WO2021015070A1 - 画素、固体撮像装置及び画素の製造方法 - Google Patents

画素、固体撮像装置及び画素の製造方法 Download PDF

Info

Publication number
WO2021015070A1
WO2021015070A1 PCT/JP2020/027504 JP2020027504W WO2021015070A1 WO 2021015070 A1 WO2021015070 A1 WO 2021015070A1 JP 2020027504 W JP2020027504 W JP 2020027504W WO 2021015070 A1 WO2021015070 A1 WO 2021015070A1
Authority
WO
WIPO (PCT)
Prior art keywords
conversion unit
light
region
pixel
photoelectric conversion
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/JP2020/027504
Other languages
English (en)
French (fr)
Japanese (ja)
Inventor
信一 寺西
篤史 小野
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shizuoka University NUC
Original Assignee
Shizuoka University NUC
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 Shizuoka University NUC filed Critical Shizuoka University NUC
Priority to JP2021533975A priority Critical patent/JP7493250B2/ja
Publication of WO2021015070A1 publication Critical patent/WO2021015070A1/ja
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/20—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from infrared radiation only
    • H04N23/21—Cameras or camera modules comprising electronic image sensors; Control thereof for generating image signals from infrared radiation only from near infrared [NIR] radiation
    • 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
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
    • H10F30/00—Individual radiation-sensitive semiconductor devices in which radiation controls the flow of current through the devices, e.g. photodetectors
    • H10F30/20—Individual radiation-sensitive semiconductor devices in which radiation controls the flow of current through the devices, e.g. photodetectors the devices having potential barriers, e.g. phototransistors
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
    • H10F39/00—Integrated devices, or assemblies of multiple devices, comprising at least one element covered by group H10F30/00, e.g. radiation detectors comprising photodiode arrays
    • H10F39/10—Integrated devices
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
    • H10F39/00—Integrated devices, or assemblies of multiple devices, comprising at least one element covered by group H10F30/00, e.g. radiation detectors comprising photodiode arrays
    • H10F39/10—Integrated devices
    • H10F39/12—Image sensors
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
    • H10F77/00—Constructional details of devices covered by this subclass
    • H10F77/40—Optical elements or arrangements

