trench element separation for the solid-state image sensor of the present embodiment.
  • FIG. 10 is a diagram showing the relationship between the angle drop rate (2a / b) and the quantum efficiency with respect to the deep trench element separation shown in FIG.
  • FIG. 11 is a diagram showing an example of spectral characteristics of a general image sensor.
  • FIG. 12 is a diagram showing incident angle characteristics in the solid-state image sensor of Patent Document 1.
  • FIGS. 1 to 12 An embodiment of the present disclosure will be described with reference to FIGS. 1 to 12. A brief description will be given before the individual embodiments are described in detail.
  • FIG. 11 shows an example of spectral characteristics of a general image sensor.
  • the quantum efficiencies for light of each wavelength are shown by the lines B, G, and R in order.
  • the absorption coefficient of the silicon substrate decreases as the wavelength becomes longer, and the quantum efficiency decreases.
  • the quantum efficiency is about 10%. Therefore, it is desirable to increase this quantum efficiency.
  • the solid-state image sensor of the present disclosure is formed on a silicon substrate 5 in which a plurality of pixels 22 having a photoelectric conversion region 21 are arranged, an insulating film 4 formed on the silicon substrate 5, and an insulating film 4.
  • the uneven pattern 3 and the flattening film 2 covering the uneven pattern 3 are provided.
  • the uneven pattern 3 is made of a high refractive index material having a higher refractive index than the flattening film 2.
  • the incident light is scattered by the uneven pattern, so that the optical path in the photoelectric conversion region becomes long and the quantum efficiency is improved. Since the uneven pattern 3 is formed on the insulating film 4, the interface state is not destabilized in the photoelectric conversion region 21 provided on the silicon substrate 5. Moreover, the incident angle characteristic does not become irregular. Therefore, deterioration of image quality can be avoided.
  • the refractive index of the high refractive index material may be 1.9 or more. As a result, the scattering of the incident angle can be ensured due to the relationship of the refractive index with the insulating film 4.
  • polysilicon may be used as a specific high refractive index material for this purpose.
  • the thickness of the convex portion of the uneven pattern 3 may be 250 nm or more and 400 nm or less. Further, the convex portion of the concave-convex pattern 3 may have an area occupancy of 7.5% or more and 30% or less in the pixel 22. In this way, better quantum efficiency is achieved.
  • a light reflection region 8 may be provided on the side opposite to the uneven pattern 3 with respect to the photoelectric conversion region 21. Thereby, the quantum efficiency can be further improved.
  • the deep trench element separation 7 may be formed so as to surround at least a part of the photoelectric conversion region 21. As a result, the incident light is reflected at the interface between the photoelectric conversion region 21 and the deep trench element separation 7, so that the quantum efficiency is further improved.
  • the deep trench element separation 7 When viewed in the direction perpendicular to the silicon substrate 5, the deep trench element separation 7 may have an octagonal pattern in which the regions near the four corners are partially excluded from the square region.
  • the length of the diagonal line of the square area is b, and the length of one of the diagonal lines not included in the octagonal pattern at one corner of the square area.
  • a be a, and 2a may be 3% or more and 9% or less with respect to b.
  • FIG. 1 is a diagram schematically showing a cross section of the solid-state image sensor of the present embodiment in a range to which two pixels 22 are included.
  • the solid-state image sensor is a back-illuminated type and is formed by using a silicon substrate 5.
  • the silicon substrate 5 is separated by a deep trench element separation 7 (Deep Trench Isolation: DTI), and forms a photoelectric conversion region 21 for each pixel 22. Although not shown in the figure, the silicon substrate 5 continues below the photoelectric conversion region 21.
  • DTI Deep Trench Isolation
  • the insulating film 4 is formed on the silicon substrate 5.
  • the insulating film 4 is formed of, for example, HfO, Al 2 O 3 , SiO 2, or the like. By forming the insulating film 4, the interface state near the surface of the silicon substrate 5 is stabilized, and the occurrence of dark current and white scratches (white spots) is suppressed.
  • the thickness of the insulating film 4 is preferably set to 30 nm or less.
  • a metal shield 6 is formed above the deep trench element separation 7 on the insulating film 4.
  • the metal shield 6 may be formed of, for example, tungsten. Further, an uneven pattern 3 made of a high refractive index material is formed in a region on the insulating film 4 where the metal shield 6 is not formed, that is, above the photoelectric conversion region 21.
  • a flattening film 2 which is a transparent material layer made of, for example, an organic polymer resin is formed on a silicon substrate 5 so as to cover the metal shield 6 and the uneven pattern 3.
  • the upper surface of the flattening film 2 is flattened by mitigating the influence of unevenness due to the metal shield 6 and the unevenness pattern 3 of the lower layer.
  • An on-chip lens 1 made of, for example, an organic polymer resin is formed on the flattening film 2 so as to correspond to each pixel 22.
  • a color filter is not provided for the pixels.
  • a color filter may be formed between the on-chip lens 1 and the silicon substrate 5.
  • FIG. 2 shows the unevenness pattern 3 when the semiconductor device of FIG. 1 is viewed from the side of the on-chip lens 1.
  • FIG. 2 shows a range corresponding to one pixel 22.
  • the convex portion of the concave-convex pattern 3 is an independent plurality of square patterns. Such convex portions are arranged in a checkerboard shape at an angle of 45 ° with respect to the boundary between the square pixels 22.
  • the incident light on the solid-state image sensor is focused on each pixel 22 (photoelectric conversion region 21) by the on-chip lens 1.
  • the collected incident light passes through the flattening film 2 and travels toward the silicon substrate 5 to irradiate the portion of the uneven pattern 3.
  • the incident light is scattered in the silicon substrate 5 (photoelectric conversion region 21) due to reflection by the side surface of the uneven pattern 3 (light 11), diffraction generated by the uneven pattern 3 (light 12), and the like, and is an effective optical path.
  • the length becomes longer. As a result, the absorption of near-infrared light on the silicon substrate 5 increases, so that the quantum efficiency increases.
  • the quantum efficiency is further improved. That is, when the deep trench element separation 7 is formed, the incident light can be totally reflected at these interfaces due to the difference in the refractive index from the silicon substrate 5 (light 13). Therefore, the incident light is further absorbed in the photoelectric conversion region 21 in the same pixel 22, so that the quantum efficiency is improved. In addition, it prevents light from entering adjacent pixels, and it is possible to suppress color mixing and the like.
  • the deep trench element separation 7 is formed by forming a groove on the silicon substrate 5 by etching or the like and filling the groove with an insulating material.
  • the flattening film 2 a material made of an organic polymer resin having a refractive index of about 1.5 to 1.6 is generally used. Further, the flattening film 2 may be combined with silicon dioxide (SiO 2 ) or the like in addition to the organic polymer resin. Therefore, it is desirable to form the uneven pattern 3 using a material having a refractive index larger than this, for example, 1.9 or more. Specific examples thereof include Si 3 N 4 , TiO 2 , Ta 2 O 5 , SiCN, and polysilicon. In this embodiment, polysilicon is used. Since the refractive index of polysilicon is about 4, a very large difference in refractive index can be obtained with respect to the flattening film 2.
  • FIG. 3 shows the relationship between the thickness of the uneven pattern 3 and the quantum efficiency. From FIG. 3, it can be seen that the quantum efficiency of 20% or more is realized in the range where the thickness of the uneven pattern 3 is about 250 nm to 400 nm. When the thickness is 250 nm or more, the improvement in quantum efficiency due to the increase in the reflection of the incident light on the side surface of the uneven pattern 3 becomes remarkable. Further, by setting the thickness to 400 nm or less, it is possible to prevent the uneven pattern 3 from absorbing the incident light and reducing the quantum efficiency. For this reason, the thickness of the uneven pattern 3 is preferably about 250 nm to 400 nm. As shown in FIG. 3, when the thickness is about 300 nm to 375 nm, the quantum efficiency is 25% or more. Further, a quantum efficiency of about 26% is realized in the vicinity of a thickness of 350 nm.
  • FIG. 4 shows the relationship between the ratio of the area occupied by the convex portion of the uneven pattern 3 (area occupancy rate) to the area of the pixel 22 and the cooking efficiency. From FIG. 4, it can be seen that the quantum efficiency of 20% or more is realized in the range where the area occupancy of the convex portion is about 7.5% to 30%. By setting the area occupancy rate to 7.5% or more, reflection and diffraction on the side surface of the uneven pattern 3 can be surely generated, and the quantum efficiency can be remarkably improved. Further, by setting the area occupancy rate to 30% or less, it is possible to prevent the uneven pattern 3 from absorbing the incident light and reducing the quantum efficiency. For this reason, it is preferable that the area occupancy of the convex portion is about 7.5% to 30%. A quantum efficiency of about 26% is realized when the area occupancy rate is around 22%.
  • a quantum efficiency of 26% was realized at a wavelength of 940 nm.
  • the quantum efficiency of a general image sensor at a wavelength of around 940 nm is about 10%, the quantum efficiency is significantly improved as compared with this.
  • FIG. 5 is a diagram showing incident angle characteristics of the solid-state image sensor of the present embodiment. That is, the difference in quantum efficiency with respect to the incident angle is shown, where 1 is the quantum efficiency when the incident angle is 0 ° (that is, light is incident perpendicularly to the surface of the silicon substrate 5).
  • the quantum efficiency decreases almost monotonously as the incident angle increases. This is desirable in order to improve the image quality of the output image. That is, as shown in FIG. 12, when the quantum efficiency increases or decreases irregularly with respect to the incident angle, irregular striped patterns such as moire are likely to occur in the output image, and the image quality of the output image deteriorates. In the solid-state image sensor of the present embodiment, such deterioration in image quality does not occur.
  • the uneven pattern 3 having the shape in the plan view as shown in FIG. 2 (the shape when viewed from the direction perpendicular to the silicon substrate 5) has been described.
  • the shape of the uneven pattern 3 is not limited to this.
  • Other shapes are illustrated in A to E of FIG.
  • A has a pattern in which the convex portion is square as in FIG. 2, but each side is arranged vertically and horizontally along the boundary between the pixels 22.
  • B is a circular pattern
  • C is an octagonal pattern
  • D is a donut-shaped (hollow circular) pattern, all of which are arranged vertically and horizontally.
  • E is a pattern in which lines arranged in parallel are orthogonal to each other in the vertical and horizontal directions to form a grid pattern.
  • the uneven pattern 3 having shapes in various plan views also realizes the effect of improving the quantum efficiency while avoiding deterioration of image quality. It should be noted that the above is also an example, and other patterns (not shown) can be used.
  • the present invention is not limited to this, and the same effect is realized in the near infrared region where the wavelength is 700 nm to 100 nm, for example.
  • FIG. 7 is a diagram corresponding to FIG. 1 and shows a solid-state image sensor of the first modification.
  • the solid-state image sensor of FIG. 7 is provided with a light reflection region 8 below the photoelectric conversion region 21 (because it is a back-illuminated solid-state image sensor, the front side thereof) with respect to the solid-state image sensor of FIG. ..
  • the quantum efficiency can be further improved.
  • the quantum efficiency of the solid-state image sensor shown in FIG. 1 was 26%, whereas the quantum efficiency of 30% was realized by providing the light reflection region 8.
  • the material of the light reflection region 8 may be any material that can reflect light, but for example, Cu, Al, or the like can be used.
  • FIG. 8 shows a plan view of the deep trench element separation 7 of the solid-state image sensor shown in FIG. 1 as viewed from a direction perpendicular to the silicon substrate 5.
  • the deep trench element separation 7 is formed for each pixel 22 and surrounds the photoelectric conversion region 21.
  • the pixel 22 is a square region, and the deep trench element separation 7 is also provided in a square grid pattern with respect to the plurality of pixels 22.
  • the solid-state image sensor of this modified example is provided with an octagonal deep trench element separation 7a in which the corners of the square are reduced.
  • the deep trench element separation 7 of FIG. 8 in the vicinity of the four corners of the square pixel 22, each triangular region is excluded, and the pattern surrounds the octagonal photoelectric conversion region 21.
  • the distance from the center of the photoelectric conversion region 21 to the deep trench element separation 7a becomes short. Therefore, the light scattered by the uneven pattern 3 is easily reflected by the deep trench element separation 7, and the quantum efficiency is improved.
  • the quantum efficiency of the solid-state image pickup apparatus provided with the deep trench element separation 7 of FIG. 8 was 26%, whereas the quantum efficiency of 28% was achieved in this modification provided with the deep trench element separation 7a of FIG. It was realized.
  • the quantum efficiency fluctuates depending on the octagonal shape. This is shown in FIG. Let b be the diagonal length of the square pixel 22, and let a be the angle drop width (the dimension of the portion of the corner of the pixel 22 that is not included in the deep trench element separation 7a in the diagonal direction). It is defined as 2a / b.
  • FIG. 10 shows the relationship between the corner drop rate (%) and the quantum efficiency.
  • the quantum efficiency is high in the range where the corner drop rate is 3% or more and 9% or less.
  • the corner drop rate needs to have a certain magnitude, and is preferably 3% or more, for example.
  • the quantum efficiency decreases as the volume of the photoelectric conversion region 21 decreases, there is an upper limit in order to avoid this, and it is preferably 9% or less, for example.
  • the deep trench element separation 7a has an octagonal pattern
  • a pattern that excludes the corner region of the pixel 22 in a curved shape may be used.
  • the pixel 22 is not limited to a square, and may be a rectangle or a shape in which a part of the quadrangle is cut out.
  • the deep trench element separation may be formed in a shape excluding the corner portion so as to reduce the distance from the center.
  • the first and second modified examples, the shape of the uneven pattern 3 in the plan view shown in FIGS. 2 and 6, and the like can be used in various combinations. Further, although the description has been made above as a back-illuminated solid-state image sensor, this is not essential. Even in the case of the surface irradiation type, the effect of improving the quantum efficiency is realized by forming the uneven pattern 3.
  • the technique of the present disclosure is useful as a solid-state image sensor because it improves quantum efficiency and obtains a good image, especially in the near infrared region.
  • On-chip lens Flattening film 3 Concavo-convex pattern 4 Insulation film 5 Silicon substrate 6 Metal shield 7 Deep trench element separation 7a Deep trench element separation 8 Light reflection area 11 Light 12 Light 13 Light 14 Light 21 Photoelectric conversion area 22 pixels