Definitions

  • the present invention relates to a pixel, a solid-state image sensor, and a method for manufacturing a pixel.
  • the pixel and solid-state image sensor are formed using a substrate made of silicon. Silicon absorbs light and generates carriers corresponding to the absorbed light. As shown in FIG. 12, the property of silicon to absorb light corresponds to the wavelength of light. For example, silicon satisfactorily absorbs visible light having a wavelength of 360 nm or more and 830 nm or less. On the other hand, silicon is difficult to absorb near-infrared light having a wavelength of 830 nm or more. Therefore, in order to obtain a good image using a pixel using silicon and a solid-state image sensor, it is necessary to improve the light receiving sensitivity.
  • the imaging device disclosed in Non-Patent Document 1 has an uneven structure provided on an incident surface of light. According to this uneven structure, the substantial light transmission distance inside the silicon is extended. As the transmission distance increases, the chances of light being absorbed by silicon increase. As a result, the light receiving sensitivity is improved.
  • the image pickup apparatus disclosed in Patent Document 1 has a configuration for converting near-infrared light into evanescent light. Evanescent light is more easily absorbed by silicon than near-infrared light. Therefore, the image pickup apparatus disclosed in Patent Document 1 has improved light receiving sensitivity.
  • Crosstalk and dark current can be mentioned as factors of noise.
  • Crosstalk is a phenomenon in which the first pixel affects a second pixel adjacent to the first pixel. For example, when the thickness of silicon is increased in order to increase the light receiving sensitivity, crosstalk is likely to occur. Further, the structure that produces evanescent light is formed of metal. The imaging device then has an interface where the semiconductor comes into contact with the metal. Such interfaces form Schottky barrier or ohmic contacts. These bonding forms are factors that generate dark current.
  • the present invention provides a pixel, a solid-state image sensor, and a method for manufacturing a pixel for near-infrared light capable of obtaining a good image.
  • a pixel according to an embodiment of the present invention includes a photoelectric conversion unit that generates a signal voltage according to the absorbed light received from the light input surface, and an optical conversion unit that receives the incident light to generate the absorbed light. ..
  • the photoelectric conversion unit has a first region having a first impurity concentration and a second region having a second impurity concentration higher than the first impurity concentration.
  • the second region includes an optical input surface and is in direct contact with the optical conversion unit.
  • the light conversion unit receives the incident light and generates absorbed light including evanescent light.
  • the light conversion unit of this pixel receives the incident light and generates the absorbed light that is output to the photoelectric conversion unit.
  • the optical input surface of the photoelectric conversion unit is in direct contact with the optical conversion unit. Therefore, the absorbed light is input to the photoelectric conversion unit without being attenuated. Then, this absorbed light is more easily absorbed by the photoelectric conversion unit than the incident light. Therefore, the light receiving sensitivity can be increased.
  • the photoelectric conversion unit has a second region including an optical input surface. The second region has a higher impurity concentration than the first region. According to this configuration, the dark current that may occur at the interface can be suppressed. That is, the pixel can increase the light receiving sensitivity and suppress the dark current. As a result, according to the pixels, a good image can be obtained.
  • the concentration of the second impurity in the second region may decrease as it approaches the first region. According to this configuration, the minority carriers generated by the absorption of the absorbed light in the second region can be quickly moved to the first region. As a result, the signal voltage due to the minority carriers is well captured. That is, the proportion of minority carriers from the absorbed light contributing to the signal voltage without being recombined increases. As a result, the light receiving sensitivity can be further increased.
  • the second region may be formed by ion implantation with a dose of 1 ⁇ 10 14 cm- 2 or more and 3 ⁇ 10 15 cm- 2 or less. Further, the second region may be formed by ion implantation in which the implantation energy is 0.2 keV or more and 1 keV or less. By lowering the energy to 1 keV or less, the density profile after annealing in the second region decreases monotonically from the light incident surface toward the inside of the photoelectric conversion unit. As will be described later, the second region may have a dose amount of 1 ⁇ 10 14 cm- 2 or more in order to prevent a dark current generated from the optical input surface.
  • the second region may have a dose amount of 3 ⁇ 10 15 cm- 2 or less in order to prevent the occurrence of crystal defects. According to this configuration, it is possible to suppress the occurrence of crystal defects in the second region. Therefore, the life of the minority carriers generated according to the absorbed light can be extended. As a result, the signal voltage caused by the minority carrier can be captured well. That is, the conversion efficiency from the absorbed light to the signal voltage due to the minority carriers is increased. As a result, the light receiving sensitivity can be further increased.
  • the solid-state image sensor which is another embodiment of the present invention, has a plurality of pixels arranged in a two-dimensional shape, a pixel portion having an isolation wall provided between pixels adjacent to each other, and an operation of the pixel portion. It includes a pixel control unit that generates a control signal to be controlled, and a signal processing unit that receives a signal voltage generated by the pixel unit.
  • the pixels are provided on a photoelectric conversion unit that generates a signal voltage according to the absorbed light received from the optical input surface and on the optical input surface of the photoelectric conversion unit so as to receive the incident light and generate the absorbed light. It also has an optical conversion unit.
  • the photoelectric conversion unit has a first region having a first impurity concentration and a second region having a second impurity concentration higher than the first impurity concentration.
  • the second region includes an optical input surface and is in direct contact with the optical conversion unit.
  • the light conversion unit receives the incident light and generates absorbed light including evanescent light.
  • This solid-state image sensor has pixels having the same configuration as the above pixels. Therefore, a good image can be obtained.
  • one optical conversion unit adjacent to each other may be separated from the other optical conversion unit.
  • the side portion of the optical conversion unit may be formed on the optical input surface of the photoelectric conversion unit.
  • the pixel portion may be provided on the sensor substrate.
  • the sensor substrate may have a voltage application unit that applies a predetermined potential to the second region.
  • the voltage applying portion may include a dicing surface of the sensor substrate formed by dicing and a conductive portion that applies a voltage to the dicing surface.
  • the dicing surface may include at least the end surface of the second region and the end surface of the conductive portion.
  • the photoelectric conversion unit may generate electrons as signal carriers by absorbing the light to be absorbed.
  • the second region may receive a negative potential.
  • the photoelectric conversion unit may generate holes as signal carriers by absorbing the light to be absorbed.
  • the second region may receive a positive potential. According to these configurations, the electric field of the photoelectric conversion unit becomes strong. In addition, the collection of minority carriers will be facilitated. As a result, good light receiving sensitivity can be obtained.
  • the second region may be electrically connected to the optical converter.
  • the first optical conversion unit included in the first pixel may be electrically connected to the second optical conversion unit included in the second pixel adjacent to the first pixel.
  • the optical conversion unit may include a plurality of convex portions protruding from the main surface of the second region. The plurality of protrusions may be electrically connected to each other. According to this configuration, a configuration capable of suppressing dark current can be obtained by a simple step.
  • the solid-state image sensor which is still another embodiment of the present invention, may include a plurality of pixels arranged in a two-dimensional manner in the light receiving region.
  • the pixels are provided on the optical input surface of the photoelectric conversion unit so as to generate a signal voltage corresponding to the absorbed light received from the optical input surface and to generate the absorbed light by receiving the incident light. It may also have an optical conversion unit.
  • the photoelectric conversion unit may come into direct contact with the light conversion unit.
  • the light conversion unit may include a plurality of convex portions arranged according to a predetermined period, and may receive incident light to generate absorbed light including evanescent light. The predetermined period may be set based on the generation condition of the evanescent light and the direction of the incident light.
  • this solid-state image sensor even if the direction of the incident light is different for each pixel, it is possible to generate the evanescent light according to the direction of the incident light. Therefore, the decrease of evanescent light due to the difference in the angle of the incident light is suppressed. As a result, the absorbed light can be suitably generated from the incident light.
  • the period when the direction of the incident light is the normal direction of the main surface of the light conversion unit, the period may be a predetermined period (L).
  • the predetermined period of the pixels arranged at the side portion of the light receiving region may be larger than the predetermined period of the pixels arranged at the central portion of the light receiving region. Also with this configuration, the absorbed light can be suitably generated from the incident light.
  • the optical conversion unit of the pixel which is one form, may be provided on the optical input surface of the photoelectric conversion unit. According to this configuration, a pixel including an optical conversion unit and a photoelectric conversion unit can be easily manufactured.
  • the optical conversion unit of the pixel may include a light receiving surface that receives incident light and may be embedded in the photoelectric conversion unit so that the light receiving surface is exposed from the photoelectric conversion unit.
  • the bottom surface of the optical conversion unit comes into contact with the photoelectric conversion unit.
  • the side surface of the optical conversion unit also comes into contact with the photoelectric conversion unit. Therefore, it becomes possible to more efficiently absorb the evanescent light generated in the light conversion unit. As a result, the photoelectric conversion efficiency is further increased. Therefore, a better image can be obtained.
  • the pixel in one form may further include a covering portion that covers the photoelectric conversion portion and the optical conversion portion.
  • the permittivity of the coating may be higher than the permittivity of air.
  • the configuration of the light conversion unit that generates the absorbed light including the evanescent light is determined by the wavelength of the incident light. According to this covering portion, it is possible to give a degree of freedom to the configuration of the light conversion portion that is uniquely determined by the wavelength of the incident light.
  • the optical conversion unit of the pixel which is one form, may include a plurality of convex portions formed apart from each other.
  • the apparent period of the plurality of convex portions in the optical conversion portion covered with the covering portion may be determined by the dielectric constant of the covering portion and the actual period of the plurality of convex portions.
  • the apparent period may satisfy the resonance condition of generating the absorbed light including the evanescent light by the incident light. According to this configuration, it is possible to make the period of the plurality of convex portions constituting the light conversion unit shorter than the period of the light conversion unit determined by the wavelength of the incident light. As a result, the resonance condition can be satisfied. In addition, more light to be absorbed can be generated. Therefore, since the photoelectric conversion efficiency is further increased, a better image can be obtained.
  • the dielectric constant of the covering portion in the pixel which is one form, may be equal to the dielectric constant of the second region of the photoelectric conversion portion. According to this configuration, the difference in permittivity around the optical conversion unit becomes small. As a result, the absorbed light including the evanescent light can be generated more efficiently in the light conversion unit.
  • the coating portion of the pixel which is one form, may be formed of any one of aluminum oxide, hafnium oxide, zirconium oxide and tantalum oxide. According to these materials, the above-mentioned covering portion can be formed satisfactorily.
  • the pixel in one form may further include a microlens arranged on the covering portion. According to this configuration, the incident light can be efficiently collected in the light conversion unit. As a result, the absorbed light including the evanescent light can be generated more efficiently in the light conversion unit.
  • Yet another embodiment of the present invention includes a photoelectric conversion unit that generates a signal voltage according to the absorbed light received from the optical input surface, and an optical conversion unit that receives the incident light to generate the absorbed light.
  • the photoelectric conversion unit is a method for manufacturing a pixel having a first region having a first impurity concentration and a second region having a second impurity concentration higher than the first impurity concentration.
  • the pixel manufacturing method includes a first step of providing a plurality of recesses for embedding the optical conversion unit in the photoelectric conversion unit, and a second step of performing a process for forming at least a second region with respect to the wall surface of the recess. It has a third step of providing a plurality of metal portions constituting the optical conversion portion in the recess.
  • an optical conversion unit embedded in the photoelectric conversion unit is formed so as to be exposed from the photoelectric conversion unit. Then, the bottom surface of the photoelectric conversion unit comes into contact with the photoelectric conversion unit. Further, the side surface of the photoelectric conversion unit also includes a metal portion that comes into contact with the photoelectric conversion unit. Therefore, it is possible to manufacture a pixel capable of more efficiently absorbing the evanescent light generated in the light conversion unit.
  • a fourth step of providing a covering portion having a dielectric constant higher than the dielectric constant of air so as to cover the photoelectric conversion portion and the optical conversion portion is further added. You may have. According to this step, it is possible to efficiently collect the incident light in the light conversion unit, and it is possible to manufacture a pixel capable of more efficiently generating absorbed light including evanescent light in the light conversion unit. ..
  • the optical conversion unit of the pixel which is one form, may include a first metal layer, a dielectric layer, and a second metal layer laminated along the direction of the incident light.
  • the second metal layer may be provided on the light input surface. According to this configuration, the absorption loss of incident light in the light conversion unit can be reduced. Therefore, the light absorption efficiency in the photoelectric conversion unit can be further increased.
  • the optical conversion unit of the pixel which is one form, includes a light receiving surface that receives incident light, and may be embedded in the photoelectric conversion unit so that the light receiving surface is exposed from the photoelectric conversion unit.
  • the light conversion unit may include a first metal layer, a dielectric layer, and a second metal layer laminated along the direction of the incident light.
  • the first metal layer may form a light receiving surface. This configuration also reduces the absorption loss of incident light in the light conversion unit. Therefore, the light absorption efficiency in the photoelectric conversion unit can be further increased.
  • a good image can be obtained.
  • FIG. 1 is a diagram schematically showing a configuration of a solid-state image sensor according to an embodiment.
  • FIG. 2 is a diagram showing an electrical configuration of pixels.
  • FIG. 3 is an example of a control signal provided to the pixel.
  • FIG. 4 is a diagram showing a pixel structure.
  • FIG. 5 is an enlarged view showing the vicinity (S1) of the optical input surface which is a part of FIG.
  • FIG. 6 is an enlarged view of the pixel and the isolation wall (S2).
  • FIG. 7 is a diagram showing a configuration of an optical conversion unit.
  • FIG. 8 is a diagram for explaining the solid-state image sensor of the first modification.
  • FIG. 9 is an enlarged view showing a part of the solid-state image sensor of the first modification.
  • FIG. 8 is a diagram for explaining the solid-state image sensor of the first modification.
  • FIG. 10 is a diagram for explaining the solid-state image sensor of the second modification.
  • FIG. 11 is a diagram for explaining the solid-state image sensor of the second modification.
  • FIG. 12 is a graph showing the light absorption rate of silicon.
  • FIG. 13 is a graph showing the attenuation of light intensity in silicon.
  • FIG. 14 is an example of the distribution of impurity concentration in the pinning region.
  • FIG. 15 is a diagram showing a pixel structure of a reference example.
  • FIG. 16 is a diagram showing a pixel structure of another reference example.
  • FIG. 17 is an enlarged view showing a main part of the solid-state image sensor of the first embodiment.
  • FIG. 18 is an enlarged view showing a main part of the solid-state image sensor of the second embodiment.
  • FIG. 19 is a diagram showing a modified example of the solid-state image sensor of the third embodiment.
  • 20 (a) and 20 (b) are diagrams showing the main steps of the method for manufacturing the solid-state image sensor of the second embodiment, following FIG. 19.
  • FIG. 21 is a diagram showing a modified example of the solid-state image sensor of the third embodiment.
  • 22 (a) and 22 (b) are enlarged views showing a main part of a solid-state image sensor having a high-dielectric film.
  • 23 (a) and 23 (b) are views showing a modified example of a solid-state image sensor having a high-dielectric film.
  • FIG. 24 is a diagram showing a modified example of the solid-state image sensor of the third embodiment.
  • FIG. 25 is an enlarged view showing a main part of the solid-state image sensor of the third modification.
  • FIG. 26 is an enlarged view showing a main part of the solid-state image sensor of the modified example 4.
  • the following description relates to increasing the sensitivity of a photodiode or a CMOS image sensor in which each pixel is equipped with a photodiode.
  • the following description relates to increasing the sensitivity in the near infrared light region.
  • the solid-state image sensor 1 of the present embodiment is a so-called back-illuminated image sensor.
  • the solid-state image sensor 1 may receive light from the opposite side.
  • the solid-state image sensor 1 may receive light from the support substrate 11 side, which will be described later.
  • the solid-state image sensor 1 includes a pixel unit 2, a pixel control unit 3, a signal processing unit 4, a horizontal control line group 6 composed of a plurality of horizontal control lines, and a vertical signal line 7.
  • the pixel unit 2, the pixel control unit 3, the signal processing unit 4, and the like are provided on the sensor substrate 12, which will be described later.
  • the sensor board 12 is attached to the support board 11.
  • a pixel portion 2 is provided in a portion of the sensor substrate 12 corresponding to the light receiving region S.
  • the pixel unit 2 has a plurality of pixels 8 arranged in a two-dimensional manner.
  • the pixel 8 outputs the signal voltage corresponding to the incident light to the vertical signal line 7.
  • the pixel control unit 3 is connected to each pixel 8 via a horizontal control line group 6.
  • the pixel control unit 3 outputs a control signal ⁇ for controlling the operation of the pixel 8.
  • the signal processing unit 4 is connected to each pixel 8 via a vertical signal line 7.
  • the signal processing unit 4 receives the signal voltage ⁇ output by the pixel 8.
  • the signal processing unit 4 generates an image signal from the signal voltage ⁇ output from the pixel 8.
  • FIG. 2 shows the electrical configuration of the pixel 8.
  • Pixel 8 is a so-called 4-transistor type CMOS image sensor.
  • the pixel 8 of this embodiment is an N-channel type. That is, the signal of the pixel 8 is carried by the signal electron.
  • the pixel 8 may be a P-channel type.
  • the pixel 8 has a photodiode PD, a floating diffusion layer FD, a transfer gate TG, a reset transistor RG, a source follower transistor SF, and a selection transistor SEL.
  • the photodiode PD is a PN junction type.
  • the photodiode PD generates signal electrons as carriers.
  • the photodiode PD accumulates the generated signal electrons.
  • the transfer gate TG, the reset transistor RG, the source follower transistor SF, and the selection transistor SEL are field effect transistors, respectively.
  • the source of the transfer gate TG is connected to the photodiode PD.
  • the drain of the transfer gate TG is connected to the floating diffusion layer FD.
  • the gate of the transfer gate TG is connected to the horizontal control line group 6.
  • the transfer gate TG receives the control signal ⁇ 1 from the horizontal control line group 6.
  • the transfer gate TG controls the transfer of signal electrons from the photodiode PD to the floating diffusion layer FD based on the control signal ⁇ 1.
  • the floating diffusion layer FD is connected to the drain of the transfer gate TG.
  • the floating diffusion layer FD is connected to the photodiode PD via the transfer gate TG.
  • the floating diffusion layer FD converts signal electrons into a signal voltage.
  • the floating diffusion layer FD is connected to the source of the reset transistor RG.
  • the floating diffusion layer FD is also connected to the gate of the source follower transistor SF.
  • the source of the reset transistor RG is connected to the floating diffusion layer FD.
  • the drain of the reset transistor RG is connected to the reset drain.
  • the gate of the reset transistor RG is connected to the horizontal control line group 6.
  • the reset transistor RG receives the control signal ⁇ 2 from the horizontal control line group 6.
  • the reset transistor RG resets the potential of the floating diffusion layer FD based on the control signal ⁇ 2.
  • the source of the source follower transistor SF is connected to the selection transistor SEL.
  • the drain of the source follower transistor SF is connected to an analog power supply.
  • the gate of the source follower transistor SF is connected to the floating diffusion layer FD.
  • the source follower transistor SF outputs a signal voltage corresponding to the voltage input to the gate via the selection transistor SEL.
  • the source of the selection transistor SEL is connected to the vertical signal line 7.
  • the drain of the selection transistor SEL is connected to the source of the source follower transistor SF.
  • the gate of the selection transistor SEL is connected to the horizontal control line group 6.
  • the selection transistor SEL outputs a signal voltage ⁇ to the vertical signal line 7 based on the control signal ⁇ 3.
  • FIG. 3 shows the control signals ⁇ 1, ⁇ 2, ⁇ 3, RS, SS output by the pixel control unit 3.
  • FIG. 3 schematically shows the timings of the control signals ⁇ 1, ⁇ 2, ⁇ 3, RS, and SS during the read period of the nth row.
  • “(H)” indicates that the signal is high, that is, on.
  • “(L)” indicates that the signal is low, that is, off.
  • the pixel control unit 3 outputs a control signal ⁇ 3 (H). As a result, the line that received the control signal ⁇ 3 is selected. Subsequently, the pixel control unit 3 outputs the control signal ⁇ 2 (H) for a predetermined period. As a result, the reset drain voltage is output to the floating diffusion layer FD. Subsequently, the pixel control unit 3 outputs the control signal RS (H) for a predetermined period. As a result, the reset level is sampled. Next, the pixel control unit 3 outputs the control signal ⁇ 1 (H) for a predetermined period. As a result, signal electrons are transferred from the photodiode PD to the floating diffusion layer FD.
  • the pixel control unit 3 outputs the control signal SS (H) for a predetermined period.
  • the signal level caused by the signal electrons stored in the floating diffusion layer FD is sampled at the timing of the control signal SS (H).
  • the pixel control unit 3 outputs the control signal ⁇ 2 (H) again for a predetermined period.
  • the reset drain voltage is output to the floating diffusion layer FD. That is, the potential of the floating diffusion layer FD is reset.
  • the pixel control unit 3 outputs the control signal ⁇ 3 (L).
  • the reading of the nth row is completed.
  • the pixel control unit 3 outputs the control signal ⁇ 3 (H) to the next n + 1 line.
  • the difference between the sampled signal level and the reset level is created. This difference is correlated double sampling. Correlated double sampling plays a role in noise reduction and offset elimination.
  • the level difference component is treated as a signal. Next, the level difference component is converted from an analog value to a digital value. Then, the component of the level difference converted into a digital value is output to the outside of the solid-state image sensor 1.
  • FIG. 4 schematically shows a cross section of the pixel portion 2.
  • the pixel portion 2 is formed on the sensor substrate 12.
  • the sensor board 12 has a wiring region 13 and an element region 14.
  • the surface of the sensor board 12 on the wiring region 13 side is joined to the support board 11.
  • the pixel unit 2 has a color filter 16 and a microlens 17 provided on the sensor substrate 12.
  • the color filter 16 plays a role of colorization.
  • the microlens 17 plays a role of condensing light.
  • the wiring area 13 has a plurality of wirings 18, a plurality of vias 19, and gates 21, 21A.
  • the wiring 18 is made of copper or aluminum.
  • the via 19 electrically connects the wirings 18 to each other. Further, the via 19 electrically connects the wiring 18 to the gates 21 and 21A formed of polysilicon.
  • the wiring region 13 also has an insulating film (not shown) that covers the wiring 18 and the via 19.
  • the element region 14 has an isolation wall 22 and pixels 8.
  • the isolation wall 22 is provided between the pixels 8 adjacent to each other.
  • the isolation wall 22 has a deep trench 23 and a p + -shaped channel stop region 24.
  • the deep trench 23 is provided on the incident side.
  • the channel stop region 24 is provided on the wiring region 13 side.
  • a charge storage portion 33a and a floating diffusion layer FD which will be described later, are provided.
  • the deep trench 23 suppresses optical crosstalk between pixels 8. Further, the deep trench 23 suppresses crosstalk due to the diffusion of signal electrons.
  • the channel stop region 24 electrically separates the pixels 8. More specifically, the channel stop region 24 electrically separates the charge storage portion 33a and the N + type region 34d.
  • Pixel 8 has a photoelectric conversion unit 26 and an optical conversion unit 27.
  • the photoelectric conversion unit 26 generates a signal voltage corresponding to the absorbed light L2 received from the optical input surface 26a.
  • the optical conversion unit 27 is provided on the optical input surface 26a of the photoelectric conversion unit 26.
  • the light conversion unit 27 receives the incident light L1 (see FIG. 5) and generates the absorbed light L2 (see FIG. 5) that is output to the photoelectric conversion unit 26.
  • the optical input surface 26a is also the main surface of the pinning region 36 described later.
  • the photoelectric conversion unit 26 has a first region 29 and a second region 31.
  • the first region 29 has a base unit 32, a charge storage unit 33a, a photodiode pinning layer 33b, and a readout unit 34.
  • the base portion 32 is P-shaped and constitutes an optical input surface 26a.
  • the substrate portion 32 absorbs the light L2 to be absorbed. As a result, signal electrons are generated.
  • the charge storage portion 33a and the photodiode pinning layer 33b are provided between the base portion 32 and the wiring region 13. In other words, the charge storage unit 33a is provided near the surface on the wiring region 13 side.
  • the charge storage unit 33a cooperates with the P-type substrate unit 32 and the photodiode pinning layer 33b to form a PN junction diode.
  • the P + type photodiode pinning layer 33b is provided between the charge storage portion 33a and the wiring region 13.
  • the photodiode pinning layer 33b prevents the generation of dark current from the interface state of the silicon surface.
  • the photodiode pinning layer 33b is in contact with the channel stop region 24 and has the same potential.
  • the reading unit 34 has a threshold value adjusting region 34a, N + type regions 34b, 34c, 34d, and a P-type well 34e.
  • the threshold adjustment region 34a is in contact with the charge storage unit 33a and the photodiode pinning layer 33b.
  • the threshold adjustment region 34a cooperates with the gate 21A to form the transfer gate TG.
  • the N + type region 34b is in contact with the threshold adjustment region 34a.
  • the N + type region 34b constitutes the floating diffusion layer FD.
  • the N + type region 34b is provided on the opposite side of the threshold adjustment region 34a from the photodiode PD. That is, the threshold adjustment region 34a is provided between the charge storage unit 33a and the floating diffusion layer FD.
  • the P-type well 34e is adjacent to the charge storage unit 33a. In other words, the P-type well 34e is adjacent to the charge storage unit 33a via the threshold adjustment region 34a.
  • the N + type region 34c is adjacent to the N + type region 34b so as to sandwich the channel region which is the P type well 34e.
  • the N + type region 34c constitutes a transfer gate TG from the charge storage unit 33a.
  • the signal electrons are read out into the N + type region 34b, which is the floating diffusion layer FD, via the threshold adjustment region 34a.
  • the signal electrons are converted into a voltage signal in this floating diffusion layer FD.
  • the N + type region 34d is adjacent to the N + type region 34c so as to sandwich the channel region which is the P type well 34e.
  • the N + type region 34d, the N + type region 34c, and the gate 21 form a source follower transistor SF.
  • the channel stop region 24 is in contact with the N + type region 34d.
  • a reset transistor RG and a selection transistor SEL are provided.
  • the P-type well 34e includes a floating diffusion layer FD, a reset transistor RG, a source follower transistor SF, and a selection transistor SEL.
  • the P-type well 34e prevents the inflow of signal electrons from the P-type substrate portion 32. Further, the P-type well 34e controls the threshold values of the reset transistor RG, the source follower transistor SF, and the selection transistor SEL.
  • FIG. 5 is an enlarged view of the area S1 of FIG. As shown in FIG. 5, the photoelectric conversion unit 26 further has a pinning region 36 formed in the second region 31.
  • the pinning region 36 is provided on the incident side of the substrate portion 32. In other words, the pinning region 36 is provided on the surface of the base portion 32 opposite to the wiring region 13 side.
  • the thickness of the pinning region 36 is 3 nm or more and 100 nm or less.
  • the pinning region 36 has a high acceptor concentration.
  • the pinning region 36 is neutralized.
  • the pinning region 36 holds holes.
  • the pinning region 36 suppresses the generation of dark current.
  • the pinning region 36 will be described in more detail.
  • the region (photoelectric conversion unit) that generates signal electrons due to light absorption functions to suppress the generation of dark current (second region 31, pinning region). 36) is included. Therefore, the layer and the film that hinder the absorbed light L2 output from the light conversion unit 27 to the photoelectric conversion unit 26 are substantially not present between the light conversion unit 27 and the photoelectric conversion unit 26.
  • the sensor will be described as an N-channel type. That is, the signal caused by the incident light L1 will be described as being carried by the signal electrons.
  • the pinning region 36 is P-shaped.
  • the pinning region 36 is a surface pinning layer on the P + type incident side.
  • the acceptor concentration in the pinning region 36 is high.
  • holes having a concentration of 1 ⁇ 10 17 cm- 3 or more are always present. These holes suppress the dark current generated due to the interface state existing on the light input surface 26a on the incident side of the photoelectric conversion unit 26.
  • the pinning region 36 is even better if it satisfies at least the first condition and further satisfies the second condition and the third condition.
  • the first condition is that the pinning region 36 neutralized by accumulating holes on the light input surface 26a on the incident side suppresses the generation of dark current on the light input surface 26a.
  • a plurality of bandgap levels are included, such as at the Si / SiO 2 interface, electrons in the valence band are excited to the conduction band via the bandgap levels. As a result, a dark current is generated.