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Power Engineering (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Electromagnetism (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Solid State Image Pick-Up Elements (AREA)
  • Light Receiving Elements (AREA)

Abstract

A solid-state imaging device according to the present invention is provided with: a silicon substrate (5) on which a plurality of pixels (22), each having a photoelectric conversion region (21), are arranged; an insulating film (4) which is formed on the silicon substrate (5); a relief pattern (3) which is formed on the insulating film (4); and a planarization film (2) which covers the relief pattern (3). The relief pattern (3) is formed of a high-refractive-index material which has a refractive index higher than the refractive index of the planarization film (2).

Description

固体撮像装置Solid-state image sensor
　本開示は、固体撮像装置に関する。 This disclosure relates to a solid-state image sensor.
　近年、近赤外光（具体的には、波長７００ｎｍ～１１００ｎｍ付近）について高感度なイメージセンサが要望されている。これは、監視用途、測距、認証、車載用途等としての利用に適しているからである。特に、波長９４０ｎｍ付近において高感度なイメージセンサの要求が大きい。これは、太陽光の波長スペクトルにおいて波長９４０ｎｍ付近の成分が小さいので、昼間の映像において太陽光の影響を受けにくいからである。 In recent years, there has been a demand for a highly sensitive image sensor for near-infrared light (specifically, a wavelength of around 700 nm to 1100 nm). This is because it is suitable for use in monitoring applications, distance measurement, authentication, in-vehicle applications, and the like. In particular, there is a great demand for a highly sensitive image sensor in the vicinity of a wavelength of 940 nm. This is because the component near the wavelength of 940 nm is small in the wavelength spectrum of sunlight, so that it is not easily affected by sunlight in daytime images.
　ところが、近赤外光はＳｉ基板には吸収されにくいので、イメージセンサにＳｉ基板を用いた場合には高い量子効率（高感度）が得られにくい。 However, since near-infrared light is not easily absorbed by the Si substrate, it is difficult to obtain high quantum efficiency (high sensitivity) when the Si substrate is used for the image sensor.
　これについて、Ｓｉ基板の表面に周期的な凹凸を形成することによりＳｉ基板表面にて光を屈折させて、Ｓｉ基板内における光路長を長くすると言う従来例がある（特許文献１）。 Regarding this, there is a conventional example in which light is refracted on the surface of a Si substrate by forming periodic irregularities on the surface of the Si substrate to lengthen the optical path length in the Si substrate (Patent Document 1).
　上記特許文献１の技術では、ＣＭＯＳセンサーであるＳｉ基板の表面をエッチングすることにより、周期的な逆ピラミッド形状の凹凸を形成している。このようなＳｉ基板表面にて光が屈折し、光路が曲げられるので、Ｓｉ基板表面が平坦な場合に比べて光路が長くなる。入射光は、酸化膜等が埋め込まれたトレンチ構造表面にて反射され、Ｓｉ基板内に閉じ込められる。この結果、入射光はＳｉ基板内にてより多く吸収され、量子効率が向上する。これは、特に近赤外光において有効である。 In the technique of Patent Document 1, the surface of the Si substrate, which is a CMOS sensor, is etched to form periodic inverted pyramid-shaped irregularities. Since light is refracted on such a Si substrate surface and the optical path is bent, the optical path becomes longer than when the Si substrate surface is flat. The incident light is reflected by the surface of the trench structure in which the oxide film or the like is embedded and is confined in the Si substrate. As a result, more incident light is absorbed in the Si substrate and the quantum efficiency is improved. This is especially effective in near infrared light.
特開２０１６－００１６３３号公報Japanese Unexamined Patent Publication No. 2016-001633
　しかしながら、特許文献１の技術では、Ｓｉ表面を直接加工するので、界面準位が不安定化して暗電流、白キズ（白点）等の増加に繋がりやすい。従って、Ｓｉ基板表面の界面準位を修理する必要がある。また、表面の形状が逆ピラミッド状であることから、入射角特性が不規則になる。これは、出力画像にシェーディング、モアレ等が生じる原因となり、良好な画像が得られにくくなる。 However, in the technique of Patent Document 1, since the Si surface is directly processed, the interface state becomes unstable, which tends to lead to an increase in dark current, white scratches (white spots), and the like. Therefore, it is necessary to repair the interface state on the surface of the Si substrate. Moreover, since the surface shape is an inverted pyramid, the incident angle characteristic becomes irregular. This causes shading, moire, etc. in the output image, and it becomes difficult to obtain a good image.
　以上から、本開示の目的は、量子効率を高めながらより良好な画像が得られる固体撮像装置を実現することである。 From the above, the object of the present disclosure is to realize a solid-state image sensor that can obtain a better image while increasing the quantum efficiency.
　前記の目的を達成するために、本開示の固体撮像装置は、光電変換領域を有する画素が複数配列されたシリコン基板と、シリコン基板上に形成された絶縁膜と、絶縁膜上に形成された凹凸パターンと、凹凸パターン上を覆う平坦化膜とを備える。凹凸パターンは、平坦化膜よりも屈折率の高い高屈折率材料からなる。 In order to achieve the above object, the solid-state image sensor of the present disclosure is formed on a silicon substrate in which a plurality of pixels having a photoelectric conversion region are arranged, an insulating film formed on the silicon substrate, and an insulating film. It includes an uneven pattern and a flattening film that covers the uneven pattern. The uneven pattern is made of a high refractive index material having a higher refractive index than the flattening film.
　本開示の固体撮像装置によると、高屈折率材料からなる凹凸パターンにより入射光が散乱されるので光電変換領域における光路長が長くなり、量子効率が向上する。 According to the solid-state image sensor of the present disclosure, the incident light is scattered by the uneven pattern made of the high refractive index material, so that the optical path length in the photoelectric conversion region becomes long and the quantum efficiency is improved.
図１は、本開示の実施形態の例示的固体撮像装置について模式的に示す断面図である。FIG. 1 is a cross-sectional view schematically showing an exemplary solid-state image sensor according to the embodiment of the present disclosure. 図２は、図１の固体撮像装置について、凹凸パターン構成を示す平面図である。FIG. 2 is a plan view showing a concave-convex pattern configuration of the solid-state image sensor of FIG. 図３は、図１の固体撮像装置について、凹凸パターンの厚さと量子効率との関係を示す図である。FIG. 3 is a diagram showing the relationship between the thickness of the uneven pattern and the quantum efficiency of the solid-state image sensor of FIG. 図４は、図１の固体撮像装置について、凸部の面積占有率と量子効率との関係を示す図である。FIG. 4 is a diagram showing the relationship between the area occupancy of the convex portion and the quantum efficiency of the solid-state image sensor of FIG. 図５は、図１の固体撮像装置について、入射角特性を示す図である。FIG. 5 is a diagram showing incident angle characteristics of the solid-state image sensor of FIG. 図６は、図２に示す凹凸パターンの他の構成例を示す図である。FIG. 6 is a diagram showing another configuration example of the uneven pattern shown in FIG. 図７は、本実施形態の第１の変形例の固体撮像装置を示す図である。FIG. 7 is a diagram showing a solid-state image sensor of a first modification of the present embodiment. 図８は、本実施形態の固体撮像装置について、画素及びディープトレンチ素子分離の構成を示す図である。FIG. 8 is a diagram showing a configuration of pixel and deep trench element separation for the solid-state image sensor of the present embodiment. 図９は、本実施形態の固体撮像装置について、画素及びディープトレンチ素子分離の他の構成を示す図である。FIG. 9 is a diagram showing another configuration of pixel and deep trench element separation for the solid-state image sensor of the present embodiment. 図１０は、図９に示すディープトレンチ素子分離に関して、角落とし率（２ａ／ｂ）と量子効率との関係を示す図である。FIG. 10 is a diagram showing the relationship between the angle drop rate (2a / b) and the quantum efficiency with respect to the deep trench element separation shown in FIG. 図１１は、一般的なイメージセンサについて、分光特性の例を示す図である。FIG. 11 is a diagram showing an example of spectral characteristics of a general image sensor. 図１２は、特許文献１の固体撮像装置において、入射角特性を示す図である。FIG. 12 is a diagram showing incident angle characteristics in the solid-state image sensor of Patent Document 1.