  • the value of the dark current is represented by equation (1) based on the Shockley-Read-Hall model.
  • Each parameter constituting the equation (1) is as follows.
  • U Recombination rate.
  • vs thermal velocity.
  • N t Bandgap level density.
  • n Electron density in the conduction band.
  • p Hole density in the valence band.
  • ni Intrinsic carrier density.
  • k Boltzmann constant.
  • T Absolute temperature. If the recombination rate is positive, it is recombination. When the recombination rate is negative, it is the dark current generation rate.
  • the equation (2) can be obtained from the equation (1).
  • Et E i
  • the recombination rate U dep becomes maximum when the interface of Si / SiO 2 is depleted. Therefore, the equation (3) can be obtained from the equation (2).
  • Equation (5) is the ratio between when the interface is depleted and when holes are accumulated.
  • the second condition is to extend the life of a small number of carriers.
  • the minority carrier here is an electron.
  • the second condition is that the electrons of the generated electron-hole pairs promptly drift through the P-type photoelectric conversion unit 26 and reach the charge storage unit 33a.
  • the lifetime of minority carriers is affected by the number of crystal defects contained in the pinning region 36. Specifically, as the number of crystal defects contained in the pinning region 36 decreases, the life of minority carriers is extended. Crystal defects are likely to occur when the impurity concentration contained in the pinning region 36 is high and the high concentration region is large. This is because the atom size of the impurity is different from the atom size of silicon.
  • the pinning region 36 is provided by ion implantation.
  • the amount of ions injected corresponds to the integral of the impurity concentration and the impurity distribution.
  • the amount of ion implantation that can suppress the occurrence of crystal defects is shown to be 3 ⁇ 10 15 cm- 2 or less.
  • ion implantation introduces impurities into silicon and at the same time brings energy into silicon. This energy is also a factor that causes crystal defects. Crystal defects are healed by an annealing treatment performed after ion implantation. However, the crystal defects do not disappear completely.
  • the formation of the pinning region 36 is performed after the wiring region 13 is formed. Therefore, the annealing method is limited to the laser annealing method and the like. Therefore, it is necessary to reduce the occurrence of crystal defects caused by ion implantation.
  • the energy brought into silicon by ion implantation is proportional to the amount of impurities and energy. As a condition for suppressing the occurrence of crystal defects, it is shown that the product of the amount of impurities and energy is 5 ⁇ 10 15 keVcm -2 or less.
  • the third condition is that the electrons of the electron-hole pairs generated by the photoelectric conversion are promptly reached to the charge storage unit 33a. After passing through the P-shaped substrate portion 32, the electrons reach the charge storage portion 33a. Such an operation is performed by the movement of electrons due to the drift motion caused by the potential gradient.
  • the P-shaped substrate portion 32 is depleted. Similarly, the portion of the pinning region 36 on the substrate portion 32 side where the impurity concentration is low is depleted. That is, a sufficient electric field can be obtained in these regions. On the other hand, holes are accumulated in the vicinity of the optical input surface 26a in the pinning region 36 in order to suppress the dark current. That is, the vicinity of the optical input surface 26a is neutralized. The difference in impurity concentration facilitates the movement of electrons in the non-depleted region.
  • the pinning region 36 has an impurity concentration distribution as shown in FIG.
  • the larger the acceptor concentration the smaller the energy difference between the Fermi level and the valence band end.
  • the impurity concentration which is an acceptor, decreases in the pinning region 36 from the surface on the incident side toward the substrate portion 32. According to this impurity concentration distribution, the signal electrons quickly move to the substrate portion 32 due to drift even in the region where holes are accumulated in the pinning region 36.
  • Such an impurity concentration distribution can be obtained by ion implantation. Specifically, it can be obtained by setting the peak position of the impurity concentration distribution in the light input surface 26a or the oxide film provided on the light input surface 26a. According to this step, a distribution of impurity concentration that monotonically decreases from the light input surface 26a toward the substrate portion 32 is formed.
  • the distribution of impurity concentration can also be realized by other conditions.
  • ion implantation is performed with a low energy of 0.2 keV.
  • the peak position in the impurity concentration distribution is about 0.5 nm.
  • the injection energy is preferably 1 keV or less, for example.
  • the method using low-energy ion implantation has an advantage that there is no complexity of removing the oxide film after ion implantation.
  • the method of forming by low-energy ion implantation will be mainly described.
  • the light conversion unit 27 is a metal film formed on the surface of the pinning region 36.
  • the optical conversion unit 27 is in direct contact with the pinning region 36.
  • the light conversion unit 27 is formed of a metal nitride containing aluminum, silver, gold, copper, and titanium nitride (TiN).
  • the thickness of the light conversion unit 27 is 10 nm or more and 30 nm or less.
  • FIG. 6 is an enlarged view of the area S2 of FIG.
  • the light conversion unit 27 of the pixel 8 is separated from the light conversion unit 27 of another pixel 8 adjacent to the pixel 8.
  • This separation means that the plasmon generated in one optical conversion unit 27 does not reach the adjacent optical conversion unit 27.
  • a gap G is provided at the boundary of the pixel 8.
  • the plasmon generated in one optical conversion unit 27 reaching the other adjacent optical conversion unit 27 is also referred to as interference of the optical conversion unit 27. According to this gap G, the interference of the optical conversion unit 27 can be suppressed between the pixels 8.
  • the electric field is strong at the edge 27e of the optical conversion unit 27.
  • the edge 27e emits a strong evanescent light L2a. Therefore, the edge 27e of the light conversion unit 27 is formed on the pinning region 36. That is, the edge 27e of the light conversion unit 27 is formed on the light input surface 26a. In other words, the edge 27e of the optical conversion unit 27 is not formed on the deep trench 23 (isolation wall 22) that separates the pixels 8. Further, the deep trench 23 is formed in the region corresponding to the gap G formed between the light conversion units 27. According to this configuration, the strong evanescent light L2a generated at the edge 27e can be absorbed in the pinning region 36 and the base portion 32.
  • the evanescent light L2a that can be converted into signal electrons can be increased. That is, it becomes possible to effectively use the evanescent light L2a.
  • the gap G also suppresses the propagation of plasmons to the adjacent pixels 8. As a result, the occurrence of crosstalk is suppressed.
  • the gap G reduces the area of the optical conversion unit 27 that generates the evanescent light L2a.
  • the solid-state image sensor 1 has a microlens 17 (see FIG. 4) provided on the light conversion unit 27. According to the microlens 17, the incident light L1 can be focused on the light conversion unit 27. As a result, it is possible to prevent a decrease in sensitivity.
  • the light conversion unit 27 receives incident light L1 such as near infrared light.
  • the light conversion unit 27 generates the absorbed light L2.
  • the absorbed light L2 includes an evanescent light L2a and a propagating light L2b.
  • the light conversion unit 27 outputs the evanescent light L2a to the pinning region 36 and the base unit 32.
  • the light conversion unit 27 may include a grain structure. The size of the grain structure is equal to or smaller than the wavelength of the incident light L1.
  • the light conversion unit 27 may include the uneven structure shown in FIG. 7.
  • the uneven structure includes a plurality of ridges 27a (a plurality of convex portions). The distance between the ridges 27a may be, for example, 100 nm or more and may be a wavelength or less.
  • the ridge 28a projects from the optical input surface 26a.
  • the light conversion unit 27 may include a periodic structure.
  • a periodic structure for example, a plurality of spherical structures or columnar structures may be used. Further, such a structure can be roughly classified into a pattern structure and a fine particle structure.
  • the pattern structure examples include a diffraction grating, a hall array, a disk array, a slit array, an antenna array, and a bullseye array.
  • the diffraction grating may be, for example, a striped one-dimensional array structure, a square lattice-like or a triangular lattice-like two-dimensional array structure.
  • the hole array may have holes of circular, rectangular and triangular shapes.
  • the disk array may have a disk shape of a disk, a rectangle, a triangle, and a hemisphere.
  • the slit array may have a one-dimensional structure, a cross-shaped structure, and an asterisk-shaped structure.
  • the slit arrays may be arranged in a square grid or a triangular grid, respectively.
  • the antenna array may have a particle pair structure, a rod pair structure and a bowtie type structure.
  • a structure including an opening and a concentric concavo-convex structure may be arranged in a square lattice or a triangular lattice.
  • fine particles formed of a metal material are exemplified.
  • a metal nitride containing aluminum, silver, gold, copper, titanium nitride and the like may be adopted.
  • the shape of the fine particles include spherical nanoparticles, metal nanoshells, metal nanorods, and metal nanowires.
  • the fine particle structure utilizes localized surface plasmon resonance. When spherical nanoparticles are used as the fine particle structure, near-infrared resonance can be obtained by applying a gap mode that works between the particles.
  • the diameter of the spherical nanoparticles and the nanoshell may be 10 nm or more and 1 ⁇ m or less.
  • the diameter of the nanorod and nanowire may be 10 nm or more and 300 nm or less.
  • the lengths of the nanorods and nanowires may be 50 nm or more and 10 ⁇ m or less.
  • nitrogen-based nanoparticles such as TiN and high refractive index nanoparticles such as Si utilizing Mie scattering are also exemplified.
  • These fine particle structures may be formed by using a chemical synthesis method, a sputtering method and a vacuum deposition method. According to the vacuum vapor deposition method, an island-shaped film having a grain structure is formed.
  • the structure utilizing surface plasmon resonance based on periodicity has a period of 100 nm or more and is equal to or less than the wavelength of incident light.
  • the structure utilizing surface plasmon resonance based on the gap mode preferably has an intermetal distance such as a particle pair equal to or less than the wavelength of incident light.
  • the distance between metals is more preferably 1 nm or more and 100 nm or less.
  • the solid-state image sensor 1 components such as a charge storage unit 33a, a photodiode pinning layer 33b, a transistor, and wiring 18 are formed on a part of the silicon wafer on the wiring region 13 side.
  • the sensor substrate 12 is formed.
  • the sensor substrate 12 is joined to the support substrate 11.
  • the surface on the side where the charge storage portion 33a or the like is formed in the previous step is joined to the support substrate 11.
  • the support substrate 11 may use a wafer on which a circuit is formed. In this case, the degree of integration can be further increased.
  • the thickness of the sensor substrate 12 is 2 ⁇ m or more and 3 ⁇ m or less in the case of normal visible light.
  • the thickness of the sensor substrate 12 is 3 ⁇ m or less, near-infrared light can be sufficiently detected.
  • the pinning region 36 is formed. Specifically, boron is ion-implanted into the optical input surface 26a of the sensor substrate 12.
  • the ion implantation energy is 0.2 keV or more and 1 keV or less.
  • laser annealing for activation is performed. Ion implantation may be performed through the oxide film. In this case, the distribution of impurity concentration has its peak formed at the interface on the incident side or in the oxide film.
  • the ion implantation amount is preferably 1 ⁇ 10 14 cm- 2 or more and 3 ⁇ 10 15 cm- 2 or less.
  • a deep trench 23 for preventing crosstalk is formed.
  • FIG. 14 is a graph showing the distribution of impurity concentrations when monovalent boron ions are ion-implanted based on predetermined conditions.
  • the predetermined conditions are that the ion implantation energy is 0.2 keV and the dose amount is 1 ⁇ 10 15 cm- 2 .
  • Graph G14a shows the concentration distribution formed under this condition.
  • the ion implantation energy is 0.5 keV and the dose amount is 1 ⁇ 10 15 cm- 2 .
  • Graph G14b shows the concentration distribution formed under this condition.
  • the position of the peak P14 is 0.5 nm.
  • the concentration of the peak P14 is about 2 ⁇ 10 + 21 cm- 3 . Considering the measurement limit, it can be said that boron ions are distributed to a depth of about 10 nm.
  • laser annealing is performed.
  • the annealing can be performed without damaging the transistor and the wiring formed on the wiring region 13 side.
  • the energy of laser annealing is reduced. That is, it does not melt the silicon. By satisfying this condition, it is possible to prevent the impurity concentration of the molten portion from becoming constant.
  • the state of insufficient impurity activity can be changed to the state of sufficient impurity activity.
  • the natural oxide film is removed.
  • the light conversion unit 27, which is a metal film is formed.
  • the light conversion unit 27 may be formed by using a method according to the above-mentioned structure. Then, the color filter 16 and the microlens 17 are formed on the light conversion unit 27.
  • the solid-state image sensor 1 of the embodiment generates evanescent light L2a in the light conversion unit 27 which is a metal film. Almost all the evanescent light L2a is absorbed by the photoelectric conversion unit 26. In other words, the evanescent light L2a is absorbed not only in the substrate portion 32 of the photoelectric conversion unit 26 but also in the pinning region 36. That is, the evanescent light L2a is converted into signal electrons in the substrate portion 32 and the pinning region 36.
  • the optical conversion unit 27 is in direct contact with the pinning region 36. In other words, there is no insulating film between the optical conversion unit 27 and the photoelectric conversion unit 26 which is composed of silicon and performs photoelectric conversion.
  • the evanescent light L2a reaches the photoelectric conversion unit 26 without being attenuated by the insulating film or the like.
  • the light conversion unit 27, which is a metal film, and the pinning region 36 are in direct contact with each other.
  • the “direct” includes not only a configuration in which no film or layer is sandwiched between the optical conversion unit 27 and the pinning region 36, but also a configuration in which a film or layer in which the attenuation of the evanescent light L2a can be ignored is sandwiched.
  • a natural oxide film exists between the light conversion unit 27 and the pinning region 36, it can be said that the light conversion unit 27 and the pinning region 36 are substantially in direct contact with each other.
  • Evancent light L2a has a rapid decay of strength in silicon.
  • the evanescent light L2a is easily absorbed by silicon. Therefore, when detecting near-infrared light, it is not necessary to increase the thickness of silicon for photoelectric conversion.
  • 1 ⁇ m is exemplified as the thickness of silicon for photoelectric conversion.
  • the depth of the trench for preventing crosstalk can be made shallow. The formation of the trench can be facilitated and the width of the trench can be reduced.
  • holes are accumulated on the surface of the pinning region 36 on the incident side. As a result, it is possible to suppress the generation of dark current due to the bandgap levels existing in large numbers at the Si / SiO 2 interface.
  • the internal quantum efficiency of the solid-state image sensor 1 having the above configuration is almost 100%.
  • the internal quantum efficiency is a ratio in which the number of photons incident on silicon is the denominator and the number of signal electrons is the numerator.
  • the internal quantum efficiency excludes losses such as reflectance on the sensor surface.
  • the solid-state image sensor 1 can be considered that a so-called dead layer is not substantially formed in the pinning region 36.
  • the pixel 8 of the embodiment has a photoelectric conversion unit 26 that generates a signal voltage corresponding to the absorbed light L2 received from the optical input surface 26a, and an absorbed light that receives the incident light L1 and is output to the photoelectric conversion unit 26.
  • An optical conversion unit 27 provided on the optical input surface 26a of the photoelectric conversion unit 26 is provided so as to generate light L2.