　本開示の実施形態について、図１～図１２を参照しながら説明する。個々の実施形態を詳しく説明する前に、概略を説明する。 An embodiment of the present disclosure will be described with reference to FIGS. 1 to 12. A brief description will be given before the individual embodiments are described in detail.
　図１１は、一般的なイメージセンサについて、分光特性の例を示す。青色光、緑色光及び赤色光に対応する画素について、各波長の光に対する量子効率を順にＢ、Ｇ及びＲの線によって示している。図１１に示す通り、近赤外領域では、波長が長くなるに従ってシリコン基板の吸収係数が小さくなり、量子効率が低くなっている。波長９４０ｎｍ付近では、量子効率は１０％程度となっている。従って、この量子効率を上げることが望ましい。 FIG. 11 shows an example of spectral characteristics of a general image sensor. For the pixels corresponding to blue light, green light, and red light, the quantum efficiencies for light of each wavelength are shown by the lines B, G, and R in order. As shown in FIG. 11, in the near-infrared region, the absorption coefficient of the silicon substrate decreases as the wavelength becomes longer, and the quantum efficiency decreases. At a wavelength of around 940 nm, the quantum efficiency is about 10%. Therefore, it is desirable to increase this quantum efficiency.
　近赤外領域における量子効率を上げるためには、シリコン基板内におけるフォトダイオードの深さを深くする方法がある。しかしながら、十分な吸収を実現するためには深さが例えば１０μｍ以上必要となる。フォトダイオードを深くすると、隣接する画素との混色が発生しやすくなるという弊害がある。 In order to increase the quantum efficiency in the near infrared region, there is a method of increasing the depth of the photodiode in the silicon substrate. However, a depth of, for example, 10 μm or more is required to achieve sufficient absorption. If the photodiode is deepened, there is an adverse effect that color mixing with adjacent pixels is likely to occur.
　また、特許文献１の技術によると、シリコン基板表面に逆ピラミッド形状が形成されているので、入射角によってシリコン基板内の光電変換領域に入射する光の量が不規則に変動する。この一例を図１２に示す。このような不規則な入射角特性は、出力画像にシェーディング、モアレ等が生じる原因となり、良好な画像が得られにくくなる。 Further, according to the technique of Patent Document 1, since the inverted pyramid shape is formed on the surface of the silicon substrate, the amount of light incident on the photoelectric conversion region in the silicon substrate fluctuates irregularly depending on the incident angle. An example of this is shown in FIG. Such irregular incident angle characteristics cause shading, moire, etc. in the output image, and it becomes difficult to obtain a good image.
　以上に対し、本開示の固体撮像装置は、光電変換領域２１を有する画素２２が複数配列されたシリコン基板５と、シリコン基板５上に形成された絶縁膜４と、絶縁膜４上に形成された凹凸パターン３と、凹凸パターン３上を覆う平坦化膜２とを備える。凹凸パターン３は、平坦化膜２よりも屈折率が高い高屈折率材料からなる。 In contrast, the solid-state image sensor of the present disclosure is formed on a silicon substrate 5 in which a plurality of pixels 22 having a photoelectric conversion region 21 are arranged, an insulating film 4 formed on the silicon substrate 5, and an insulating film 4. The uneven pattern 3 and the flattening film 2 covering the uneven pattern 3 are provided. The uneven pattern 3 is made of a high refractive index material having a higher refractive index than the flattening film 2.
　このような固体撮像装置によると、凹凸パターンによって入射光が散乱されるので、光電変換領域における光路が長くなり、量子効率が向上する。凹凸パターン３は絶縁膜４上に形成されているので、シリコン基板５に設けられた光電変換領域２１において、界面準位の不安定化等を生じることはない。また、入射角特性が不規則になることもない。従って、画質の低下を避けることができる。 According to such a solid-state image sensor, the incident light is scattered by the uneven pattern, so that the optical path in the photoelectric conversion region becomes long and the quantum efficiency is improved. Since the uneven pattern 3 is formed on the insulating film 4, the interface state is not destabilized in the photoelectric conversion region 21 provided on the silicon substrate 5. Moreover, the incident angle characteristic does not become irregular. Therefore, deterioration of image quality can be avoided.
　高屈折率材料の屈折率は、１．９以上としても良い。これにより、絶縁膜４との屈折率の関係から、入射角の散乱を寄り確実にすることができる。このための具体的な高屈折率材料としては、ポリシリコンとしても良い。 The refractive index of the high refractive index material may be 1.9 or more. As a result, the scattering of the incident angle can be ensured due to the relationship of the refractive index with the insulating film 4. As a specific high refractive index material for this purpose, polysilicon may be used.
　凹凸パターン３の凸部の厚さは、２５０ｎｍ以上で且つ４００ｎｍ以下としても良い。また、凹凸パターン３の凸部は、画素２２内において、７．５％以上で且つ３０％以下の面積占有率を有するようにしても良い。このようにすると、より優れた量子効率が実現する。 The thickness of the convex portion of the uneven pattern 3 may be 250 nm or more and 400 nm or less. Further, the convex portion of the concave-convex pattern 3 may have an area occupancy of 7.5% or more and 30% or less in the pixel 22. In this way, better quantum efficiency is achieved.
　光電変換領域２１に対して凹凸パターン３とは反対側に、光反射領域８が設けられていても良い。これにより、更に量子効率を向上させることができる。 A light reflection region 8 may be provided on the side opposite to the uneven pattern 3 with respect to the photoelectric conversion region 21. Thereby, the quantum efficiency can be further improved.
　画素２２において、光電変換領域２１の少なくとも一部を取り囲むようにディープトレンチ素子分離７が形成されていても良い。これにより、光電変換領域２１とディープトレンチ素子分離７との界面において入射光が反射されるので、量子効率が更に向上する。 In the pixel 22, the deep trench element separation 7 may be formed so as to surround at least a part of the photoelectric conversion region 21. As a result, the incident light is reflected at the interface between the photoelectric conversion region 21 and the deep trench element separation 7, so that the quantum efficiency is further improved.
　シリコン基板５に垂直な方向に見たとき、ディープトレンチ素子分離７は、正方形の領域から４つの角付近の領域をそれぞれ部分的に除外した八角形のパターンを有していても良い。 When viewed in the direction perpendicular to the silicon substrate 5, the deep trench element separation 7 may have an octagonal pattern in which the regions near the four corners are partially excluded from the square region.
　また、シリコン基板５に垂直な方向に見たとき、前記正方形の領域の対角線の長さをｂ、対角線のうち、前記正方形の領域の１つの角において八角形のパターンに含まれない部分の長さをａとして、ｂに対して２ａが３％以上で且つ９％以下となるようにしても良い。 Further, when viewed in the direction perpendicular to the silicon substrate 5, the length of the diagonal line of the square area is b, and the length of one of the diagonal lines not included in the octagonal pattern at one corner of the square area. Let a be a, and 2a may be 3% or more and 9% or less with respect to b.
　このようにすると、量子効率を更に向上することができる。 By doing so, the quantum efficiency can be further improved.
　　（第１の実施形態）
　第１の実施形態について、図面を参照して説明する。図１は、本実施形態の固体撮像装置について、画素２２が２つ含まれる程度の範囲の断面を模式的に示す図である。 (First Embodiment)
The first embodiment will be described with reference to the drawings. FIG. 1 is a diagram schematically showing a cross section of the solid-state image sensor of the present embodiment in a range to which two pixels 22 are included.