  • the photoelectric conversion unit 26 has a first region 29 having a first acceptor concentration (first impurity concentration) and a second acceptor concentration (second impurity concentration) higher than the first acceptor concentration. It has 2 regions 31 and 2.
  • the second region 31 includes the light input surface 26a and comes into direct contact with the light conversion unit 27.
  • the light conversion unit 27 receives the incident light L1 and generates the absorbed light L2 including the evanescent light L2a.
  • the solid-state image sensor 1 of the embodiment controls the operation of the pixel unit 2 having the isolation wall 22 provided between the plurality of pixels 8 arranged in a two-dimensional shape and the pixels 8 adjacent to each other, and the pixel unit 2. It includes a pixel control unit 3 that generates a control signal to be generated, and a signal processing unit 4 that receives a signal voltage generated by the pixel unit 2.
  • the pixel 8 generates a photoelectric conversion unit 26 that generates a signal voltage corresponding to the absorbed light L2 received from the optical input surface 26a and an absorbed light L2 that receives the incident light L1 and is output to the photoelectric conversion unit 26. As described above, it has an optical conversion unit 27 provided on the optical input surface 26a of the photoelectric conversion unit 26.
  • the photoelectric conversion unit 26 has a first region 29 having a first acceptor concentration and a second region 31 having a second acceptor concentration higher than the first acceptor concentration.
  • the second region 31 includes the light input surface 26a and comes into direct contact with the light conversion unit 27.
  • the light conversion unit 27 receives the incident light L1.
  • the light conversion unit 27 generates the absorbed light L2 including the evanescent light L2a.
  • the light conversion unit 27 of the pixel 8 receives the incident light L1.
  • the light conversion unit 27 generates the absorbed light L2 output to the photoelectric conversion unit 26.
  • the optical input surface 26a of the photoelectric conversion unit 26 is in direct contact with the optical conversion unit 27. Therefore, the absorbed light L2 is input to the photoelectric conversion unit 26 without being attenuated.
  • the absorbed light L2 is more easily absorbed by the photoelectric conversion unit 26 than the incident light L1. Therefore, the light receiving sensitivity can be increased.
  • the photoelectric conversion unit 26 has a second region 31 including an optical input surface 26a. Therefore, the second region 31 has a higher acceptor concentration than the first region 29. According to this configuration, the dark current generated at the interface can be suppressed. That is, according to the pixel 8, the light receiving sensitivity can be increased and the dark current can be suppressed. Therefore, a good image can be obtained by using the pixel 8.
  • the second acceptor concentration in the second region 31 decreases as it approaches the first region 29. According to this configuration, the minority carriers generated by the absorption of the absorbed light L2 in the second region 31 can be quickly moved to the first region 29. As a result, the minority carriers do not disappear, so that the signal voltage caused by the minority carriers can be captured well. That is, the conversion efficiency from the absorbed light L2 to the signal voltage due to the minority carriers is increased. Therefore, the light receiving sensitivity can be further increased.
  • the second region 31 is formed by doping having a dose amount of 1 ⁇ 10 14 cm- 2 or more and 3 ⁇ 10 15 cm- 2 or less. Further, the second region 31 is formed by doping so that the injection energy is 0.2 keV or more and 1 keV or less. According to this configuration, it is possible to suppress the occurrence of crystal defects in the second region 31. Therefore, the life of the minority carriers generated according to the absorbed light L2 can be extended. As a result, minority carriers will not disappear. Therefore, the signal voltage caused by the minority carrier can be captured well. That is, the conversion efficiency from the absorbed light L2 to the signal voltage due to the minority carriers is increased. Therefore, the light receiving sensitivity can be further increased.
  • the isolation wall 22 is provided between the photoelectric conversion units 26 in the pixels 8 adjacent to each other.
  • One optical conversion unit 27 adjacent to each other is separated from the other optical conversion unit 27.
  • the side portion of the optical conversion unit 27 is formed on the optical input surface 26a of the photoelectric conversion unit 26. According to this configuration, it is possible to prevent the absorbed light L2 generated by the light conversion unit 27 of one pixel 8 from reaching another adjacent pixel 8. Therefore, the occurrence of crosstalk can be suitably suppressed.
  • the pixel 8 and the solid-state image sensor 1 according to the present embodiment are not limited to the above embodiment.
  • ⁇ Modification example 1> 8 and 9 are diagrams for explaining the solid-state image sensor 1A according to the first modification.
  • the incident light L1 is not parallel light.
  • the incident angle ⁇ of the incident light L1 with respect to the pixel 8 arranged near the center and the incident angle ⁇ of the incident light L1 with respect to the pixel 8 arranged away from the center may be different from each other. is there. That is, the main ray of the incident light L1 extends from the center of the camera lens 90 toward the solid-state image sensor 1A.
  • the incident angle ⁇ is 0 degrees.
  • the incident angle ⁇ is large in the peripheral portion of the solid-state image sensor 1A.
  • the distance between the camera lens 90 and the solid-state image sensor 1A becomes narrower. According to this configuration, the incident angle ⁇ in the peripheral portion of the solid-state image sensor 1A is further increased.
  • the incident angle ⁇ with respect to the pixel 8 is different, the incident angle ⁇ with respect to the light conversion unit 27 is different.
  • the period of the ridge 27a is determined according to the wavelength of the incident light L1.
  • the period of the light conversion unit 27 corresponds to the wavelength of the incident light L1.
  • the period of the optical conversion unit 27 may not correspond to the wavelength of the incident light L1. That is, when the incident angle ⁇ is 0 degrees and the period is L, the substantial period when the incident angle ⁇ is L ⁇ cos ⁇ .
  • the incident angle ⁇ is an angle with respect to the normal direction A on the light input surface 26a.
  • FIG. 9A is an enlarged view showing the relationship between the pixel 8 arranged near the center of the light receiving surface in the solid-state image sensor 1A and the incident light L1 and the optical conversion unit 27.
  • FIG. 9B is an enlarged view showing the relationship between the pixel 8 arranged near the right end of the light receiving surface in the solid-state image sensor 1A and the incident light L1 and the optical conversion unit 27.
  • the period LS of the optical conversion unit 27 of the pixel 8 at a certain position can be obtained by the following equation (6).
  • L is the period of the optical conversion unit 27 when the incident angle ⁇ is zero.
  • the period of the optical conversion unit 27 included in the pixel 8 arranged at the edge of the light receiving surface is larger than the period of the optical conversion unit 27 included in the pixel 8 arranged in the center of the light receiving surface.
  • the solid-state image sensor 1A may adjust the position of the microlens 17 so as to correspond to the incident angle ⁇ of the incident light L1. In this case, the decrease in the amount of light in the pixel 8 having a large incident angle ⁇ is suppressed.
  • the solid-state image sensor 1A may be provided with a configuration in which a desired potential is applied to the optical conversion unit 27 included in the solid-state image sensor 1B of the modification 2 described later. This technique can be widely applied to the solid-state image pickup apparatus shown in FIGS. 15 and 16.
  • ⁇ Modification 2> 10 and 11 are diagrams for explaining the solid-state image sensor 1B according to the second modification.
  • the pinning region 36 and the optical conversion unit 27B are in direct electrical contact with each other.
  • the contact state between the pinning region 36 and the optical conversion unit 27B is ohm contact due to the high impurity concentration in the pinning region 36. Therefore, the potential of the optical conversion unit 27B is the same as the potential of the pinning region 36 where the holes are accumulated.
  • the potential of the pinning region 36 where holes are accumulated may be the ground potential.
  • the P-type photoelectric conversion unit 26 may be completely depleted and the electric field may be strengthened toward the charge storage unit 33a. In this case, the potential of the pinning region 36 where the holes are accumulated is a negative potential. When holes are used as signal carriers, a positive potential is applied.
  • the solid-state image sensor 1B according to the second modification has a configuration in which a desired potential is applied to the light conversion unit 27B.
  • FIG. 10 is a cross-sectional view schematically showing a peripheral portion of the sensor chip.
  • the solid-state image sensor 1B according to the second modification includes a voltage application unit 40 in addition to the configuration of the solid-state image sensor 1.
  • the voltage application portion 40 includes a wiring 18B provided in the wiring region 13, a dicing region 41 (conductive portion) having a large acceptor concentration provided in the base portion 32, and a dicing surface 26b formed on the end surface of the base portion 32. ,including.
  • the end of the wiring 18B which is an aluminum pad, is electrically connected to the dicing region 41.
  • the wiring 18B is adjacent to the guard ring region S4 including the N + type region 42 provided on the base portion 32.
  • the dicing region 41 is a P + type.
  • One end of the dicing region 41 forms a part of the dicing surface 26b.
  • the dicing region 41 is, for example, a region doped with boron by ion implantation. As a result, the electrical resistance of the dicing region 41 is small.
  • the dicing surface 26b is the end surface of the sensor chip.
  • the sensor chip is an element in which a support substrate 11 and a sensor substrate 12 on which a pixel portion 2 or the like is formed are bonded to each other.
  • a plurality of solid-state image sensors 1B are formed on a silicon wafer and are separated by dicing.
  • the dicing surface 26b is a surface formed by this dicing. That is, the dicing surface 26b is a cut surface generated when the solid-state image sensor 1B is cut out from the silicon wafer. Due to dicing, a plurality of defects are generated on the dicing surface 26b. This defect aids electrical conductivity. That is, the dicing surface 26b has electrical conductivity.
  • the dicing surface 26b includes at least a dicing region 41, a base portion 32, and a pinning region 36 included in the base portion 32.
  • the dicing surface 26b may include an optical conversion unit 27B.
  • the dicing region 41 is electrically connected to the pinning region 36 via the dicing surface 26b. That is, a negative potential can be applied to the pinning region 36. Then, the pinning region 36 forms ohmic contact with the optical conversion unit 27B. Therefore, the negative potential is also applied to the optical conversion unit 27B via the dicing region 41, the dicing surface 26b, and the pinning region 36.
  • the optical conversion unit 27B has a configuration in which the ridges 27b are electrically connected to each other. Further, the optical conversion units 27B included in the pixels 8 adjacent to each other may be electrically connected to each other.
  • the solid-state image sensor 1B has a bridge 45 that electrically connects the optical conversion units 27B adjacent to each other. The bridge 45 may connect two optical conversion units 27B. The bridge 45 may connect two or more optical conversion units 27B. As a result, even when the isolation wall 22 is present, the pinning region 36 and the optical conversion unit 27B are electrically connected between the pixels 8. That is, a negative potential is supplied to the pinning region 36 and the optical conversion unit 27B of all the pixels 8.
  • the step of removing the natural oxide film immediately before forming the light conversion unit 27B, which is a metal film, may be omitted in order to shorten the step. Further, if the film thickness of the natural oxide film is thick, for example, if about 1 nm remains, the light conversion unit 27B may be electrically suspended. However, there is no problem due to the light conversion unit 27B being electrically suspended. In the natural oxide film, there is a slight loss of evanescent light L2a. However, the loss of evanescent light L2a due to the natural oxide film can be ignored.
  • the solid-state imaging device of the second modification has a plurality of pixels arranged in a two-dimensional shape, a pixel portion having an isolation wall provided between the pixels adjacent to each other, and an operation of the pixel portion. It includes a pixel control unit that generates a control signal to be controlled, and a signal processing unit that receives a signal voltage generated by the pixel unit.
  • the pixels are a photoelectric conversion unit that generates a signal voltage according to the absorbed light received from the optical input surface, and the optical input surface of the photoelectric conversion unit so as to receive the incident light and generate the absorbed light. It has an optical conversion unit provided in.
  • the photoelectric conversion unit includes the optical input surface and comes into direct contact with the optical conversion unit.
  • the light conversion unit receives the incident light and generates the absorbed light including the evanescent light.
  • the optical conversion unit receives a negative potential when the signal carrier is an electron and a negative potential when the signal carrier is a hole.
  • the optical conversion unit 27B receives a negative potential. According to this configuration, the photoelectric conversion unit 26 is completely depleted. Further, according to this configuration, the electric field toward the charge storage unit 33a becomes large. As a result, the signal carriers are quickly drifted to the charge storage unit 33a without recombination. Therefore, the sensitivity is improved and the reaction speed is increased.
  • one optical conversion unit 27B adjacent to each other is electrically connected to the other optical conversion unit 27B by the bridge 45. According to this configuration, a configuration capable of suppressing dark current can be obtained by a simple step.
  • FIG. 12 is a graph showing the relationship between the absorption coefficient of silicon and the wavelength. As shown in the graph G12a, the absorption coefficient of silicon sharply decreases from the visible light band G12b (360 nm to 830 nm) to the near infrared light band G12c (830 nm or more). As a result, in order to sufficiently absorb near-infrared light, it is necessary to increase the thickness of silicon.
  • FIG. 13 is a graph showing the relationship between the incident depth and the intensity of near-infrared light. According to FIG. 13, the degree of attenuation of the light intensity of each wavelength of near-infrared light due to the light absorption generated inside the silicon can be seen.
  • the horizontal axis of FIG. 13 is the depth inside the silicon.
  • the vertical axis is the light intensity.
  • graph G13a shows the attenuation of light intensity when the wavelength is 800 nm.
  • Graph G13b shows the attenuation of light intensity when the wavelength is 850 nm.
  • Graph G13c shows the attenuation of light intensity when the wavelength is 900 nm.
  • Graph G13d shows the attenuation of light intensity when the wavelength is 950 nm.
  • Graph G13e shows the attenuation of light intensity when the wavelength is 1000 nm.
  • Graph G13f shows the attenuation of light intensity when the wavelength is 1050 nm.
  • graph G13b is light used in a gesture input device, and shows the state of attenuation when the wavelength is 850 nm.
  • a 12 ⁇ m silicon layer is required for half of the incident light L1 to be absorbed by the silicon.
  • the graph G13d is light used for an in-vehicle device, and shows the state of attenuation when the wavelength is 950 nm.
  • a silicon layer of 24 ⁇ m is required in order for 1/2 of the incident light L1 to be absorbed by silicon. When the incident light L1 is further absorbed, it is necessary to further increase the thickness of the silicon layer.
  • the thickness of the silicon layer such as 12 ⁇ m and 24 ⁇ m and the thickness of the silicon layer larger than these are large considering that the sensor thickness of the image sensor for visible light is about 3 ⁇ m.
  • the thickness of such a silicon layer is large even considering that the pixel size of the image sensor for surveillance is about 3 ⁇ m.
  • a PN junction type diode is mainly used for the photodiode.
  • the photodiode is photoelectrically converted in a reverse bias state.
  • the depletion layer extends from the PN junction surface to the N-type region and the P-type region.