　固体撮像装置は、裏面照射型であって、シリコン基板５を用いて形成されている。シリコン基板５は、ディープトレンチ素子分離７（Deep Trench Isolation：ＤＴＩ）によって区切られ、画素２２毎の光電変換領域２１を形成している。尚、図には示していないが、光電変換領域２１の下方にもシリコン基板５は続いている。 The solid-state image sensor is a back-illuminated type and is formed by using a silicon substrate 5. The silicon substrate 5 is separated by a deep trench element separation 7 (Deep Trench Isolation: DTI), and forms a photoelectric conversion region 21 for each pixel 22. Although not shown in the figure, the silicon substrate 5 continues below the photoelectric conversion region 21.
　シリコン基板５上には絶縁膜４が形成されている。絶縁膜４は、例えば、ＨｆＯ、Ａｌ２Ｏ３、ＳｉＯ２等を材料として形成される。絶縁膜４を形成することにより、シリコン基板５の表面付近における界面準位を安定化し、暗電流及び白キズ（白点）の発生を抑制している。尚、絶縁膜４の厚さは、３０ｎｍ以下に設定するのが好ましい。 An insulating film 4 is formed on the silicon substrate 5. The insulating film 4 is formed of, for example, HfO, Al 2 O 3 , SiO 2, or the like. By forming the insulating film 4, the interface state near the surface of the silicon substrate 5 is stabilized, and the occurrence of dark current and white scratches (white spots) is suppressed. The thickness of the insulating film 4 is preferably set to 30 nm or less.
　絶縁膜４上において、ディープトレンチ素子分離７の上方に、メタルシールド６が形成されている。メタルシールド６は、例えばタングステンにより形成すれば良い。また、絶縁膜４上のメタルシールド６が形成されない領域、つまり、光電変換領域２１の上方には、高屈折率材料からなる凹凸パターン３が形成されている。 A metal shield 6 is formed above the deep trench element separation 7 on the insulating film 4. The metal shield 6 may be formed of, for example, tungsten. Further, an uneven pattern 3 made of a high refractive index material is formed in a region on the insulating film 4 where the metal shield 6 is not formed, that is, above the photoelectric conversion region 21.
　メタルシールド６及び凹凸パターン３を覆うように、シリコン基板５上に、例えば有機高分子樹脂からなる透明材料層である平坦化膜２が形成されている。当該平坦化膜２の上面は、下層のメタルシールド６及び凹凸パターン３等による凹凸の影響を緩和して平坦になっている。平坦化膜２上には、画素２２毎に対応するように例えば有機高分子樹脂からなるオンチップレンズ１が形成されている。 A flattening film 2 which is a transparent material layer made of, for example, an organic polymer resin is formed on a silicon substrate 5 so as to cover the metal shield 6 and the uneven pattern 3. The upper surface of the flattening film 2 is flattened by mitigating the influence of unevenness due to the metal shield 6 and the unevenness pattern 3 of the lower layer. An on-chip lens 1 made of, for example, an organic polymer resin is formed on the flattening film 2 so as to correspond to each pixel 22.
　本実施形態では、近赤外光による撮像を行う撮像装置の構造を例示しているので、画素に対してカラーフィルタは設けられていない。しかし、可視光による撮像も行う撮像装置の場合、オンチップレンズ１とシリコン基板５との間にカラーフィルタを形成しても良い。 In this embodiment, since the structure of the imaging device that performs imaging with near-infrared light is illustrated, a color filter is not provided for the pixels. However, in the case of an imaging device that also performs imaging with visible light, a color filter may be formed between the on-chip lens 1 and the silicon substrate 5.
　次に、図２は、図１の半導体装置をオンチップレンズ１の側から見たときの凹凸パターン３を示している。図２では、１つの画素２２に対応する範囲を示す。図２の例では、凹凸パターン３の凸部は、独立した複数の正方形のパターンである。このような凸部が、正方形である画素２２同士の境界に対して４５°傾いてチェッカーボード状に配置されている。 Next, FIG. 2 shows the unevenness pattern 3 when the semiconductor device of FIG. 1 is viewed from the side of the on-chip lens 1. FIG. 2 shows a range corresponding to one pixel 22. In the example of FIG. 2, the convex portion of the concave-convex pattern 3 is an independent plurality of square patterns. Such convex portions are arranged in a checkerboard shape at an angle of 45 ° with respect to the boundary between the square pixels 22.
　固体撮像装置に対する入射光は、オンチップレンズ１によってそれぞれの画素２２（光電変換領域２１）に対して集光される。集光された入射光は、平坦化膜２を透過して、シリコン基板５に向かって進み、凹凸パターン３の部分に照射される。入射光は、凹凸パターン３の側面による反射（光１１）、凹凸パターン３により発生する回折（光１２）等により、シリコン基板５（光電変換領域２１）内に対して散乱し、実効的な光路長が長くなる。この結果、シリコン基板５における近赤外光の吸収が増えるので、量子効率が大きくなる。 The incident light on the solid-state image sensor is focused on each pixel 22 (photoelectric conversion region 21) by the on-chip lens 1. The collected incident light passes through the flattening film 2 and travels toward the silicon substrate 5 to irradiate the portion of the uneven pattern 3. The incident light is scattered in the silicon substrate 5 (photoelectric conversion region 21) due to reflection by the side surface of the uneven pattern 3 (light 11), diffraction generated by the uneven pattern 3 (light 12), and the like, and is an effective optical path. The length becomes longer. As a result, the absorption of near-infrared light on the silicon substrate 5 increases, so that the quantum efficiency increases.
　また、画素２２間にディープトレンチ素子分離７を形成しておくことにより、量子効率は更に向上する。つまり、ディープトレンチ素子分離７が形成されていると、シリコン基板５との屈折率の違いにより、これらの界面において入射光を全反射させ得る（光１３）。従って、入射光は同じ画素２２内の光電変換領域２１において更に吸収されるので、量子効率が向上する。また、光が隣接する画素に浸入することを防ぐことになり、混色等を抑制することもできる。 Further, by forming the deep trench element separation 7 between the pixels 22, the quantum efficiency is further improved. That is, when the deep trench element separation 7 is formed, the incident light can be totally reflected at these interfaces due to the difference in the refractive index from the silicon substrate 5 (light 13). Therefore, the incident light is further absorbed in the photoelectric conversion region 21 in the same pixel 22, so that the quantum efficiency is improved. In addition, it prevents light from entering adjacent pixels, and it is possible to suppress color mixing and the like.
　尚、ディープトレンチ素子分離７は、シリコン基板５にエッチング等により溝を形成し、当該溝に絶縁材料を充填することにより形成される。 The deep trench element separation 7 is formed by forming a groove on the silicon substrate 5 by etching or the like and filling the groove with an insulating material.
　凹凸パターン３による反射、回折を増やすためには、凹凸パターン３と、平坦化膜２との屈折率の差を大きくすることが望ましい。平坦化膜２については、一般には、有機高分子樹脂からなる屈折率が１．５～１．６程度の材料が用いられる。また、平坦化膜２について、有機高分子樹脂の他に二酸化シリコン（ＳｉＯ２）等を組み合わせる場合もある。
そこで、屈折率がこれよりも大きく、例えば１．９以上である材料を用いて凹凸パターン３を形成することが望ましい。具体的にはＳｉ３Ｎ４、ＴｉＯ２、Ｔａ２Ｏ５、ＳｉＣＮ、ポリシリコンなどが挙げられる。本実施形態では、ポリシリコンを用いる。ポリシリコンの屈折率は４程度であるから、平坦化膜２に対して非常に大きな屈折率差が得られる。 In order to increase reflection and diffraction by the uneven pattern 3, it is desirable to increase the difference in refractive index between the uneven pattern 3 and the flattening film 2. As the flattening film 2, a material made of an organic polymer resin having a refractive index of about 1.5 to 1.6 is generally used. Further, the flattening film 2 may be combined with silicon dioxide (SiO 2 ) or the like in addition to the organic polymer resin.
Therefore, it is desirable to form the uneven pattern 3 using a material having a refractive index larger than this, for example, 1.9 or more. Specific examples thereof include Si 3 N 4 , TiO 2 , Ta 2 O 5 , SiCN, and polysilicon. In this embodiment, polysilicon is used. Since the refractive index of polysilicon is about 4, a very large difference in refractive index can be obtained with respect to the flattening film 2.
　次に、図３は、凹凸パターン３の厚さと量子効率との関係を示している。図３から、凹凸パターン３の厚さが２５０ｎｍ～４００ｎｍ程度の範囲において、２０％以上の量子効率が実現されていることが分かる。厚さを２５０ｎｍ以上とすると、凹凸パターン３の側面における入射光の反射が増加することによる量子効率の向上が顕著になる。また、厚さを４００ｎｍ以下とすることにより、凹凸パターン３が入射光を吸収してしまって量子効率を減少させることを抑制できる。このような理由から、凹凸パターン３の厚さを２５０ｎｍ～４００ｎｍ程度とするのが良い。尚、図３に示す通り、厚さを３００ｎｍ～３７５ｎｍ程度とすると、量子効率は２５％以上となる。また、厚さ３５０ｎｍ付近において、量子効率２６％程度が実現している。 Next, FIG. 3 shows the relationship between the thickness of the uneven pattern 3 and the quantum efficiency. From FIG. 3, it can be seen that the quantum efficiency of 20% or more is realized in the range where the thickness of the uneven pattern 3 is about 250 nm to 400 nm. When the thickness is 250 nm or more, the improvement in quantum efficiency due to the increase in the reflection of the incident light on the side surface of the uneven pattern 3 becomes remarkable. Further, by setting the thickness to 400 nm or less, it is possible to prevent the uneven pattern 3 from absorbing the incident light and reducing the quantum efficiency. For this reason, the thickness of the uneven pattern 3 is preferably about 250 nm to 400 nm. As shown in FIG. 3, when the thickness is about 300 nm to 375 nm, the quantum efficiency is 25% or more. Further, a quantum efficiency of about 26% is realized in the vicinity of a thickness of 350 nm.
　次に、図４は、画素２２の面積に対して、凹凸パターン３の凸部が占有している面積の比率（面積占有率）と、料理効率との関係を示している。図４から、凸部の面積占有率が７．５％～３０％程度の範囲において、２０％以上の量子効率が実現されていることが分かる。面積占有率を７．５％以上とすることにより、凹凸パターン３の側面における反射、回折を確実に生じさせて、量子効率を顕著に向上できる。また、面積占有率を３０％以下とすることにより、凹凸パターン３が入射光を吸収してしまって量子効率を減少させることを抑制できる。このような理由から、凸部の面積占有率を７．５％～３０％程度とするのが良い。尚、面積占有率２２％付近において、量子効率２６％程度が実現している。 Next, FIG. 4 shows the relationship between the ratio of the area occupied by the convex portion of the uneven pattern 3 (area occupancy rate) to the area of the pixel 22 and the cooking efficiency. From FIG. 4, it can be seen that the quantum efficiency of 20% or more is realized in the range where the area occupancy of the convex portion is about 7.5% to 30%. By setting the area occupancy rate to 7.5% or more, reflection and diffraction on the side surface of the uneven pattern 3 can be surely generated, and the quantum efficiency can be remarkably improved. Further, by setting the area occupancy rate to 30% or less, it is possible to prevent the uneven pattern 3 from absorbing the incident light and reducing the quantum efficiency. For this reason, it is preferable that the area occupancy of the convex portion is about 7.5% to 30%. A quantum efficiency of about 26% is realized when the area occupancy rate is around 22%.
　以上のように、本実施形態の例の固体撮像装置では、波長９４０ｎｍにおいて量子効率２６％が実現された。図１１に示した通り、一般的なイメージセンサにおいて波長９４０ｎｍ付近における量子効率は１０％程度であるから、これに比べると大幅に量子効率が改善している。 As described above, in the solid-state image sensor of the example of the present embodiment, a quantum efficiency of 26% was realized at a wavelength of 940 nm. As shown in FIG. 11, since the quantum efficiency of a general image sensor at a wavelength of around 940 nm is about 10%, the quantum efficiency is significantly improved as compared with this.
　次に、図５は、本実施形態の固体撮像装置について、入射角特性を示す図である。つまり、入射角０°の（つまり、シリコン基板５の表面に対して垂直に光が入射する）際の量子効率を１として、入射角に対する量子効率の違いを示す。 Next, FIG. 5 is a diagram showing incident angle characteristics of the solid-state image sensor of the present embodiment. That is, the difference in quantum efficiency with respect to the incident angle is shown, where 1 is the quantum efficiency when the incident angle is 0 ° (that is, light is incident perpendicularly to the surface of the silicon substrate 5).
　図５に示される通り、入射角の増加に従って量子効率は概ね単調に低下している。これは、出力画像の画質を向上させるために望ましい。つまり、図１２に示すように入射角に対して量子効率が不規則に増減する場合には、出力画像にモアレ等の不規則な縞状パターンが発生しやすく、出力画像の画質が低下する。本実施形態の固体撮像装置では、このような画質低下は生じない。 As shown in FIG. 5, the quantum efficiency decreases almost monotonously as the incident angle increases. This is desirable in order to improve the image quality of the output image. That is, as shown in FIG. 12, when the quantum efficiency increases or decreases irregularly with respect to the incident angle, irregular striped patterns such as moire are likely to occur in the output image, and the image quality of the output image deteriorates. In the solid-state image sensor of the present embodiment, such deterioration in image quality does not occur.
　また、本実施形態の固体撮像装置において、暗電流及び白キズ（白点）の増加も無いことが確認された。これは、シリコン基板５の表面には凹凸の加工等をしていないので、界面準位が安定していることによると考えられる。シリコン基板５上に絶縁膜４を形成し、その上に凹凸パターン３を形成する構成であることから、暗電流、白キズを増加させること無く量子効率を向上させることができる。 It was also confirmed that there was no increase in dark current and white scratches (white spots) in the solid-state image sensor of the present embodiment. It is considered that this is because the surface of the silicon substrate 5 is not processed to have irregularities, so that the interface state is stable. Since the insulating film 4 is formed on the silicon substrate 5 and the uneven pattern 3 is formed on the insulating film 4, the quantum efficiency can be improved without increasing the dark current and the white scratches.
　尚、以上では、図２に示すような平面図における形状（シリコン基板５に対して垂直な方向から見た場合の形状）を有する凹凸パターン３を説明した。しかし、凹凸パターン３の形状としては、これには限らない。図６のＡ～Ｅには、他の形状を例示する。 In the above, the uneven pattern 3 having the shape in the plan view as shown in FIG. 2 (the shape when viewed from the direction perpendicular to the silicon substrate 5) has been described. However, the shape of the uneven pattern 3 is not limited to this. Other shapes are illustrated in A to E of FIG.
　Ａは、図２と同様に凸部が正方形のパターンであるが、各辺は画素２２同士の境界に沿う向きに縦横に並んでいる。Ｂは円形のパターン、Ｃは八角形のパターン、Ｄはドーナツ状（中抜きの円形）のパターンであり、いずれも縦横に並んでいる。Ｅは、平行に並ぶ線が縦横に直交して格子状になったパターンである。 A has a pattern in which the convex portion is square as in FIG. 2, but each side is arranged vertically and horizontally along the boundary between the pixels 22. B is a circular pattern, C is an octagonal pattern, and D is a donut-shaped (hollow circular) pattern, all of which are arranged vertically and horizontally. E is a pattern in which lines arranged in parallel are orthogonal to each other in the vertical and horizontal directions to form a grid pattern.
　このような様々な平面図における形状を有する凹凸パターン３によっても、画質の低下を避けながら量子効率を向上する効果が実現する。尚、以上も例示であって、図示していない他のパターンを用いることも可能である。 The uneven pattern 3 having shapes in various plan views also realizes the effect of improving the quantum efficiency while avoiding deterioration of image quality. It should be noted that the above is also an example, and other patterns (not shown) can be used.
　また、以上では特に波長９４０ｎｍの場合について説明したが、これには限定されず、例えば波長７００ｎｍ～１００ｎｍの近赤外領域において同様の効果が実現する。 Further, the case where the wavelength is 940 nm has been described above, but the present invention is not limited to this, and the same effect is realized in the near infrared region where the wavelength is 700 nm to 100 nm, for example.
　　（第１の変形例）
　次に、実施形態の第１の変形例について説明する。図７は、図１に対応する図であって、第１の変形例の固体撮像装置を示す。図７の固体撮像装置は、図１の固体撮像装置に対して、光電変換領域２１の下方（裏面照射型の固体撮像装置であるから、その表側）に、光反射領域８が設けられている。これにより、凹凸パターン３によって散乱された光が光電変換領域２１から漏れ出した場合に、その光を光反射領域８によって光電変換領域２１内に再入射させることができる（光１４）。この結果、量子効率を更に向上することができる。例えば、図１の固体撮像装置では量子効率が２６％であったのに対し、光反射領域８を備えることにより量子効率３０％が実現された。 (First modification)