  • An electric field is formed in the depletion layer. According to this electric field, among the electron-hole pairs generated by light absorption, electrons move to the N-type region. Then, the holes move to the P-type region.
  • CMOS image sensors electrons are generally used as signals.
  • Non-Patent Document 1 discloses a technique for bending light that has been incident substantially vertically by a pyramid-shaped uneven structure provided on the surface of an incident surface. The bending of the light beam in the pixel substantially increases the thickness of the silicon.
  • DTI Deep Trench Isolation
  • This DTI acts as a mirror. According to the above structure, it is reported that the quantum efficiency is improved about twice in the wavelength region of 850 nm or more and 950 nm or less. However, the degree of improvement in quantum efficiency is only about double.
  • a process for creating a pyramid-shaped structure and a process for creating a DTI are required.
  • Reference 1 discloses another technique.
  • Reference 1 discloses a technique of utilizing plasmons to improve the sensitivity of near-infrared light.
  • Reference 1 discloses a technique of utilizing plasmons to improve the sensitivity of near-infrared light.
  • Reference 1 discloses a technique of utilizing plasmons to improve the sensitivity of near-infrared light.
  • the solid-state image sensor disclosed in Reference 1 generally has the structure shown in FIG.
  • the solid-state image sensor 100 includes a wiring region 101, a photodiode pinning layer 102, an N-type storage layer 103, a base portion 104, an insulating film 105, a negative charge holding region 106, and a fine metal structure layer 107.
  • the N-type storage layer 103 cooperates with the photodiode pinning layer 102 and the P-type substrate portion 104 to form a PN junction photodiode PD.
  • the fine metal structure layer 107 generates plasmons.
  • the fine metal structure layer 107 is formed on the photodiode PD.
  • An insulating film 105 is provided on the optical input surface 104a.
  • Examples of the insulating film 105 include a silicon oxide film.
  • the film thickness of the insulating film 105 is about 1 nm or more and 2 nm or less.
  • a negative charge holding region 106 is formed on the insulating film 105.
  • Examples of the negative charge holding region 106 include a hafnium oxide film having a film thickness of 11 nm formed by an atomic layer deposition method (Atomic Layer Deposition: ALD). Further, as the negative charge holding region 106, a hafnium oxide film having a film thickness of 50 nm formed by a physical vapor deposition method (Physical Vapor Deposition: PVD) can be mentioned.
  • a fine metal structure layer 107 is provided on the negative charge holding region 106. Further, a color filter and a microlens are formed on the color filter.
  • the fine metal structural layer 107 is formed of aluminum, silver, gold, copper, a metal nitride containing titanium nitride, or the like.
  • the thickness of the fine metal structure layer 107 is 10 nm or more and 30 nm or less.
  • the fine metal structure layer 107 exhibits a mesh-like periodic structure or a strip-like periodic structure. When light is incident on such a fine metal structure layer 107, a strong electric field is generated at the edge portion of the metal structure.
  • the solid-state image sensor 100 of Reference 1 includes an insulating film 105 having a size of 1 nm or more and 2 nm or less.
  • the film thickness of the negative charge holding region 106 is 50 nm or more.
  • the insulating film 105 and the negative charge holding region 106 attenuate the evanescent light. 80% of the generated evanescent light is lost due to the attenuation of the insulating film 105 and the negative charge holding region 106. Therefore, the effect of reducing the thickness of the substrate portion 104, which is the light receiving layer, is limited. Furthermore, it was necessary to use an expensive atomic layer deposition method for forming the negative charge holding region 106.
  • FIG. 16 is an example of another solid-state image sensor 200 using plasmon.
  • the solid-state image sensor 200 has a wiring region 201, a photodiode pinning layer 202, an N-type storage layer 203, a base portion 204, an insulating film 205, and a fine metal structure layer 207. That is, the solid-state image sensor 200 omits the negative charge holding region 106 from the solid-state image sensor 100.
  • An insulating film 205 is formed on the optical input surface 204a of the solid-state image sensor 200. The film thickness of the insulating film 205 is 2 nm.
  • the insulating film 205 is composed of a silicon oxide film and a hafnium oxide film.
  • a fine metal structure layer 207 is provided on the insulating film 205.
  • the fine metal structural layer 207 is formed of a metal nitride containing aluminum, silver, gold, copper and titanium nitride.
  • the thickness of the fine metal structure layer 207 is 10 nm or more and 30 nm or less.
  • the fine metal structure layer 207 exhibits a mesh-like periodic structure or a strip-like periodic structure.
  • a negative voltage of -2V or more and -10V or less is applied to the fine metal structure layer 207.
  • the fine metal structure layer 207, the insulating film 205, and the P-type substrate portion 204 form a MOS structure.
  • a negative voltage is applied to the fine metal structure layer 207, holes are accumulated in the region of the substrate portion 204 in the vicinity of the insulating film 205. These holes suppress the dark current generated from the interface state existing on the optical input surface 204a of the substrate portion 204.
  • a strong electric field is generated at the edge portion of the fine metal structure layer 207. This strong electric field produces exponentially decaying evanescent light and propagating light. Evanescent light is almost completely absorbed by the time it reaches a depth of 1 ⁇ m. As a result, the thickness of the light receiving layer in which the substrate portion 204 and the N-type storage layer 203 are combined can be reduced.
  • the evanescent light is attenuated by the insulating film 105 having a film thickness of 2 nm.
  • insulating films 105 and 205 were provided between the region for generating the evanescent light and the region for absorbing the evanescent light. Since the evanescent light is attenuated by the insulating films 105 and 205, the improvement of the light receiving sensitivity is hindered. As a result, it was difficult to capture near-infrared light favorably.
  • the back-illuminated CMOS image sensor of the reference example as shown in FIGS. 15 and 16 has base portions 104 and 204 made of silicon.
  • the thickness of the base portions 104 and 204 is about 3 ⁇ m.
  • the wavelength when the wavelength is 950 nm, the light intensity is attenuated to 91% at the position where the depths of the base portions 104 and 204 are 3 ⁇ m.
  • the substrate portions 104 and 204 absorb 9% of light.
  • the wavelength is 850 nm
  • the light intensity is attenuated to 76% at the position where the depths of the substrate portions 104 and 204 are 3 ⁇ m.
  • the substrate portions 104 and 204 absorb 24% of the light.
  • the unabsorbed light passes through the substrate portions 104 and 204.
  • a microlens, an antireflection film, and the like are formed on the optical input surfaces 104a and 204a, which are the back surfaces of the solid-state image sensors 100 and 200. According to the antireflection film, the reflection of the incident light L1 is suppressed.
  • the reflectance on the light input surface is 0%. Further, it is assumed that 100% of the electrons generated by the light absorbed inside the silicon are utilized as a signal. In other words, we assume that the internal quantum efficiency is the same as the external quantum efficiency. These assumptions lead to higher sensitivities. Based on these assumptions, the quantum efficiency (number of signal electrons / number of incident photons) is 9% when the wavelength of light is 950 nm. Further, when the wavelength of light is 850 nm, it is 24%.
  • Non-Patent Document 1 scatters incident light to substantially increase the thickness of silicon.
  • the quantum efficiency is improved to 16% at a wavelength of 950 nm and 33% at a wavelength of 850 nm.
  • the quantum efficiency is 16% when the wavelength of the incident light L1 is 950 nm. Further, when the wavelength of the incident light L1 is 850 nm, the quantum efficiency is 33%.
  • Total film thickness of composite film 54 nm (silicon oxide film (1 nm), hafnium oxide film formed by ALD (3 nm), hafnium oxide film formed by PVD (sputtering) (50 nm)) Depth at which the intensity of the evanescent light L2a is attenuated to 1 / e: 45 nm (wavelength 950 nm), 36 nm (wavelength 850 nm). Further, the evanescent light is attenuated by the hafnium oxide film and the silicon oxide film. Further, the ratio of evanescent light reaching the silicon film is 23% when the wavelength of the incident light is 950 nm. Further, when the wavelength of the incident light is 850 nm, it is 30%. The film thickness of silicon is 3 ⁇ m. It is assumed that all evanescent light is absorbed in silicon. Reflections on the plane of incidence are assumed to be ignored.
  • the quantum efficiency based on evanescent light when the wavelength of the incident light is 950 nm is 23%. Further, when the wavelength of the incident light L1 is 850 nm, the quantum efficiency based on the evanescent light is 30%. Far-field light is not absorbed by the oxide film and hafnium oxide. As a result, the light in the distant field reaches the silicon film. Then, the light in the distant field is partially absorbed by the silicon film having a thickness of 3 ⁇ m. In this configuration, the far-field quantum efficiency is 9% when the wavelength of the incident light is 950 nm. Further, when the wavelength of the incident light is 850 nm, the quantum efficiency in the far field is 24%.
  • the intensity ratio of evanescent light to distant light is 3.5: 1.
  • the solid-state image sensor 1 of the present embodiment does not have an insulating film that attenuates the evanescent light L2a.
  • the quantum efficiency based on the evanescent light L2a will be examined.
  • the quantum efficiency of the evanescent light L2a generated when the wavelength of the incident light L1 is 950 nm is 100%.
  • the quantum efficiency of the evanescent light L2a generated when the wavelength of the incident light L1 is 850 nm is also 100%.
  • the quantum efficiency based on the light in the distant field will be examined.
  • the quantum efficiency of the far-field light generated when the wavelength of the incident light L1 is 950 nm is 9%.
  • the quantum efficiency of the far-field light generated when the wavelength of the incident light L1 is 850 nm is 24%. Further, the intensity ratio of the evanescent light L2a and the distant field light L1a is 3.5: 1.
  • the total quantum efficiency is 80% when the wavelength of the incident light L1 is 950 nm. Further, when the wavelength of the incident light L1 is 850 nm, the total quantum efficiency is 83%. Therefore, the quantum efficiency is significantly improved as compared with the solid-state image sensor that employs the techniques described in Non-Patent Documents 1 to 4.
  • Reference 3 discloses an image sensor having a metal film provided directly on silicon.
  • the technique disclosed in Reference 3 is a Schottky barrier type infrared CCD image sensor.
  • ⁇ Reference 3> Masafumi Kimata, Munetaka Ueno, "PtSi Schottky Barrier Infrared Image Sensor", Journal of the Infrared Society of Japan, Vol. 14, No. 2, pp. 17-21, Japan Infrared Society, 2005.
  • Reference 3 Fig. 2 (b) shows a cross-sectional view of the pixels of the Schottky barrier type infrared CCD image sensor.
  • the Schottky barrier type infrared CCD image sensor is simply referred to as "SBIRCCD".
  • the SBIRCCD has a detection unit using a Schottky barrier and a reading unit composed of a transfer gate and a CCD.
  • the detection unit has a Schottky barrier.
  • the Schottky barrier has a P-type substrate and a film made of platinum silicide (PtSi).
  • the film thickness of platinum silicide is about 5 nm.
  • the barrier height of the Schottky barrier is 0.2 eV.
  • the barrier height of the Schottky barrier is small. Therefore, the detector exhibits ohmic properties at room temperature. That is, the dark current is large.
  • the SBIRCCD is cooled using liquid nitrogen. Specifically, the SBIRCCD is cooled to about -196 ° C. As a result, the dark current is reduced, so that the Schottky characteristics can be confirmed.
  • the SBIRCCD receives light from the back side.
  • the light received by the SBIRCCD is mid-infrared light having a wavelength of 3 ⁇ m or more and 5 ⁇ m or less.
  • the energy of mid-infrared light is smaller than the silicon bandgap of 1.1 eV. Therefore, mid-infrared light is not absorbed by silicon.
  • Part of the mid-infrared light is absorbed by the platinum silicide film.
  • high-energy electrons and holes are generated by the photoelectric effect of the platinum silicide film. Holes with an energy of 0.2 eV or higher are injected into silicon across the Schottky barrier. That is, the holes become signals.
  • Reference 3 Fig. A potential diagram is shown in 1.
  • a pinning region cannot be provided in order to realize the Schottky barrier bonding. That is, it is not possible to reduce the dark current using the pinning region.
  • the platinum silicide film is formed on the silicon surface on the side opposite to the incident side of infrared light.
  • a reverse bias voltage is applied between the platinum silicide film and the substrate made of P-type silicon.
  • the SBIRCCD does not have a configuration corresponding to an optical conversion layer. That is, the SBIRCCD cannot utilize the plasmon effect.
  • FIG. 17 shows an enlarged view of the main parts of the optical conversion unit 27 and the photoelectric conversion unit 26 of the above embodiment (hereinafter referred to as “first embodiment”).
  • the optical conversion unit 27 of the first embodiment includes a plurality of ridges 27a.
  • the ridge 27a was provided so as to be in contact with the optical input surface 26a. That is, the ridge bottom surface 27a1 of the ridge 27a is in direct contact with the optical input surface 26a. In other words, the ridge bottom surface 27a1 is in direct contact with the surface of the pinning region 36.
  • the evanescent light generated in the optical conversion unit 27 is generated at the edge 27a2 sandwiching the ridge bottom surface 27a1.
  • the evanescent light is directly incident on the pinning region 36 located in the vicinity of the edge 27a2. That is, the light conversion unit 27 is in direct contact with the photoelectric conversion unit 26 without a layer such as an insulating layer. As a result, the attenuation of the evanescent light from the light conversion unit 27 to the photoelectric conversion unit 26 is suppressed. Therefore, the light receiving sensitivity can be increased.
  • the incident light L1 is incident on the pixel 8.
  • the incident direction of the incident light L1 is orthogonal to the direction in which the plurality of ridges 27a are separated from each other.
  • the incident light L1 is an electromagnetic wave. Therefore, the incident light L1 generates an electric field E1.
  • the direction of the electric field E1 is orthogonal to the traveling direction of the incident light L1. That is, the direction of the electric field E1 coincides with the direction in which the plurality of ridges 27a are separated from each other.
  • the electric field E1 causes a bias of free electrons. More specifically, the free electrons collect on the ridge side surface 27a3.
  • a strong electric field E2 is generated around the ridge 27a from one ridge side surface 27a3 toward the other ridge side surface 27a3.
  • Surface plasmon has a so-called electric field enhancing effect.
  • This electric field enhancing effect strongly depends on the vibration direction of the electric field E2. That is, a strong electric field E2 is generated in the vicinity of the ridge side surface 27a3 arranged in the vibration direction of the electric field E2. As a result, the surface plasmon is enhanced by this electric field E2. That is, the electric field E2 overlapping the region where the surface plasmon is generated enhances the surface plasmon.
  • a part of such an electric field E2 is referred to as an enhanced electric field E2a.
  • the position where the surface plasmon is generated is in the vicinity of the edge 27a2. Then, the enhanced electric field E2a is generated in the vicinity of the edge 27a2.
  • the inventors paid attention to the relationship between the position where the surface plasmon is generated, the position where the electric field E2 is generated, and the position of the pinning region 36 which absorbs the evanescent light. If the silicon sensitivity layer (including the pinning region 36) is arranged so as to overlap the region where the surface plasmon is enhanced, the evanescent light due to the enhanced surface plasmon can be efficiently absorbed. That is, the relationship between the position where the surface plasmon is generated and the position where the electric field E2 is generated is set to the positional relationship so that the electric field enhancing effect can be more preferably obtained. Further, the pinning region 36 is arranged at a position where the evanescent light due to the enhanced surface plasmon is efficiently absorbed.
  • the structures of the optical conversion unit 27C and the photoelectric conversion unit 26C included in the pixels 8C of the solid-state image sensor 1C of the second embodiment will be described in detail with reference to FIG.