Next, a first modification of the embodiment will be described. FIG. 7 is a diagram corresponding to FIG. 1 and shows a solid-state image sensor of the first modification. The solid-state image sensor of FIG. 7 is provided with a light reflection region 8 below the photoelectric conversion region 21 (because it is a back-illuminated solid-state image sensor, the front side thereof) with respect to the solid-state image sensor of FIG. .. As a result, when the light scattered by the uneven pattern 3 leaks from the photoelectric conversion region 21, the light can be re-entered into the photoelectric conversion region 21 by the light reflection region 8 (light 14). As a result, the quantum efficiency can be further improved. For example, the quantum efficiency of the solid-state image sensor shown in FIG. 1 was 26%, whereas the quantum efficiency of 30% was realized by providing the light reflection region 8.
　尚、光反射領域８の材料としては、光を反射させることができる材料であれば良いが、例えばＣｕ、Ａｌ等を用いることができる。 The material of the light reflection region 8 may be any material that can reflect light, but for example, Cu, Al, or the like can be used.
　　（第２の変形例）
　次に、実施形態の第２の変形例について説明する。本変形例では、ディープトレンチ素子分離の平面図における形状について、図１の場合とは異なっている。 (Second modification)
Next, a second modification of the embodiment will be described. In this modification, the shape of the deep trench element separation in the plan view is different from that in FIG.
　具体的に、図８は、図１の示す固体撮像装置のディープトレンチ素子分離７について、シリコン基板５に対して垂直な方向から見た平面図におけるを示す。ディープトレンチ素子分離７は、画素２２毎に形成され、光電変換領域２１を取り囲んでいる。また、画素２２は正方形の領域であり、ディープトレンチ素子分離７も複数の画素２２に対して正方形の格子状に設けられている。 Specifically, FIG. 8 shows a plan view of the deep trench element separation 7 of the solid-state image sensor shown in FIG. 1 as viewed from a direction perpendicular to the silicon substrate 5. The deep trench element separation 7 is formed for each pixel 22 and surrounds the photoelectric conversion region 21. Further, the pixel 22 is a square region, and the deep trench element separation 7 is also provided in a square grid pattern with respect to the plurality of pixels 22.
　このようなディープトレンチ素子分離７の場合、画素２２の中心付近から、正方形の頂点までの距離（図８において矢印により示す対角線方向の距離）が長い。このことから、凹凸パターン３によって散乱された光がディープトレンチ素子分離７によって反射されにくい。 In the case of such a deep trench element separation 7, the distance from the vicinity of the center of the pixel 22 to the apex of the square (diagonal distance indicated by the arrow in FIG. 8) is long. For this reason, the light scattered by the uneven pattern 3 is less likely to be reflected by the deep trench element separation 7.
　これに対して、本変形例の固体撮像装置では、図９に示すように、正方形の角を落とした八角形のディープトレンチ素子分離７ａを備えている。図８のディープトレンチ素子分離７と比較すると、正方形である画素２２の４つの角付近において、それぞれ三角形の領域を除外し、八角形の光電変換領域２１を取り囲むパターンとなっている。 On the other hand, as shown in FIG. 9, the solid-state image sensor of this modified example is provided with an octagonal deep trench element separation 7a in which the corners of the square are reduced. Compared with the deep trench element separation 7 of FIG. 8, in the vicinity of the four corners of the square pixel 22, each triangular region is excluded, and the pattern surrounds the octagonal photoelectric conversion region 21.
　このようにすると、光電変換領域２１の中心から、ディープトレンチ素子分離７ａまでの距離が短くなる。従って、凹凸パターン３によって散乱された光がディープトレンチ素子分離７によって反射されやすくなり、量子効率が向上する。具体的に、図８のディープトレンチ素子分離７を備える固体撮像装置では量子効率が２６％であったのに対し、図９のディープトレンチ素子分離７ａを備える本変形例では、量子効率２８％が実現された。 In this way, the distance from the center of the photoelectric conversion region 21 to the deep trench element separation 7a becomes short. Therefore, the light scattered by the uneven pattern 3 is easily reflected by the deep trench element separation 7, and the quantum efficiency is improved. Specifically, the quantum efficiency of the solid-state image pickup apparatus provided with the deep trench element separation 7 of FIG. 8 was 26%, whereas the quantum efficiency of 28% was achieved in this modification provided with the deep trench element separation 7a of FIG. It was realized.
　尚、八角形の形状により量子効率は変動する。これについて図１０に示す。正方形である画素２２の対角方向の長さをｂとし、角落とし幅（対角線方向について、画素２２の角におけるディープトレンチ素子分離７ａに含まれない部分の寸法）をａとして、角落とし率を２ａ／ｂと定義する。図１０には、角落とし率（％）と、量子効率との関係を示す。 The quantum efficiency fluctuates depending on the octagonal shape. This is shown in FIG. Let b be the diagonal length of the square pixel 22, and let a be the angle drop width (the dimension of the portion of the corner of the pixel 22 that is not included in the deep trench element separation 7a in the diagonal direction). It is defined as 2a / b. FIG. 10 shows the relationship between the corner drop rate (%) and the quantum efficiency.
　図１０に示されるように、角落とし率が３％以上で且つ９％以下の範囲において、量子効率が高くなっている。このように、実質的な効果を得るためには角落とし率は一定の大きさを要し、例えば３％以上とするのが良い。また、光電変換領域２１の体積が減少すると量子効率が低下するので、これを避けるためには上限があり、例えば９％以下とするのが好ましい。 As shown in FIG. 10, the quantum efficiency is high in the range where the corner drop rate is 3% or more and 9% or less. As described above, in order to obtain a substantial effect, the corner drop rate needs to have a certain magnitude, and is preferably 3% or more, for example. Further, since the quantum efficiency decreases as the volume of the photoelectric conversion region 21 decreases, there is an upper limit in order to avoid this, and it is preferably 9% or less, for example.
　尚、ここではディープトレンチ素子分離７ａが八角形のパターンとなる例を示したが、これには限らない。画素２２の角の領域を曲線状に除外するパターン（正方形の角を丸めたようなパターン）等としても良い。また、画素２２についても、正方形には限定されず、長方形であったり、四角形の一部が切り欠かれた形状であったりしても良い。この場合にも、中心からの距離を小さくするように角の部分を除外するような形状にディープトレンチ素子分離を形成しても良い。 Here, an example in which the deep trench element separation 7a has an octagonal pattern is shown, but the present invention is not limited to this. A pattern that excludes the corner region of the pixel 22 in a curved shape (a pattern in which the corners of a square are rounded) may be used. Further, the pixel 22 is not limited to a square, and may be a rectangle or a shape in which a part of the quadrangle is cut out. Also in this case, the deep trench element separation may be formed in a shape excluding the corner portion so as to reduce the distance from the center.
　また、第１及び第２の変形例、図２及び図６に示す平面図における凹凸パターン３の形状等は、様々に組み合わせて用いることができる。更に、以上では裏面照射型の固体撮像装置であるものとして説明したが、このことは必須ではない。表面照射型の場合にも、凹凸パターン３を形成することにより量子効率を向上させる効果は実現する。 Further, the first and second modified examples, the shape of the uneven pattern 3 in the plan view shown in FIGS. 2 and 6, and the like can be used in various combinations. Further, although the description has been made above as a back-illuminated solid-state image sensor, this is not essential. Even in the case of the surface irradiation type, the effect of improving the quantum efficiency is realized by forming the uneven pattern 3.
　本開示の技術は、特に近赤外領域において量子効率を向上し良好な画像を得られるので、固体撮像装置として有用である。 The technique of the present disclosure is useful as a solid-state image sensor because it improves quantum efficiency and obtains a good image, especially in the near infrared region.
　１　　　オンチップレンズ
　２　　　平坦化膜
　３　　　凹凸パターン
　４　　　絶縁膜
　５　　　シリコン基板
　６　　　メタルシールド
　７　　　ディープトレンチ素子分離
　７ａ　　ディープトレンチ素子分離
　８　　　光反射領域
１１　　　光
１２　　　光
１３　　　光
１４　　　光
２１　　　光電変換領域
２２　　　画素 1 On-chip lens 2 Flattening film 3 Concavo-convex pattern 4 Insulation film 5 Silicon substrate 6 Metal shield 7 Deep trench element separation 7a Deep trench element separation 8 Light reflection area 11 Light 12 Light 13 Light 14 Light 21 Photoelectric conversion area 22 pixels

Claims (9)

  1. 　光電変換領域を有する画素が複数配列されたシリコン基板と、
    　前記シリコン基板上に形成された絶縁膜と、
    　前記絶縁膜上に形成された凹凸パターンと、
    　前記凹凸パターン上を覆う平坦化膜とを備え、
    　前記凹凸パターンは、前記平坦化膜よりも屈折率の高い高屈折率材料からなることを特徴とする固体撮像装置。     A silicon substrate in which a plurality of pixels having a photoelectric conversion region are arranged, and
    The insulating film formed on the silicon substrate and
    The uneven pattern formed on the insulating film and
    A flattening film that covers the uneven pattern is provided.
    A solid-state image sensor, wherein the uneven pattern is made of a high-refractive index material having a higher refractive index than the flattening film.
  2. 　請求項１において、
    　前記高屈折率材料の屈折率は、１．９以上であることを特徴とする固体撮像装置。     In claim 1,
    A solid-state image sensor characterized in that the refractive index of the high-refractive index material is 1.9 or more.
  3. 　請求項１又は２において、
    　前記高屈折率材料は、ポリシリコンであることを特徴とする固体撮像装置。     In claim 1 or 2,
    A solid-state image sensor, wherein the high-refractive index material is polysilicon.
  4. 　請求項１～３のいずれか１つにおいて、
    　前記凹凸パターンの凸部の厚さは、２５０ｎｍ以上で且つ４００ｎｍ以下であることを特徴とする固体撮像装置。     In any one of claims 1 to 3,
    A solid-state image sensor, characterized in that the thickness of the convex portion of the uneven pattern is 250 nm or more and 400 nm or less.
  5. 　請求項１～４のいずれか１つにおいて、
    　前記凹凸パターンの凸部は、前記画素内において、７．５％以上で且つ３０％以下の面積占有率を有することを特徴とする固体撮像装置。     In any one of claims 1 to 4,
    A solid-state image sensor, wherein the convex portion of the uneven pattern has an area occupancy of 7.5% or more and 30% or less in the pixel.
  6. 　請求項１～５のいずれか１つにおいて、
    　前記光電変換領域に対して前記凹凸パターンとは反対側に、光反射領域が設けられていることを特徴とする固体撮像装置。     In any one of claims 1 to 5,
    A solid-state image sensor, characterized in that a light reflection region is provided on the side opposite to the uneven pattern with respect to the photoelectric conversion region.
  7. 　請求項１～６のいずれか１つにおいて、
    　前記画素において、前記光電変換領域の少なくとも一部を取り囲むようにディープトレンチ素子分離が形成されていることを特徴とする固体撮像装置。     In any one of claims 1 to 6,
    A solid-state image pickup device in which a deep trench element separation is formed so as to surround at least a part of the photoelectric conversion region in the pixel.
  8. 　請求項７において、
    　前記シリコン基板に垂直な方向に見たとき、前記ディープトレンチ素子分離は、正方形の領域から４つの角付近の領域をそれぞれ部分的に除外した八角形のパターンを有することを特徴とする固体撮像装置。     In claim 7,
    When viewed in a direction perpendicular to the silicon substrate, the deep trench element separation is characterized by having an octagonal pattern in which regions near four corners are partially excluded from a square region. ..
  9. 　請求項８において、
    　前記シリコン基板に垂直な方向に見たとき、
    　前記正方形の領域の対角線の長さをｂ、
    　前記対角線のうち、前記正方形の領域の１つの角において前記八角形のパターンに含まれない部分の長さをａとして、
    　ｂに対して２ａが３％以上で且つ９％以下となることを特徴とする固体撮像装置。     In claim 8.
    When viewed in the direction perpendicular to the silicon substrate
    The diagonal length of the square area is b,
    Let a be the length of a portion of the diagonal line that is not included in the octagonal pattern at one corner of the square area.
    A solid-state image sensor, characterized in that 2a is 3% or more and 9% or less with respect to b.
PCT/JP2019/027807 2019-07-12 2019-07-12 Solid-state imaging device WO2021009816A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2019/027807 WO2021009816A1 (en) 2019-07-12 2019-07-12 Solid-state imaging device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2019/027807 WO2021009816A1 (en) 2019-07-12 2019-07-12 Solid-state imaging device