  • the other configurations of the pixels 8C of the solid-state image sensor 1C of the second embodiment are the same as those of the pixels 8 of the solid-state image sensor 1 of the first embodiment. Therefore, detailed description about them will be omitted.
  • FIG. 18 shows an enlarged view of the main part of the pixel 8C of the solid-state image sensor 1C of the second embodiment.
  • the pixel 8C of the solid-state image sensor 1C includes an optical conversion unit 27C and a photoelectric conversion unit 26C. Then, similarly to the pixel 8 of the solid-state image sensor 1 of the first embodiment, the optical conversion unit 27C is in direct contact with the photoelectric conversion unit 26C. On the other hand, the optical conversion unit 27C of the second embodiment is embedded in the photoelectric conversion unit 26C.
  • the photoelectric conversion unit 26C includes a groove 51 (recess).
  • the groove 51 is provided on the main surface 50a that receives the incident light L1 in the substrate constituting the photoelectric conversion unit 26C.
  • the groove 51 is open to the main surface 50a.
  • the groove 51 is surrounded by a pair of groove wall surfaces 52 and a groove bottom surface 53.
  • a pinning region 36 (second region) having a high impurity concentration is formed at a predetermined depth from the main surface 50a, the groove wall surface 52, and the groove bottom surface 53. Therefore, the groove 51 is surrounded by the groove wall surface 52 and the groove bottom surface 53, which are the surfaces of the pinning region 36.
  • a metal layer 61 (metal portion) having a metal nanostructure is formed in the groove 51.
  • the plurality of metal layers 61 constitute the optical conversion unit 27C. That is, the metal layer 61 corresponds to the ridge 27a.
  • the side surface 62 of the metal layer 61 is in contact with the groove wall surface 52.
  • the bottom surface 63 of the metal layer 61 is in contact with the bottom surface 53 of the groove.
  • the upper surface 64 of the metal layer 61 is flush with the main surface 50a. That is, the metal layer 61 is surrounded by three surfaces except the upper surface 64.
  • the distance between one groove 51 and another groove 51 adjacent to the groove 51 matches the distance between the metal layers 61.
  • the spacing between the grooves 51 is, for example, greater than the thickness of the pinning region 36. Then, it can be said that the metal layers 61 and the photoelectric conversion unit 26C are alternately arranged along the direction orthogonal to the incident direction. Further, it can be said that the metal layer 61 is sandwiched between the pinning regions 36 along the direction orthogonal to the incident direction.
  • the light conversion unit 27C generates evanescent light from the side surface 62 of the metal layer 61.
  • An electric field E2 is generated in the vicinity of the side surface 62. Therefore, the region where evanescent light is generated by the surface plasmon and the enhanced electric field E2a that enhances the surface plasmon overlap. As a result, evanescent light is generated according to the enhanced surface plasmon. Then, the evanescent light is incident on the pinning region 36 in contact with the side surface 62.
  • the optical input surface of the photoelectric conversion unit 26C is the groove wall surface 52 of the groove 51.
  • the optical conversion unit 27C of the pixel 8C included in the solid-state image sensor 1C of the second embodiment includes the upper surface 64 which is a light receiving surface for receiving the incident light L1 and is photoelectric so that the upper surface 64 is exposed from the photoelectric conversion unit 26C. It is embedded in the conversion unit 26C. According to this configuration, the bottom surface 63 of the optical conversion unit 27C comes into contact with the photoelectric conversion unit 26C. Further, the side surface 62 of the optical conversion unit 27C also comes into contact with the photoelectric conversion unit 26C. Then, in the vicinity of the side surface 62, the surface plasmon is enhanced by the enhanced electric field E2a. Therefore, it becomes possible to more efficiently absorb the evanescent light enhanced by the enhanced electric field E2a. As a result, the photoelectric conversion efficiency is further increased. Therefore, a better image can be obtained.
  • the light conversion unit 27C has been described as being composed of a striped metal layer 61.
  • the metal nanostructure constituting the light conversion unit 27C is not limited to the striped metal layer 61.
  • the light conversion unit 27 may be composed of a plurality of rectangular parallelepiped or cubic metal layers arranged in a grid pattern.
  • the light conversion unit 27 may be composed of a metal layer which is a plurality of cylinders arranged in a grid pattern. That is, various metal nanostructures capable of generating evanescent light by surface plasmons caused by incident light L1 may be appropriately adopted.
  • the substrate 50 is prepared (process ST1).
  • This step ST1 may include a step of forming the charge storage unit 33a, the photodiode pinning layer 33b, the transistor, and the wiring 18. Further, the step ST1 may include a step of joining the sensor substrate 12 to the support substrate 11. The step ST1 may include a step of adjusting the thickness of the sensor substrate 12.
  • step ST2 first step.
  • An etching method may be used for forming the groove 51, for example.
  • the pinning region 36 is formed (step ST3: second step).
  • an ion implantation method may be used for the formation of the pinning region 36. Specific conditions such as the ion implantation energy and the ion implantation amount may be the same as the method for manufacturing the solid-state image sensor 1 of the first embodiment.
  • the injection directions of the impurity ions include a plurality of directions (arrows B1 and B2) different from each other. For example, when forming the pinning region 36 including the main surface 50a and the groove bottom surface 53, the injection direction of the impurity ions is set to the normal direction of the main surface 50a (see arrow B1).
  • the injection direction of the impurity ions is set in an oblique direction inclined with respect to the normal of the main surface 50a (see arrow B2).
  • laser annealing is performed. This laser annealing activates the ion-implanted boron.
  • the specific conditions for this laser annealing may be the same as the method for manufacturing the solid-state image sensor 1 of the first embodiment.
  • an intermediate cambium 66 to be a metal layer 61 is formed (step ST4).
  • a method such as plating, thin film deposition and sputtering may be used.
  • a part of the intermediate cambium 66 formed outside the groove 51 is removed (step ST5). Chemical mechanical polishing (CMP) may be used for this removal. As a result, only the portion of the groove 51 of the intermediate cambium 66 remains, and that portion becomes the metal layer 61.
  • CMP Chemical mechanical polishing
  • the solid-state image sensor 1C can be obtained by the above steps ST1 to ST5.
  • the evanescent light is emitted from the side surface 62 of the metal layer 61, it seems that the more the number of the metal layers 61 is, the more evanescent light is emitted.
  • increasing the number of metal layers 61 affects the spacing (cycle) of the metal layers 61.
  • the period of the metal layer 61 is determined according to the wavelength of the incident light L1 and the resonance condition of the surface plasmon. Then, if the number of the metal layers 61 is increased and the period of the metal layers 61 is changed, the resonance condition of the surface plasmon may not be satisfied.
  • the effective wavelength which is the apparent wavelength of the incident light L1
  • the refractive index of the medium is n and the wavelength of the incident light L1 is ⁇
  • the effective wavelength of the incident light L1 in the medium is ⁇ / n.
  • the wavelength of the incident light L1 and the period of the metal layer 61 satisfy the resonance condition. That is, the effective wavelength ( ⁇ / n) of the incident light L1 in the medium and the period (p) of the metal layer 61 covered with the medium need only satisfy the resonance condition.
  • the refractive index (n) of the medium may be set so that That is, by adjusting the refractive index (n) of the medium of the incident light L1 around the metal layer 61, the resonance condition of the surface plasmon is satisfied and the period (p) of the metal layer 61 is narrowed to increase the amount of light absorption. be able to.
  • the pixel 8D of the solid-state image sensor 1D of the third embodiment has a high-dielectric film 71 (coating portion) that covers the light conversion unit 27C.
  • the high dielectric film 71 may be formed as a fourth step (ST6) after the above-mentioned step ST5.
  • the refractive index is related to the dielectric constant of the material.
  • a material having a high refractive index also has a high dielectric constant. That is, the refractive index of the high-dielectric film 71 is larger than the refractive index of air.
  • the material of the high dielectric film 71 for example, aluminum oxide (AlO 3: 10), hafnium oxide (HfO: 19), zirconium oxide (ZrO 2: 12), tantalum oxide (TaO 5: 22), and the like.
  • AlO 3: 10 aluminum oxide
  • hafnium oxide (HfO: 19) hafnium oxide
  • zirconium oxide (ZrO 2: 12) tantalum oxide (TaO 5: 22), and the like.
  • the numerical value written inside the parentheses is an approximate relative permittivity of each material.
  • materials such as silicon oxide (SiO 2 : 3.9), silicon nitride (SiN), and acrylic organic substances can also be adopted as the material of the high dielectric film 71. That is, as the material of the high dielectric film 71, a material having a higher dielectric constant than vacuum or air may be adopted.
  • the relationship between the refractive index of the high-dielectric film 71 that fills the light conversion unit 27C and the refractive index of air or vacuum is important.
  • the refractive index of the high-dielectric film 71 may be larger or smaller than the refractive index of the photoelectric conversion unit 26.
  • the refractive index of the high-dielectric film 71 may be equal to the refractive index of the photoelectric conversion unit 26.
  • the refractive index of the high-dielectric film 71 is equal to the refractive index of the photoelectric conversion unit 26, for example, a more effective effect is exhibited in the pixel 8E of the solid-state image sensor 1E having the structure shown in FIG. 22A. ..
  • the pixel 8E of the solid-state image sensor 1E has the same structure as the optical conversion unit 27 and the photoelectric conversion unit 26 of the first embodiment. Further, the pixel 8E of the solid-state image sensor 1E has a high-dielectric film 71E in which the light conversion unit 27 is embedded. The high-dielectric film 71E is embedded between the ridges 27a adjacent to each other. Further, the high dielectric film 71 also covers the upper surface 27a4 of the ridge. The thickness of such a high-dielectric film 71E is thicker than the height of the metal layer 61. Therefore, the high dielectric film 71E has a flat light receiving surface 71Ea.
  • the refractive index (n) affects the effective wavelength ( ⁇ / n) of the incident light L1. It is assumed that there is a difference between the refractive index of the material in the vicinity of the ridge upper surface 27a4 and the refractive index of the material in the vicinity of the ridge bottom surface 27a1. Under this assumption, the resonance condition is satisfied in the vicinity of the upper surface of the ridge 27a4. However, in the vicinity of the ridge bottom surface 27a1, the effective wavelength changes due to the difference in the refractive index of the material. Therefore, there may be cases where the resonance condition is not satisfied.
  • the difference between the refractive index of the material in the vicinity of the ridge upper surface 27a4 and the refractive index of the material in the vicinity of the ridge bottom surface 27a1 becomes small.
  • deviation from the resonance condition is suppressed around the ridge 27a, so that the light conversion unit 27 can efficiently generate evanescent light.
  • the high-dielectric film is not limited to the configuration shown in FIG. 22 (a). That is, the high-dielectric film may be arranged in the vicinity of the ridge 27a where the surface plasmon is generated.
  • FIG. 22B shows a solid-state image sensor 1F provided with a high-dielectric film 71F having a configuration different from that of the high-dielectric film shown in FIG. 22A.
  • the high dielectric film 71F covers the main surface 36a of the pinning region 36, the ridge upper surface 27a4, and the ridge side surface 27a3.
  • the thickness of such a high-dielectric film 71F is thinner than the period of the ridge 27a.
  • the thickness of the high-dielectric film 71F is thinner than the height of the ridge 27a.
  • the ridges 27a adjacent to each other are not embedded by the high dielectric film 71F. Therefore, the light receiving surface of the pixel 8F has an uneven shape.
  • the solid-state image sensor having a high dielectric film may further include a microlens.
  • the pixels 8G of the solid-state image sensor 1G shown in FIG. 23 (a) are obtained by further providing a microlens 72 on the solid-state image sensor 1E shown in FIG. 22 (a).
  • the microlens 72 may be arranged for each ridge 27a, for example.
  • the pixels 8H of the solid-state image sensor 1H shown in FIG. 23 (b) are obtained by further providing a microlens 72 on the solid-state image sensor 1F shown in FIG. 22 (b).
  • the solid-state image sensor 1H it further has a flattening film 73 in which a high-dielectric film 71F is embedded to form a flat surface.
  • the microlens 72 is provided on the light receiving surface of the flattening film 73.
  • the material of the flattening film 73 for example, glass and silicon oxide (SiO 2 ) may be adopted.
  • the pixels 8K of the solid-state image sensor 1K shown in FIG. 24 are obtained by further providing the microlens 72 on the solid-state image sensor 1D shown in FIG. These microlenses 72 may be provided for each pixel.
  • the optical conversion unit 27 included in the solid-state image sensor 1 is provided with a metal ridge 27a.
  • the entire ridge 27a was made of a metallic material.
  • the ridge 81a constituting the optical conversion unit 27P may adopt a structure in which layers made of different materials are laminated, as in the pixel 8P of the solid-state image sensor 1P shown in FIG. 25.
  • the ridge 81a constituting the optical conversion unit 27P has the same outer shape as the ridge 27a of the embodiment. Specifically, the width of the ridge 81a is the same as that of the ridge 27a of the embodiment. Further, the distance between the ridges 81a adjacent to each other is also the same as that of the ridge 27a of the embodiment.
  • the ridge 81a of the modified example 3 has a structure in which a metal material and a dielectric material are laminated in the traveling direction of the incident light L1. Such a laminated structure is a so-called MIM (metal-insulator-metal: metal-insulator-metal) structure.
  • the ridge 81a has a first metal layer 82, a dielectric layer 83, and a second metal layer 84.
  • the second metal layer 84 is in contact with the optical input surface 26a. That is, the lower surface of the second metal layer 84 is the ridge bottom surface 81a1.
  • the lower surface of the dielectric layer 83 is in contact with the upper surface of the second metal layer 84.
  • the lower surface of the first metal layer 82 is in contact with the upper surface of the dielectric layer 83.
  • the ridge side surface 81a2 of the ridge 81a includes the side surfaces of the first metal layer 82, the dielectric layer 83, and the second metal layer 84.
  • the thickness of the first metal layer 82 may be the same as the thickness of the second metal layer 84.
  • the thickness of the first metal layer 82 and the second metal layer 84 may be thinner than the thickness of the dielectric layer 83.
  • the central portion of the ridge 81a that does not contribute to plasmon vibration is formed by the dielectric layer 83.
  • the area physically occupied by the metal material that causes absorption loss in the ridge 81a becomes smaller.
  • the absorption loss caused by the metal material in the ridge 81a is reduced.
  • an enhanced electric field E2a due to surface plasmon resonance is formed in the photoelectric conversion unit 26.
  • the absorption efficiency in the photoelectric conversion unit 26, which is a silicon sensitivity layer can be further improved.
  • the ridge 81a adopting the MIM structure may be embedded in the photoelectric conversion unit 26C like the optical conversion unit 27C included in the solid-state image sensor 1C shown in FIG. That is, the solid-state image sensor 1Q of the modification 4 shown in FIG. 26 has pixels 8Q, and the pixels 8Q include an optical conversion unit 27Q.
  • the optical conversion unit 27Q is composed of a ridge 81a.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Toxicology (AREA)
  • Solid State Image Pick-Up Elements (AREA)
  • Light Receiving Elements (AREA)
  • Transforming Light Signals Into Electric Signals (AREA)
PCT/JP2020/027504 2019-07-22 2020-07-15 画素、固体撮像装置及び画素の製造方法 Ceased WO2021015070A1 (ja)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2021533975A JP7493250B2 (ja) 2019-07-22 2020-07-15 画素、固体撮像装置及び画素の製造方法

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
JP2019-134322 2019-07-22
JP2019134322 2019-07-22
JP2020018082 2020-02-05
JP2020-018082 2020-08-27

Publications (1)

Publication Number Publication Date
WO2021015070A1 true WO2021015070A1 (ja) 2021-01-28

Family

ID=74193473

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/027504 Ceased WO2021015070A1 (ja) 2019-07-22 2020-07-15 画素、固体撮像装置及び画素の製造方法

Country Status (2)

Country Link
JP (1) JP7493250B2 (https=)
WO (1) WO2021015070A1 (https=)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006509358A (ja) * 2002-12-09 2006-03-16 クォンタム セミコンダクター リミテッド ライアビリティ カンパニー Ｃｍｏｓ画像センサー
JP2010263158A (ja) * 2009-05-11 2010-11-18 Sony Corp ２次元固体撮像装置、及び、２次元固体撮像装置における偏光光データ処理方法
JP2012064703A (ja) * 2010-09-15 2012-03-29 Sony Corp 撮像素子および撮像装置
JP2015232599A (ja) * 2014-06-09 2015-12-24 ソニー株式会社 光学フィルタ、固体撮像装置、および電子機器
WO2016136502A1 (ja) * 2015-02-26 2016-09-01 ソニー株式会社 固体撮像素子、および電子装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006509358A (ja) * 2002-12-09 2006-03-16 クォンタム セミコンダクター リミテッド ライアビリティ カンパニー Ｃｍｏｓ画像センサー
JP2010263158A (ja) * 2009-05-11 2010-11-18 Sony Corp ２次元固体撮像装置、及び、２次元固体撮像装置における偏光光データ処理方法
JP2012064703A (ja) * 2010-09-15 2012-03-29 Sony Corp 撮像素子および撮像装置
JP2015232599A (ja) * 2014-06-09 2015-12-24 ソニー株式会社 光学フィルタ、固体撮像装置、および電子機器
WO2016136502A1 (ja) * 2015-02-26 2016-09-01 ソニー株式会社 固体撮像素子、および電子装置

Also Published As

Publication number Publication date
JPWO2021015070A1 (https=) 2021-01-28
JP7493250B2 (ja) 2024-05-31

Similar Documents

Publication Publication Date Title
JP7386830B2 (ja) 感光撮像素子および関連方法
US10484855B2 (en) Process module for increasing the response of backside illuminated photosensitive imagers and associated methods
US12501724B2 (en) Shallow trench textured regions and associated methods
US8569855B2 (en) Two-dimensional solid-state imaging device
US9373657B2 (en) System and method for fabricating a 3D image sensor structure
TWI841659B (zh) 用於檢測不同波長的光檢測裝置
US11258971B2 (en) Multi-function transfer gate electrode for a photodetector and methods of operating the same
JP2012151452A (ja) 裏面に設けられた金属テクスチャリングによって最適化された光検出器
EP3794643B1 (en) Integration of a short-wave infrared detector with cmos compatible substrates
JP7493250B2 (ja) 画素、固体撮像装置及び画素の製造方法
EP3714487B1 (en) Short-wave infrared detector array and fabrication methods thereof
CN117178367A (zh) 宽带图像装置及其制造方法
CN116724401A (zh) 固体拍摄装置以及固体拍摄装置的制造方法
Jansz et al. Extrinsic evolution of the stacked gradient poly-homojunction photodiode genre

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: 20844244

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021533975

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20844244

Country of ref document: EP

Kind code of ref document: A1