Publications (1)

Publication Number Publication Date
WO2021009816A1 true WO2021009816A1 (en) 2021-01-21

Family

ID=74209762

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2019/027807 WO2021009816A1 (en) 2019-07-12 2019-07-12 Solid-state imaging device

Country Status (1)

Country Link
WO (1) WO2021009816A1 (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009004533A (en) * 2007-06-21 2009-01-08 Toshiba Corp Array photodetector and light converging method
JP2011210981A (en) * 2010-03-30 2011-10-20 Sony Corp Solid-state image pickup device, method for manufacturing the same, and electronic apparatus
JP2015032636A (en) * 2013-07-31 2015-02-16 株式会社東芝 Manufacturing method of solid-state imaging apparatus, and solid-state imaging apparatus
JP2019114642A (en) * 2017-12-22 2019-07-11 キヤノン株式会社 Solid state image sensor, electronic equipment and transport equipment

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009004533A (en) * 2007-06-21 2009-01-08 Toshiba Corp Array photodetector and light converging method
JP2011210981A (en) * 2010-03-30 2011-10-20 Sony Corp Solid-state image pickup device, method for manufacturing the same, and electronic apparatus
JP2015032636A (en) * 2013-07-31 2015-02-16 株式会社東芝 Manufacturing method of solid-state imaging apparatus, and solid-state imaging apparatus
JP2019114642A (en) * 2017-12-22 2019-07-11 キヤノン株式会社 Solid state image sensor, electronic equipment and transport equipment

Similar Documents

Publication Publication Date Title
TWI686939B (en) Backside-illuminated color image sensors with crosstalk-suppressing color filter array and method for manufacturing the same
TWI633675B (en) Image sensor device and method of fabricating the same
JP6168331B2 (en) Imaging device and imaging apparatus
US8711258B2 (en) Solid-state imaging device and method for manufacturing the same
JP4702384B2 (en) Solid-state image sensor
JP6813319B2 (en) An image sensor including a color separation element, and an imaging device including an image sensor.
US9853073B2 (en) Image sensor for producing vivid colors and method of manufacturing the same
US20140339615A1 (en) Bsi cmos image sensor
KR101232282B1 (en) Image sensor and method for fabricating the same
JP5269409B2 (en) Image sensor element and method of manufacturing the same
TW202139450A (en) Method and structure to improve image sensor crosstalk
JP2011258728A (en) Solid state image sensor and electronic information apparatus
US10910425B2 (en) Solid-state image sensor
JP2002064193A (en) Solid-state imaging device and manufacturing method thereof
US9060144B2 (en) Image sensor and manufacturing method thereof
US9685472B2 (en) Image sensor with reduced spectral and optical crosstalk and method for making the image sensor
WO2013021541A1 (en) Solid-state image pickup device
KR100900682B1 (en) Image Sensor and Method for Manufacturing thereof
CN114556573A (en) Image sensor and imaging apparatus
JP5409087B2 (en) Solid-state image sensor
JP7270679B2 (en) image sensor
JP2009146957A (en) Solid-state imaging apparatus, and manufacturing method of solid-state imaging apparatus
WO2021009816A1 (en) Solid-state imaging device
US20210408099A1 (en) Imaging device
JP4702484B2 (en) Solid-state image sensor

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19937399

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19937399

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP