JP6010465B2 - Solid-state imaging device - Google Patents

Description

  The present invention relates to a solid-state imaging device having asymmetric pixels used for focus detection and the like.

  Camera modules and digital cameras mounted on mobile devices such as mobile phones and smartphones are equipped with an autofocus function almost as standard. A plurality of methods for realizing the autofocus function are known. For example, in recent camera modules and digital cameras, a contrast detection method (so-called contrast AF) that performs focus adjustment so as to maximize the contrast of an image, A phase difference detection method (hereinafter referred to as phase difference AF) that performs focus adjustment based on a phase difference due to parallax is often used.

  For the phase difference AF, for example, as in the solid-state imaging device described in Patent Document 1, focus detection pixels formed in an asymmetric shape in the left-right direction at a predetermined position in a light receiving region in which a plurality of pixels are arranged. A solid-state imaging device having (hereinafter referred to as phase difference pixel) is used. The phase difference pixel calculates the phase difference (parallax) based on the signal obtained by selectively receiving light incident from either the left or right direction, and drives the imaging lens so that the phase difference is reduced. To adjust the focus. Further, in Patent Document 1, the phase difference pixel is formed by making the color filter asymmetric in the left-right direction. However, the aperture is decentered with respect to the center of the photodiode, or the intralayer lens is asymmetrically formed. It is also known to form a phase difference pixel by doing so.

  A pixel formed in an asymmetric shape in the left-right direction as used for phase difference AF may also be used in a monocular 3D imaging device. The monocular 3D imaging device is an imaging device that simultaneously obtains two images for stereoscopic viewing with parallax on the left and right by one solid-state imaging device, and asymmetric pixels similar to phase difference pixels are arranged on the entire surface of the imaging region. A phase difference pixel is formed by providing two photodiodes for one on-chip microlens (Patent Document 2).

JP 2012-003009 A International Publication 2012/026292 Pamphlet

  As described above, since the phase difference pixel is formed in an asymmetric shape, a signal obtained is different from a normal symmetric shape pixel (hereinafter referred to as a normal pixel). For this reason, when capturing an image of a subject after phase difference AF, the pixel value of the phase difference pixel cannot be used as it is. It is necessary to substitute a pixel value calculated by interpolation using a normal pixel. As described above, when the pixel value of the phase difference pixel is gain-corrected or calculated by interpolation, the sensitivity or resolution at the portion where the phase-difference pixel is present depends on the specific aspect of the subject, the accuracy of gain correction or interpolation, and the like. May fall.

  Furthermore, in recent camera modules, digital cameras, and the like, there are cases where a drive mode for taking a moving image or adding pixels is prepared, but in a solid-state imaging device having a phase difference pixel, a phase difference pixel is used as described above. In some cases, the image quality may be deteriorated, so that these drive modes may not be adopted. For this reason, in addition to enabling accurate focus adjustment by phase difference AF, other conveniences may be impaired.

  Further, in the 3D imaging device in which the phase difference pixels are arranged on the entire surface, the phase difference AF can be performed by using signals obtained from the phase difference pixel for the right eye and the phase difference pixel for the left eye, respectively. When a 2D image is obtained by the apparatus, it is necessary to use only one of the right-eye image and the left-eye image, average the pixel values of the left and right phase difference pixels, or add them up. However, the number of pixels of the image obtained by these methods is half the number of pixels arranged in the imaging region, and the resolution of the captured image is lowered. In addition, since the right-eye or left-eye images all use pixel values obtained from phase difference pixels, the sensitivity is poor compared to the case of imaging with normal pixels that receive light uniformly from all directions.

  An object of the present invention is to provide a solid-state imaging device capable of suppressing a decrease in sensitivity and resolution of a captured image while being a solid-state imaging device capable of performing phase difference AF.

The solid-state imaging device of the present invention includes a first hybrid pixel, a second hybrid pixel, and a dimming control unit. The first hybrid pixel covers a part of an incident surface of a photodiode that photoelectrically converts incident light from a subject, and the amount of transmitted light changes according to an applied voltage, and the incidence of the photodiode A second dimming element that covers the remaining portion of the surface and changes the amount of transmitted light according to the applied voltage is provided. The second hybrid pixel is a pixel in which the positions of the first dimming element and the second dimming element are interchanged with respect to the first hybrid pixel. The dimming control unit applies a voltage to at least one of the first dimming element and the second dimming element, and makes one of the first dimming element and the second dimming element transparent and the other non-transparent. Thus, the first hybrid pixel and the second hybrid pixel are caused to function as a pair of phase difference pixels. The first hybrid pixel and the second hybrid pixel are provided at least in the central part and the peripheral part of the light receiving region in which the plurality of pixels are arranged, and the first hybrid pixel and the second hybrid pixel are arranged in the peripheral part in a predetermined first direction with respect to the central part. In one hybrid pixel, the area of the first dimming element is larger than the area of the second dimming element, and the second hybrid pixel disposed in the peripheral portion in the first direction with respect to the central portion is the first dimming element. Is smaller than the area of the second light control device.

  The first dimming element and the second dimming element include a pair of transparent electrodes and a layer provided between the pair of transparent electrodes and in which oriented particles are dispersedly arranged in a predetermined transparent substrate. is there. In this case, the dimming control unit adjusts the amount of light transmitted through the first dimming element and the second dimming element by adjusting the voltage applied between the pair of transparent electrodes to change the orientation direction of the oriented particles.

  In addition, the first dimming element and the second dimming element include a first transparent electrode, an oxidized color layer that is colored by being oxidized, and an insulating ion conductive material that allows ions involved in the coloring of the oxidized color layer to pass through. An electrochromic element formed by laminating a layer, a reduced coloring layer colored by being reduced through an ion conductive layer, and a second transparent electrode in this order may be used. In this case, the dimming control unit adjusts the voltage applied between the first transparent electrode and the second transparent electrode to control the coloring of the oxidation coloring layer and the reduction coloring layer, thereby controlling the first tuning electrode. Adjust the amount of light transmitted through the optical element and the second dimmer element

Further, the first dimming element and the second dimming element may include a pair of transparent electrodes and a polymer liquid crystal composite layer having a structure in which liquid crystal molecules are filled in a gap between polymer networks. In this case, the dimming control unit adjusts the amount of light transmitted through the first dimming element and the second dimming element by adjusting the voltage applied between the pair of transparent electrodes to change the alignment direction of the liquid crystal molecules .

  The first dimming element and the second dimming element include a first polarizing plate, a first transparent electrode, a liquid crystal layer composed of liquid crystal molecules, a second transparent electrode, and a second polarizing plate. A so-called liquid crystal display device-like device may be used. In this case, the dimming control unit adjusts the voltage applied to the first transparent electrode and the second transparent electrode, and controls the alignment direction of the liquid crystal molecules, thereby controlling the first dimming element and the second dimming element. Adjust the amount of transmitted light.

  In addition, the first hybrid pixel arranged in the peripheral portion in the direction opposite to the first direction with respect to the central portion has an area of the first dimmer element smaller than that of the second dimmer element, Thus, it is preferable that the second hybrid pixel disposed in the peripheral portion in the direction opposite to the first direction has a larger area of the first dimming element than that of the second dimming element.

  Note that the first hybrid pixel and the second hybrid pixel may be included in at least two colors of red, green, and blue.

  Further, all the pixels in the light receiving region where a plurality of pixels are arranged may be hybrid pixels.

  The first hybrid pixel and the second hybrid pixel may have a photoelectric conversion region divided into a plurality of element isolation regions as photodiodes. In this case, the first dimming element and the second dimming element are provided corresponding to each photoelectric conversion region. The element isolation region included in one first hybrid pixel or second hybrid pixel preferably has a larger amount of additive than the element isolation region for isolation from other pixels.

  Since the present invention includes a hybrid pixel that can be used as a normal pixel and functions as a phase difference pixel when performing phase difference AF, it is possible to perform imaging while being a solid-state imaging device capable of performing phase difference AF. Reduction in image sensitivity and resolution can be suppressed.

It is explanatory drawing which shows a solid-state imaging device. It is explanatory drawing which shows the arrangement | sequence of a color filter. It is sectional drawing of a solid-state imaging device. It is explanatory drawing which shows the light control element of a hybrid pixel, its wiring, etc. It is sectional drawing which shows schematic structure of a light control element. It is explanatory drawing which shows the effect | action of a solid-state imaging device. It is explanatory drawing which shows the solid-state imaging device using four light control elements. It is explanatory drawing in the case of obtaining the parallax on either side. It is explanatory drawing in the case of obtaining the upper and lower parallax. It is explanatory drawing in the case of obtaining diagonal parallax. It is explanatory drawing in the case of obtaining diagonal parallax by another method. It is sectional drawing showing the incident light to the hybrid pixel made into opaque on the left side. It is sectional drawing showing the incident light to the hybrid pixel made right side opaque. It is a graph showing the sensitivity ratio of the hybrid pixel with respect to a normal pixel. It is explanatory drawing which shows a mode that the width | variety (area) of a light control element is adjusted according to the position of a hybrid pixel. It is a graph showing the sensitivity ratio at the time of adjusting the width | variety (area) of a light control element according to the position of a hybrid pixel. It is explanatory drawing which shows a mode that the width | variety (area) of a light control element is adjusted according to the position of a hybrid pixel, when using four light control elements. It is explanatory drawing in the case of providing a hybrid pixel also in a red pixel and a blue pixel. It is explanatory drawing of the example which made all the pixels the hybrid pixel. It is explanatory drawing which shows another example which makes all the pixels a hybrid pixel. It is explanatory drawing which shows the example which provided two PD in 1 pixel. It is explanatory drawing which shows the example which made p ++ between two PD of 1 pixel. It is explanatory drawing which shows the solid-state imaging device of a honeycomb arrangement | sequence. It is explanatory drawing which shows the solid-state imaging device of another honeycomb arrangement | sequence. It is explanatory drawing which shows the solid-state imaging device of which a color filter is a Bayer arrangement. It is sectional drawing of a surface irradiation type solid-state imaging device. It is sectional drawing of the electrochromic element used as a light control element. It is sectional drawing of PNLC used as a light control element. It is sectional drawing of the liquid crystal device used as a light control element.

[First Embodiment]
As shown in FIG. 1, the solid-state imaging device 10 is a CMOS image sensor, and includes a light receiving region 11, a vertical scanning circuit 13, a horizontal scanning circuit 14, an output circuit 16, a control circuit 17, and the like.

  The light receiving region 11 is a region in which a plurality of pixels 21 are squarely arranged along the horizontal direction (X direction) and the vertical direction (Y direction). An image of a subject is formed in the light receiving region 11 by an imaging lens, and each pixel 21 accumulates electric charge corresponding to the amount of incident light by photoelectric conversion. Here, the horizontal direction (left-right direction) and the vertical direction (up-down direction) are the X-axis direction when the light receiving surface of the image sensor is regarded as the XY plane, and the Y-axis is the vertical direction or left-right direction. The vertical direction.

  The pixel 21 includes a photodiode (hereinafter referred to as PD) 22, a reset transistor (hereinafter referred to as Tr) 23, an amplifier transistor (hereinafter referred to as Ta) 24, and a selection transistor (hereinafter referred to as Ts) 25. Tr23, Ta24, and Ts25 are, for example, n-type MOS transistors. Each pixel 21 is provided with a color filter 41 (see FIG. 2) of any one of red (R), green (G), and blue (B). 21 photoelectrically converts the light of the color of the corresponding color filter 41.

  Note that the pixels 21 arranged in the light receiving region 11 are either normal pixels or hybrid pixels. The normal pixel is a pixel formed symmetrically in the horizontal direction and the vertical direction, and an imaging signal based on the electric charge accumulated in the normal pixel is used for forming a captured image. On the other hand, the hybrid pixel functions as a phase difference pixel for obtaining a parallax in the left-right direction (horizontal direction) when performing phase difference AF, and the structure of the hybrid pixel will be described later. It is a pixel that functions as a symmetric normal pixel. The normal pixel and the hybrid pixel are different in three-dimensional structure (specifically, the presence or absence of the light control elements 55A and 55B (see FIG. 3)), but the basic pixel for photoelectrically converting and reading out the charges is used. The electrical configuration is the same. For this reason, in FIG. 1, all the pixels in the light receiving region 11 are pixels 21 without distinguishing between normal pixels and hybrid pixels.

  The PD 22 is an element that generates a charge corresponding to the amount of incident light by photoelectric conversion. The anode is connected to the ground, and the cathode is connected to the gate electrode of Ta24. A connection portion between the cathode of the PD 22 and the gate electrode of the Ta 24 is a floating diffusion (hereinafter referred to as FD) 26, and charges generated by the PD 22 are accumulated therein.

  The source electrode of Tr23 is connected to FD26, and power supply voltage VDD (not shown) is applied to the drain electrode. The gate electrode of Tr 23 is connected to the reset line 27. The Tr 23 is turned on when a reset pulse is applied from the vertical scanning line 13 to the gate electrode via the reset line 26. When the Tr 23 is turned on, the power supply voltage VDD is applied to the FD 26, and charges accumulated in the FD 26 are discarded.

  As for Ta24, the gate electrode is connected to FD26, and the power supply voltage VDD is applied to the drain electrode. The source electrode from which the signal voltage is output is connected to the drain electrode of Ts25. The Ta 24 outputs a signal voltage corresponding to the amount of charge accumulated in the FD 26 to the source electrode.

  The drain electrode of Ts25 is connected to the source electrode of Ta24, and the source electrode is connected to the signal line. The gate electrode of Ts25 is connected to the row selection line 29. Ts 25 is turned on when a selection pulse is input from the vertical scanning circuit 13 via the row selection line 29, and outputs the signal voltage output from the Ta 24 to the signal line 28.

  The vertical scanning circuit 13 is a drive circuit for the pixels 21, and a reset line 27 and a row selection line 29 for each row are connected to each other. The vertical scanning circuit 13 inputs a reset pulse to the reset line 27 of the selected row, and also inputs a selection pulse to the row selection line 29 to control the operation of Tr23 and Ts25, as described above. The pixel 21 is operated.

  The horizontal scanning circuit 14 selects a column from which a signal voltage is read by turning on one of the column selection transistors (Tc) 32 provided on each signal line 28.

  The signal line 28 is a wiring for reading a signal voltage from each pixel 21, and is provided for each column of the pixels 21. A correlated double sampling (CDS) circuit 31 and a column selection transistor 32 are provided at the end of the signal line 28. The CDS circuit 31 operates based on the clock signal input from the control circuit 17 and is a voltage signal so that noise accompanying readout is removed from the pixel 21 on the row selection line 29 selected by the vertical scanning circuit 13. Is sampled and held. When the column selection transistor 32 is turned on by the horizontal scanning circuit 14, the voltage signal held by the CDS circuit 31 is output to the output circuit 16 via the output bus line 33.

  The output circuit 16 includes an amplifier 34 that amplifies the voltage signal input from the CDS 31 and an A / D conversion circuit 35 that converts the voltage signal amplified by the amplifier 34 into digital data. The gain of the amplifier 34 is variable and is appropriately adjusted according to the setting such as the shooting mode.

  The control circuit 17 is connected to each unit such as the vertical scanning circuit 13 and the horizontal scanning circuit 14, and comprehensively controls each unit of the solid-state imaging device 10. For example, the operations of the vertical scanning circuit 13 and the horizontal scanning circuit 14 operate based on a control signal such as a clock signal input from the control circuit 17. The operation of the CDS 31 and the gain of the amplifier 34 are also controlled by the control circuit 17.

  As shown in FIG. 2, the color filter 41 of the solid-state imaging device 10 has a long periodicity with a color arrangement of 6 × 6 pixels as one unit. More specifically, the color filter 41 is an array in which two types of subunits 41a and 41b each having a color array of 3 × 3 pixels are arranged at diagonal positions. The first subunit 41a has a green (G) filter arranged at the center and diagonal positions of 3 × 3 pixels, a blue (B) filter at the left and right center, and a red (R) filter at the top and bottom center. Is a unit. The second subunit 41b is the same as the first subunit 41a in that G filters are arranged at the center and diagonal positions of 3 × 3 pixels, but the B filter and the R filter are the same as the first subunit 41a. Are reversed, an R filter is disposed at the center of the left and right, and a B filter is disposed at the center of the top and bottom.

  For example, the color filter 41 has a long periodicity as compared with a Bayer arrangement in which the color arrangement of 2 × 2 pixels is one unit. Therefore, the solid-state imaging device 10 suppresses the occurrence of moire even without using an optical low-pass filter. be able to. Further, in the solid-state imaging device 10 employing the color filter 41, since RGB pixels always exist in the vertical direction (vertical direction) and the horizontal direction (horizontal direction), false colors can be suppressed and accurate color reproduction is possible. is there.

  The hybrid pixels 42a and 42b function as both normal pixels and phase difference pixels. For example, the hybrid pixels 42a and 42b are formed in the lower left G pixels of the upper right first subunit 41a and the upper left second subunit 41b. Pixels other than the hybrid pixels 42a and 42b are normal pixels 43 that uniformly receive light incident from the upper, lower, left, and right directions.

  When the hybrid pixels 42a and 42b function as phase difference pixels in order to perform phase difference AF, based on a voltage signal with a parallax obtained from the pair of hybrid pixels 42a and 42b (or a pixel value based on this voltage signal). Focus assessment is performed. Then, the position of the imaging lens is automatically adjusted so as to achieve the best focus according to the result of the focus evaluation. When the hybrid pixels 42a and 42b function as normal pixels in order to capture an image of the subject, the pixel values of the hybrid pixels 42a and 42b are captured as they are without gain correction or interpolation as with the other normal pixels 43. Used for image generation.

  Although the hybrid pixels 42a and 42b are provided uniformly over the entire surface of the light receiving region 11, the hybrid pixels 42a and 42b are not necessarily provided in the 6 × 6 pixel unit of the plurality of color filters 41. When each pixel 21 in the light receiving region 11 is viewed in units of 6 × 6 pixels of the color filter 41, there is a unit in which the hybrid pixels 42a and 42b are not provided. In the case of a unit in which the hybrid pixels 42a and 42b are provided, the unit is provided at the above location. In the unit where the hybrid pixels 42 a and 42 b are not provided, all are normal pixels 43.

  As shown in FIG. 3, the solid-state imaging device 10 is a backside illuminated (BSI) type CMOS image sensor, and includes a support substrate 51, a wiring layer 52, a p-type semiconductor substrate 53, a light control layer 54, and a color. The filter 41, the micro lens 57, and the like are included. The side where the wiring layer 52 is provided with respect to the p-type semiconductor substrate 53 is the “front surface” of the solid-state imaging device 10, and the side where the color filter 41, the light control layer 54 and the microlens 57 are provided is the “back surface”. is there. Light enters the solid-state imaging device 10 from the back side through the microlens 57, the color filter 41, and the light control layer 54. That is, the “back surface” is the light incident surface.

  The support substrate 51 is, for example, a silicon substrate. In the process of manufacturing the back-illuminated solid-state imaging device 10, a wiring layer is formed to expose the back surface of the p-type semiconductor substrate 53 and reduce the thickness of the p-type semiconductor substrate 53. Bonded to the surface of 52.

  The wiring layer 52 is formed on the surface of the p-type semiconductor substrate 53, and the transistors (Tr23, Ta24) that form each pixel 21 (the normal pixel 43 and the hybrid pixels 42a, 42b) by the wiring 52a provided in the wiring layer 52. , Ts25), vertical scanning circuit 13 for driving each pixel 21, horizontal scanning circuit 14, output circuit 16, control circuit 17, various circuits such as CDS 31, reset line 27, signal line 28, row selection line 29, Various wirings such as the output bus line 33 are formed.

  The PD 22 is formed by a PN junction of a p-type semiconductor substrate 53 and an n-type semiconductor region formed in the p-type semiconductor substrate 53. The p-type semiconductor substrate 53 is thinned, and the photoelectric conversion region of the PD 22 indicated by a broken line reaches the vicinity of the back surface of the p-type semiconductor substrate 53. The charges generated by the photoelectric conversion of the PD 22 are accumulated near the interface with the wiring layer 52. Each PD 22 is separated by a separation layer (not shown) (for example, a p + layer).

  The light control layer 54 is provided on the back surface side (incident surface side) of the p-type semiconductor substrate 53 and includes light control elements 55 </ b> A and 55 </ b> B and an insulating film 56. The light control elements 55A and 55B are elements that can adjust the amount of transmitted visible light (transmittance). For example, the light control elements 55A and 55B are almost transparent when a predetermined voltage is applied. When the voltage gradually becomes opaque and no voltage is applied (or the voltage is not applied by setting the electric potential of the dimming element 55A to ground or the like), the transmittance of the dimming elements 55A and 55B is almost 0%. (About several percent). The dimming elements 55A and 55B are provided on the hybrid pixels 42a and 42b so as to cover the PD 22 in a state where they can be controlled independently. When the light control elements 55A and 55B are transparent, the light incident through the microlens 57 and the color filter 41 passes through the light control elements 55A and 55B and reaches the PD 22. When the dimming elements 55A and 55B are opaque, the incident light is blocked by the dimming elements 55A and 55B, is at least dimmed, and enters the PD 22.

  The dimming elements 55A and 55B are provided symmetrically between the hybrid pixel 42a and the hybrid pixel 42b that form a pair. Specifically, the hybrid pixel 42a is provided with a dimming element 55A on the right half of the incident surface of the PD 22 and a dimming element 55B on the left half. On the other hand, in the hybrid pixel 42b, the dimming element 55A is provided in the left half of the incident surface of the PD 22, and the dimming element 55B is provided in the right half.

  The insulating film 56 is also a planarizing film that insulates the light control elements 55A and 55B from the p-type semiconductor substrate 53 and flattens the surfaces of the light control elements 55A and 55B. For this reason, the light control layer 54 on the normal pixel 43 is formed of only the insulating film 56. The insulating film 56 is made of, for example, a transparent and non-conductive material such as BPSG. The insulating film 56 is formed by integrating an insulating film below the light control elements 55A and 55B and a planarizing film between the light control elements 55A and 55B and on the light control elements 55A and 55B.

  The color filter 41 is provided on the light control layer 54 so that each color segment corresponds to the PD 22. The color arrangement of the color filter 41 is as described above, and the light beam condensed by the micro lens 57 passes through the color filter 41 and becomes one of R, G, and B and enters the PD 22.

  The microlens 57 is provided on the color filter 41 so as to correspond to each PD 22 and has a substantially hemispherical shape. The microlens 57 collects incident light on the corresponding PD 22. The center of the microlens 57 substantially coincides with the center of the PD 22 at the center of the light receiving region 11, but the closer to the periphery of the light receiving region 11, the closer to the center of the light receiving region 11 depending on the chief ray angle. It is arranged with offset (scaling). Thereby, in the peripheral portion of the light receiving region 11, the light beam incident on the PD 22 obliquely is also efficiently collected on the corresponding PD 22.

  As shown in FIG. 4, the solid-state imaging device 10 includes a first dimming control unit 61, a second dimming control unit 62, and wirings 63 and 64 that connect the dimming elements 55A and 55B, respectively. The voltage applied to the light control element 55A and the light control element 55B can be controlled independently.

  The first dimming control unit 61 is a circuit for applying a voltage to the dimming element 55A as needed via the wiring 63. The first dimming control unit 61 applies the dimming element 55A to the dimming element 55A based on a control signal input from the control unit 17. Adjust the timing of applying the voltage and the magnitude of the applied voltage. In the present embodiment, the first dimming control unit 61 keeps the voltage applied to the dimming element 55A at a predetermined voltage both when performing phase difference AF and when imaging a subject. For this reason, the light control element 55A is always in a transparent state and transmits almost all incident light.

  The second dimming control unit 62 is a circuit for applying a voltage to the dimming element 55B as needed via the wiring 64, and is input from the control unit 17 in the same manner as the first dimming control unit 62. The timing of applying a voltage to the light control element 55B and the magnitude of the applied voltage are adjusted based on the control signal. In the present embodiment, when performing the phase difference AF, the second dimming control unit 62 sets the voltage applied to the dimming element 55B to zero to make it opaque, and reduces the amount of light transmitted through the dimming element 55B. On the other hand, when the subject is imaged after performing the phase difference AF, the second dimming control unit 62 applies a predetermined voltage to the dimming element 55B to make it transparent, and transmits incident light to the dimming element 55B. Let

  The first dimming control unit 61 and the second dimming control unit 62 are formed by the wiring 52a of the wiring layer 52 and the like, similar to the control unit 17. On the other hand, the wirings 63 and 64 that connect the first dimming control unit 61 and the second dimming control unit 62 and the dimming elements 55A and 55B, respectively, are formed in the dimming layer 54 by, for example, polysilicon. Therefore, the first dimming control unit 61 and the second dimming control unit 62 are connected to the wirings 63 and 64 through via holes (also referred to as through holes, not shown) that penetrate the p-type semiconductor substrate 53.

  As shown in FIG. 5, the dimming element 55 </ b> A is, for example, an SPD (Suspended Particle Device, also referred to as a suspended particle device), and includes a pair of transparent electrodes 71 a and 71 b and an SPD layer 72. The transparent electrodes 71a and 71b are made of, for example, ITO (indium tin oxide). The transparent electrode 71a is a lower electrode provided on the p-type semiconductor substrate 53 side, and the transparent electrode 71b is an upper electrode provided on the color filter 41 and microlens 57 side. The first dimming control unit 61 adjusts the amount of light transmitted through the SPD layer 72 by applying a predetermined voltage between the transparent electrodes 71a and 71b.

  The SPD layer 72 is a layer in which microcapsules 72b are dispersedly arranged in a transparent resin 72a. The resin 72a is made of, for example, an ultraviolet curable resin, and oriented particles 72c (for example, liquid crystal molecules) are sealed in the microcapsule 72b. When a predetermined voltage is applied to the transparent electrodes 71a and 71b sandwiching the upper and lower sides of the SPD layer 72, as shown in FIG. 5A, the orientation directions of the oriented particles 72c are substantially aligned, and the incident light is dimmed. The light passes through the element 55A. On the other hand, as shown in FIG. 5B, when no voltage is applied to the transparent electrodes 71a and 71b, the orientation directions of the oriented particles 71a and 71b become almost random. For this reason, when a voltage is not applied between the transparent electrodes 71a and 71b, the incident light to the light control element 55A is scattered by the oriented particles 72c, so that the light control element 55A is in an opaque state. By adjusting the voltage applied to the transparent electrodes 71a and 71b between a zero voltage and a predetermined voltage, the degree of alignment of the alignment particles 72c in the alignment direction can be adjusted, and the opacity of the light control element 55A can be adjusted. . Although FIG. 5 shows the light control element 55A, the structure of the light control element 55B is the same as that of the light control element 55A.

  Hereinafter, the operation of the solid-state imaging device 10 configured as described above will be described. As shown in FIG. 6A, when a predetermined voltage is applied to the dimming elements 55A and 55B of the hybrid pixels 42a and 42b, the dimming elements 55A and 55B are transparent. For this reason, the incident light 71 to the hybrid pixels 42a and 42b passes through the light control elements 55A and 55B and reaches the PD 22. Therefore, when a predetermined voltage is applied to the dimming elements 55A and 55B, the hybrid pixels 42a and 42b function as the normal pixels 43.

  On the other hand, as shown in FIG. 6B, when a predetermined voltage is applied to the light control element 55A and the voltage applied to the light control element 55B is zero, the light control element 55A is transparent, but the light control element 55B becomes opaque. For this reason, a part of the incident light 71 to the hybrid pixels 42a and 42b is blocked by the opaque dimmer 55B, passes only through the transparent dimmer 55A, and reaches the PD 22. The light reaching the PD 22 is converted into electrons and output as a signal voltage as described above, and the signal voltage becomes a phase difference signal corresponding to the phase difference component of the incident light. Therefore, the hybrid pixel functions as a phase difference pixel by controlling the voltage applied to the light control element.

  In the case of the hybrid pixel 42a, the dimming element 55A is arranged on the right side (X direction positive side) and the dimming element 55B is arranged on the left side (X direction negative side). Selectively reaches PD22. Therefore, by making the light control element 55A transparent and the light control element 55B opaque, the hybrid pixel 42a functions as a phase difference pixel that selectively receives a light beam from the right side. On the other hand, in the case of the hybrid pixel 42b, the dimming element 55B is arranged on the right side and the dimming element 55A is arranged on the left side, so that the left half 72b of the incident light 71 is opposite to the hybrid pixel 42a. Selectively reaches PD22. Therefore, similarly to the hybrid pixel 42a, when the dimming element 55A is transparent and the dimming element 55B is opaque, the hybrid pixel 42b functions as a phase difference pixel that selectively receives a light beam from the left side.

  As described above, the solid-state imaging device 10 adjusts the hybrid pixels 42a and 42b when performing phase difference AF by independently adjusting the transmitted light amounts of the dimming elements 55A and 55B of the hybrid pixels 42a and 42b. The hybrid pixels 42 a and 42 b can function as the normal pixels 43 when functioning as phase difference pixels and imaging a subject after phase difference AF.

  In particular, in the solid-state imaging device 10, since the hybrid pixels 42a and 42b themselves function as normal pixels, the pixel values of the hybrid pixels 42a and 42b can be used as they are for generating a captured image when imaging a subject. . For this reason, it is necessary to correct the pixel value of the phase difference pixel so that it can be used for the captured image by interpolation, gain adjustment, etc., as in the case of imaging with a solid-state imaging device provided with a conventional phase difference pixel. There is no decrease in sensitivity and resolution of captured images. In addition, the solid-state imaging device 10 has no drive mode limitation due to the provision of the phase difference pixels and the like, and a driving mode for taking a moving image and adding pixels that cannot be conventionally used in the solid-state imaging device having the phase difference pixels. Can also be used with a solid imaging value of 10.

  In the first embodiment described above, the dimming elements 55A and 55B are provided so as to divide the hybrid pixels 42a and 42b in the left-right direction, and the phase difference AF is performed in the hybrid pixels 42a and 42b in the left-right direction parallax. However, the direction in which the light control elements 55A and 55B are provided is arbitrary. For example, by providing dimming elements 55A and 55B side by side in the vertical direction, information regarding vertical parallax may be obtained by the hybrid pixels 42a and 42b when performing phase difference AF. Further, by providing the dimming elements 55A and 55B side by side in the oblique 45 degree direction (or 135 degree direction), when performing phase difference AF, the hybrid pixels 42a and 42b may obtain information on the oblique parallax. good.

  In the first embodiment described above, the voltage applied to the dimming element 55B is made zero by applying a predetermined voltage to the dimming element 55A to make it opaque, and the hybrid pixel 42a selects the light flux from the right side. In contrast, the hybrid pixel 42b functions as a phase difference pixel that selectively receives a light flux from the left side, but conversely, the voltage applied to the dimming element 55A is set to zero. It may be made opaque, and a predetermined voltage may be applied to the light control element 55B to maintain transparency. In this case, the hybrid pixel 42a functions as a phase difference pixel that selectively receives a light beam from the left side, and the hybrid pixel 42b functions as a phase difference pixel that selectively receives a light beam from the right side. As described above, when the hybrid pixels 42a and 42b are phase difference pixels that selectively receive light beams from the left side and the right side, respectively, in the same manner as described above based on the signals obtained from the hybrid pixels 42a and 42b. Phase difference AF can be performed.

  In the first embodiment described above, when the hybrid pixels 42a and 42b function as phase difference pixels, a predetermined voltage is applied to the dimming element 55A to make it transparent, and the dimming element 55B to zero voltage to make it opaque. However, if the voltage applied to the light control element 55B is lower than that of the light control element 55A, the voltage applied to the light control elements 55A and 55B is arbitrary. For example, a voltage lower than a predetermined voltage is applied to the light control element 55A, and the light control element 55B is not set to zero voltage, but may be applied within a range lower than the voltage applied to the light control element 55A. good. That is, if the amount of transmitted light is different between the dimming element 55A and the dimming element 55B, information on the left and right parallax can be obtained from these, so the magnitude of the voltage applied to the dimming elements 55A and 55b itself is It can be set arbitrarily. However, as in the first embodiment, a predetermined voltage is applied to the light control element 55A so that the transmitted light amount is maximized, and the light control element 55B is set to zero voltage so that the transmitted light amount is minimized. In this case, parallax information is most accurate. For this reason, when making the hybrid pixels 42a and 42b function as phase difference pixels, it is preferable to apply a predetermined voltage to the dimmer 55A and set the dimmer 55B to zero voltage.

  In the first embodiment described above, in the 6 × 6 pixel unit of the color filter 41, the lower left G pixel of the upper right first subunit 41a and the lower left G pixel of the upper left second subunit 41b are combined into the hybrid pixel 42a. , 42b, but the position of the G pixel used as the hybrid pixel 42a, 42b is arbitrary. For example, the center G pixel of the upper right first subunit 41a and the center G pixel of the upper left second subunit 41b may be hybrid pixels 42a and 42b, respectively.

[Second Embodiment]
In the first embodiment described above, the two dimming elements 55A and 55B are provided in the hybrid pixels 42a and 42b. However, like the solid-state imaging device 80 shown in FIG. Four controllable light control elements may be provided.

  The solid-state imaging device 80 includes four hybrid pixels 82a, 82b, 82c, and 82d in the 6 × 6 pixel unit of the color filter 41.

  The hybrid pixel 82a is the lower left G pixel of the upper right first subunit 41a, and the four dimming elements 85A, 85B, 85C, and 85D each adjust the amount of transmitted light. In the case of the hybrid pixel 82a, the dimming elements 85A, 85B, 85C, and 85D are provided at the upper right, upper left, lower left, and lower right, respectively.

  The hybrid pixel 82b is the lower left G pixel of the upper left second subunit 41b, and has four dimming elements 85A, 85B, 85C, and 85D, similar to the hybrid pixel 82a. However, the hybrid pixel 82b is different from the hybrid pixel 82a in the arrangement of the dimming elements 85A to 85D, and is arranged symmetrically in the horizontal direction with respect to the hybrid pixel 82a. That is, in the hybrid pixel 82b, the dimming element 85A is arranged at the upper left, the dimming element 85B is arranged at the upper right, the dimming element 85C is arranged at the lower right, and the dimming element 85D is arranged at the lower left.

  The hybrid pixel 82c is the lower left G pixel of the lower left first subunit 41a, and the hybrid pixel 82d is the lower left G pixel of the lower right second subunit 41b. And these also have four light control elements 85A-85D like the hybrid pixels 82a and 82b. However, these sequences are different.

  The dimming elements 85A to 85D of the hybrid pixel 82c are arranged at the lower left, lower right, upper right, and upper left, respectively. That is, the hybrid pixel 82c is vertically symmetrical with the hybrid pixel 82b in which the arrangement of the light control elements 85A to 85D is above the same column. Further, as compared with that of the hybrid pixel 82a, the arrangement of the dimming elements 85A to 85D of the hybrid pixel 82c is symmetrical in the oblique direction (about 135 degrees direction).

  The dimming elements 85A to 85D of the hybrid pixel 82d are arranged at the lower right, the lower left, the upper left, and the upper right, respectively, and are arranged vertically symmetrically with respect to that of the hybrid pixel 82a in the upper row. In addition, as compared with that of the hybrid pixel 82b, the arrangement of the dimming elements 85A to 85D of the hybrid pixel 82d is a symmetrical arrangement in the oblique direction (about 45 degrees direction).

  These hybrid pixels 82a to 82d are connected to each pixel through wirings 89a, 89b, 89c, and 89d, such as a first dimming control unit 91, a second dimming control unit 92, and a third dimming control unit. 93 and the fourth dimming control unit 94, respectively. The wirings 89a to 89d are provided in the light control layer 54 using polysilicon or the like as in the first embodiment. The first to fourth dimming control units 91 to 94 are circuits for independently applying voltages to the dimming elements 85A to 85D through the wirings 89a to 89d as necessary. The first to fourth dimming control units 91 to 94 are formed by the wiring 52a and the like of the wiring layer 52 as in the first embodiment, and are connected to the wirings 89a to 89d through via holes (not shown). The first to fourth dimming control units 91 to 94 adjust the timing of applying a voltage to the dimming elements 85A to 85D and the magnitude of the applied voltage based on the control signal input from the control unit 17, respectively. To do.

  As shown in FIG. 8A, in the solid-state imaging device 80 configured as described above, when all of the dimming elements 85A to 85D are applied with a predetermined voltage, since the dimming elements 85A to 85D are all transparent, The pixels 82 a to 82 d function as the normal pixel 43. For this reason, by applying a predetermined voltage to the dimming elements 85A to 85D and imaging the subject, the pixel values of the hybrid pixels 82a to 82d can be used as they are without gain correction or the like to generate an image of the subject. .

  On the other hand, when a predetermined voltage is applied to the dimming element 85A and the dimming element 85D to make it transparent, and the dimming element 85B and the dimming element 85C are set to zero voltage to be opaque, the region through which the incident light 71 can be transmitted is dimmed. It is limited to the regions 96R and 96L corresponding to the elements 85A and 85D. That is, when viewed in units of 6 × 6 pixels of the color filter 41, the two hybrid pixels 82a and 82d on the right side limit the region through which the incident light 71 can be transmitted to the right half region 96R. Functions as a phase difference pixel that selectively receives a light beam incident from the right side and performs photoelectric conversion. On the other hand, in the two hybrid pixels 82b and 82c on the left side, the region through which the incident light 71 can be transmitted is limited to the left half region 96L. Functions as a phase difference pixel.

  Accordingly, by applying a predetermined voltage to the dimming element 85A and the dimming element 85D to make it transparent, and making the dimming element 85B and the dimming element 85C to be zero voltage and making them opaque, the solid-state imaging device 80 can make a right and left parallax. Based on the phase difference AF. The focus evaluation is performed based on, for example, signals obtained from a pair of the hybrid pixel 82a and the hybrid pixel 82b or a pair of the hybrid pixel 82d and the hybrid pixel 82c.

  As shown in FIG. 9, when a predetermined voltage is applied to the dimming element 85A and the dimming element 85B to make it transparent, and the dimming element 85C and the dimming element 85D are made zero voltage to be opaque, the incident light 71 can be transmitted. This area is limited to the areas 96U and 96D corresponding to the light control elements 85A and 85B. That is, when viewed in units of 6 × 6 pixels of the color filter 41, the two hybrid pixels 82a and 82b on the upper side limit the region through which the incident light 71 can be transmitted to the upper half region 96U. Functions as a phase difference pixel that selectively receives and photoelectrically converts a light beam incident from above. On the other hand, in the two hybrid pixels 82c and 82d on the lower side, the range through which the incident light 71 can be transmitted is limited to the lower half region 96D, so that these selectively receive the light beam incident from the lower side. And function as a phase difference pixel for photoelectric conversion.

  Therefore, by applying a predetermined voltage to the dimming element 85A and the dimming element 85B to make it transparent, and making the dimming element 85C and the dimming element 85D zero voltage to make them opaque, the solid-state imaging device 80 can change the upper and lower parallaxes. Based phase difference AF can be performed. Also, for example, in the case of a subject that changes in the vertical direction but has little change in the horizontal direction, such as the horizon, the phase difference AF cannot be accurately performed based on the left and right parallax as described above. In this case, the solid-state imaging device 80 can easily switch to the phase difference AF based on the vertical parallax by changing the pattern of the voltage applied to the transparent electrodes 85A to 85D.

  In addition, in a conventional solid-state imaging device that creates asymmetry by using a light shielding film or the like, in the case where both information about the parallax in the horizontal direction and the vertical direction can be obtained, a phase difference pixel that detects parallax in the horizontal direction; Phase difference pixels for detecting the parallax in the vertical direction must be arranged in advance. For this reason, since it is necessary to provide in advance a phase difference pixel that is twice that in the case of obtaining only the parallax in one direction in the left-right direction or the up-down direction, the number of pixels that have to be subjected to gain correction or interpolation is doubled. Deterioration of sensitivity and resolution tends to become more remarkable. Compared with this conventional solid-state imaging device, the solid-state imaging device 80 can arbitrarily obtain information on the suggestion in the left-right direction and the up-down direction, and the hybrid pixels 82a to 82d are normal pixels when imaging a subject. Therefore, it is possible to obtain a captured image with particularly little deterioration in sensitivity and resolution.

  When phase difference AF is performed based on vertical parallax, the focus evaluation is performed based on, for example, a signal obtained from a pair of hybrid pixel 82a and hybrid pixel 82d or a pair of hybrid pixel 82b and hybrid pixel 82c. Done.

  As shown in FIG. 10, in the solid-state imaging device 80, the incident light 71 can be transmitted by applying a predetermined voltage to the dimming element 85A to make it transparent and making the dimming elements 85B to 85D to be zero voltage to be opaque. This area can be limited to the areas 96UR, 96UL, 96DL, and 96DR corresponding to the dimming element 85A. In this case, the hybrid pixels 82a to 82d function as phase difference pixels that selectively receive light beams incident from the upper right, upper left, lower left, and lower right directions and perform photoelectric conversion. For this reason, for example, by performing phase difference AF based on a signal obtained from a pair of hybrid pixel 82a and hybrid pixel 82c or a pair of hybrid pixel 82b and hybrid pixel 82d, phase difference AF based on diagonal parallax is obtained. It can be carried out.

  In FIG. 10, only the dimming element 85A is made transparent, but as shown in FIG. 11, a predetermined voltage is applied to the dimming elements 85A, 85B, and 85D to make it transparent, and only the dimming element 85C is set to zero voltage. And may be opaque. In this case, the area where each of the hybrid pixels 82a to 82d can transmit the incident light 71 is confined to L-shaped areas 97UR, 97UL, 97DL, and 97DR. Also in this case, as described above, the phase difference AF is performed based on the signals obtained from the pair of the hybrid pixel 82a and the hybrid pixel 82c or the pair of the hybrid pixel 82b and the hybrid pixel 82d. Phase difference AF can be performed. Further, for example, the dimming element 85A is made transparent by applying a predetermined voltage, the dimming element 85C is made zero by making it opaque, and an intermediate voltage is applied to the dimming elements 85B and 85D to make it semitransparent. The same applies to such cases.

[Third Embodiment]
In the first and second embodiments described above, the influence of the arrangement of the hybrid pixels in the light receiving region 11 is not mentioned, but the chief ray angle generally increases from the central portion to the peripheral portion of the light receiving region 11. The light is obliquely incident on the peripheral pixels. For this reason, as described above, the arrangement of the microlenses 57 is scaled so that obliquely incident light can be efficiently condensed on the PD 22. More specifically, the microlens 57 is scaled based on the incident efficiency of light to the PD 22 of the normal pixel 43 having a symmetric structure.

  On the other hand, a hybrid pixel in the case of functioning as a phase difference pixel is a pixel having a substantially asymmetric structure because a region through which incident light 71 can be transmitted is limited. However, a pixel having such an asymmetric structure is The degree of decrease in sensitivity due to oblique incidence of light is larger than that of the normal pixel 43. For this reason, simply scaling the microlens 57 in accordance with the normal pixel 43 reduces the sensitivity of the hybrid pixel when functioning as a phase difference pixel. Therefore, the hybrid pixel (phase difference in the periphery of the light receiving region 11) is reduced. If phase difference AF is performed using (pixel), accuracy may be lost.

  Specifically, as shown in FIG. 12, when the hybrid pixel 42a is compared between the central portion and the peripheral portion of the light receiving area 11, the chief ray angle is small in the central hybrid pixel 42aC, and light is incident almost from the front. For this reason, when the light control element 55B is made opaque, for example, about 2/3 of the right light beam 101R and about 1/3 of the left light beam 101L of the incident light pass through the transparent light control element 55A. That is, as a whole, about 1/2 of the incident light is incident on the PD 22.

  Further, in the hybrid pixels 42aR and 42aL in the periphery of the light receiving region 11, the chief ray angle is large and light is incident obliquely. For this reason, when the dimming element 55B is made opaque, about 2/3 of the right beam 101R and about 2/3 of the left beam 101L are modulated in the hybrid pixel 42aR on the right side (X direction positive side) peripheral portion. Since the light passes through the optical element 55A, about 2/3 of the incident light is incident on the PD 22 as a whole. Further, in the hybrid pixel 42aL on the left side (X direction negative side) peripheral portion, about 1/3 of the right light beam 101R and about 1/3 of the left light beam 101L pass through the dimming element 55A. In this case, about 1/3 of the incident light is incident on the PD 22.

  Moreover, as shown in FIG. 13, among the hybrid pixels 42b in which the arrangement of the dimming elements 55A and 55B is reversed, the dimming element 55B of the hybrid pixel 42bC in the center of the light receiving region 11 becomes opaque. . About half of the incident light is incident on the PD 22. However, in the hybrid pixel 42bR in the right peripheral part of the light receiving region 11, about 1/3 of the incident light enters the PD 22, and in the hybrid pixel 42bL in the left peripheral part of the light receiving region 11, about 2 / of the incident light. 3 enters the PD 22. In FIGS. 12 and 13, the microlens 57 is not scaled for simplicity, but this tendency is the same when the microlens 57 is scaled.

  Therefore, as shown in FIG. 14, in the periphery of the light receiving region 11, the sensitivity of the hybrid pixels 42 a and 42 b with respect to the normal pixel 43 is unbalanced. In particular, when functioning as a phase difference pixel, if the sensitivity of the hybrid pixels 42a and 42b is low in the peripheral portion of the light receiving region 11, it becomes difficult to obtain a signal sufficient for the calculation of the phase difference AF. However, there are cases where phase difference AF cannot be performed.

Therefore, as shown in FIG. 15, it is preferable to adjust the respective areas according to the positions of the hybrid pixels 42a and 42b by changing the widths (lengths in the X direction) of the light control elements 55A and 55B. Specifically, the hybrid pixel 42aR disposed on the right side of the light-receiving region 11, extending the width _{W 1} even dimming element 55B than the width _{W 0} of the light control device 55B of hybrid pixel 42aC of the central portion _(W 1> W ₀ ), the width W ₂ of the dimming element 55A is made smaller than the width W ₀ of the dimming element 55A of the center hybrid pixel 42aC (W ₂ <W ₀ ). In the hybrid pixel 42bR disposed on the right side of the light-receiving region 11, the width of the light control device 55A to _{W 1,} the width of the light control device 55A to _{W 2.} The reverse is true for the hybrid pixels 42aL and 42bL on the left periphery. In this way, the sensitivity of the hybrid pixels 42a and 42b is made uniform as shown in FIG. 16, and accurate phase difference AF can be performed even when the peripheral hybrid pixels 42a and 42b are used.

  Here, the hybrid pixels 42a and 42b of the first embodiment are taken as an example. However, when four transparent electrodes 85A to 85D are used like the hybrid pixels 82a to 82d of the second embodiment, they are shown in FIG. Thus, what is necessary is just to adjust the area of light control element 85A-85D. In FIG. 17, only the dimming element 85A is provided with a reference numeral, but the positions of the other dimming elements 85B to 85D of the hybrid pixels 82a to 82d are as described above (see FIG. 7).

  In the first to third embodiments described above, the G pixel is a hybrid pixel. However, as shown in FIG. 18, the R pixel or the B pixel may be a hybrid pixel. Of course, all the hybrid pixels may be formed as R pixels, or all the hybrid pixels may be formed as B pixels. Further, only two of R, G, and B may be provided with hybrid pixels.

  However, for example, when the hybrid pixels 42a and 42b are formed for two or more colors of R, G, and B, the amount of decrease in sensitivity in the peripheral portion 11b differs for each color, so the areas of the light control elements 55A and 55B for each color It is preferable to change the enlargement ratio (reduction ratio). If the area enlargement ratio of the light control elements 55A and 55B is changed for each color, the sensitivity of the hybrid pixel of each color can be made uniform with respect to the sensitivity of the normal pixel 43, so it is necessary to perform shading correction or the like for each color. Therefore, accurate and accurate phase difference AF can be performed. The enlargement ratio (reduction ratio) of the light control elements 55A and 55B for each color is determined according to the characteristics of the color filter.

  In the first to third embodiments described above, some of the pixels in the light receiving region 11 are hybrid pixels. However, since the hybrid pixel also functions as a normal pixel, as shown in FIG. All the pixels in 11 may be hybrid pixels. In a solid-state imaging device including a conventional phase difference pixel, since the phase difference pixel is provided in a predetermined position in advance, the subject in an area that does not include the phase difference pixel cannot be focused by the phase difference AF. However, if all the pixels in the light receiving region 11 are hybrid pixels as described above, it is possible to focus on a subject at an arbitrary position with the phase difference AF.

  In FIG. 19, the hybrid pixel 42a in which the dimming element 55A is arranged on the right side and the hybrid pixel 42b in which the dimming element 55A is arranged on the left side are mixed, but as shown in FIG. The pixel may be a hybrid pixel 42a in which the dimming element 55A is arranged on the right side. In this case, for example, in a certain frame, a predetermined voltage is applied to the dimming element 55A to make it transparent, and the dimming element 55B is made to be zero voltage to make it opaque, and the light beam from the right side is selectively received by all the pixels for imaging. In the next frame, the dimming element 55A is set to zero voltage to make it opaque, a predetermined voltage is applied to the dimming element 55B to make it transparent, and light is picked up from all the pixels selectively from the left side for imaging. . In this way, the phase difference AF can be performed using the pixel value obtained in the first frame and the pixel value obtained in the second frame. If the image obtained in the first frame is the right-eye image and the image obtained in the second frame is the left-eye image, a stereoscopic image can be obtained. That is, 2D shooting (when all pixels function as normal pixels) and 3D shooting can be easily switched, and the resolution of the 2D shot image and the resolution of the 3D shot image are high-resolution images using all pixels. is there.

  The light control elements 55A and 55B are transparent when a predetermined voltage is applied, but the transparency may not be a transparency that transmits 100% of visible light. For this reason, if dimming elements 55A and 55B are provided only in the hybrid pixels 42a and 42b, even if the dimming elements 55A and 55B are made transparent so that the hybrid pixels 42a and 42b function as normal pixels, The sensitivity may be lower than that of the normal pixel 43 that is not provided. For this reason, dimming elements 55A and 55B may be provided in all the pixels, and the sensitivity of all the pixels when functioning as a normal pixel may be made uniform.

  In the first to third embodiments described above, one PD 22 is provided for one pixel. However, as shown in FIG. 21, the color segments of the color filter 41 and the microlens 75 are provided for the two PDs 22. One pixel may include two PDs 22, and each pixel may output a voltage signal proportional to the total value of the charges generated by the two PDs 22. In this case, in the hybrid pixel 142a, the dimming element 55A is provided on one PD22, and the dimming element 55B is provided on the other PD22. If a predetermined voltage is applied to the dimming element 55A to make it transparent, and the dimming element 55B is made zero voltage to make it opaque, the hybrid pixel 142a can function as a phase difference pixel. When using four dimming elements 85A to 85D as in the second embodiment, four PDs 22 may be included in one pixel. As described above, when a plurality of PDs 22 are included in one pixel, since each PD 22 is independent, there is no movement of charges between the PDs 22 and there is little noise, so that more accurate phase difference AF can be performed. it can.

  Thus, when a plurality of PDs 22 are included in one pixel, as shown in FIG. 22, element isolation regions 151 such as between pixels are formed by p +, and further, element isolation regions 152 between PDs 22 are formed. Is preferably formed with p ++ with more additives. By doing so, the potential of the element isolation region 152 becomes higher than that of the normal element isolation region 151, so that the movement of charges between the two PDs 22 can be more reliably prevented. Similarly, as shown in FIG. 22, it is further preferable to provide an FD 26 and a read gate 153 for each PD 22 so as to read charges for each PD 22 without mixing the charges of the two PDs 22.

  22 exemplifies a case where two PDs 22 are provided for one pixel. However, as in the first to third embodiments, even when one PD 22 is provided for one normal pixel 43, a hybrid pixel is used. As described above, the PD 22 may be divided into two as described above.

  In the first to third embodiments described above, the pixels of the solid-state imaging device 10 are arranged in a square, but the arrangement of the pixels is arbitrary. For example, the pixel array may be a honeycomb array as shown in FIGS. Further, the color arrangement of the color filter when the pixel arrangement is the honeycomb arrangement is arbitrary, but for example, the color arrangement shown in FIGS. 23 and 24 can be used. The color arrangement in FIG. 23 is an example in which columns of G pixels and columns in which two R pixels and two B pixels are alternately arranged are alternately arranged in a 45-degree oblique direction.

  In the first to third embodiments described above, when the pixels of the solid-state imaging device 10 are arranged in a square, the color filter 41 using 6 × 6 pixels as a unit is used. Is optional. For example, when the pixels are arranged in a square array, as shown in FIG. 25, a so-called Bayer array in which 2 × 2 pixels surrounded by a broken line are arranged vertically and horizontally may be used.

  In the first to third embodiments described above, the hybrid pixels that are paired when functioning as the phase difference pixels are provided in the same row (the same row and the same column in the second embodiment). These two hybrid pixels may be provided in different rows (columns).

  In the above-described embodiment, the three transistors Tr23, Ta24, and Ts25 form a pixel (see FIG. 1), but the pixel 21 (the normal pixel 43 and the phase difference pixels 42a and 42b) receives incident light respectively. The number of transistors and the like are arbitrary as long as photoelectric conversion can be performed and signals necessary for formation of a captured image and phase difference AF can be output. For example, the pixel 21 may be configured by providing a transistor for transfer between the PD 22 and the FD and using four transistors.

  In the first to third embodiments described above, the light control layer 54 is provided on the back surface of the p-type semiconductor substrate 53. However, if the distribution of the transmitted light amount can be formed by the light control element, the light control layer 54 is provided. The order of stacking is arbitrary. For example, a light control layer 54 may be provided between the color filter 41 and the microlens 57. However, since the controllability is better as the distance between the light control element and the PD 22 is shorter, the light control layer 54 is preferably provided directly on the p-type semiconductor substrate 53 as in the first to third embodiments described above. .

  In the first to third embodiments described above, the solid-state imaging device 10 is a CMOS image sensor, but the solid-state imaging device 10 may be a CCD image sensor. In the first to third embodiments described above, the solid-state imaging device 10 is a backside illumination type. However, as shown in FIG. 26, the solid-state imaging device 10 is a front side illuminated (FSI) type image sensor. But you can. The front-illuminated image sensor has a structure in which a wiring layer 52, a color filter 41, and a microlens 57 are laminated in this order on a p-type semiconductor substrate 53, and light enters the PD 22 from the surface side through the wiring layer 52. This type of image sensor. In the case of the surface irradiation type, the light control layer 54 may be provided, for example, between the p-type semiconductor substrate 53 and the wiring layer 52 (on the surface of the p-type semiconductor substrate 53). However, if the distribution of the amount of transmitted light can be formed by the light control elements 55A and 55B of the light control layer 54, the light control layer 54 is under the PD 22 (on the back side of the p-type semiconductor substrate 53) even in the case of the front surface irradiation type. May be provided.

  In the first to third embodiments described above, the SPD is used as the dimming element. However, a device other than the SPD may be used instead as long as the amount of transmitted light can be adjusted. For example, the electrochromic element 170 shown in FIG. 27 can be used as a light control element instead of the SPD. The electrochromic element 170 is an element whose color is changed by controlling the applied voltage. For example, a lower electrode 171, an oxidation coloring layer 172, an ion conductive layer 173, a reduction coloring layer 174, and an upper electrode 175 are laminated in this order. Formed.

The lower electrode 171 and the upper electrode 175 are both transparent electrodes, and are made of, for example, indium tin oxide (ITO). SnO ₂ or In ₂ O may be used for the lower electrode 171 and the upper electrode 175.

  The oxidation coloring layer 172 is originally transparent, but is a layer that develops color when oxidized. An oxide or hydroxide such as iridium, nickel, chromium, vanadium, ruthenium, or rhodium can be used for the oxidation coloring layer 172. When iridium oxide is used for the oxidation coloring layer 172, it becomes gray when colored.

The ion conductive layer 173 is an insulating film, but is a good conductor (or supplier) of ions involved in the coloring of the oxidation coloring layer 172 and the reduction coloring layer 174. The ion conductive layer 173 is made of, for example, tantalum pentoxide (Ta ₂ O ₅ ). For the ion conductive layer 173, an oxide such as silicon, titanium, aluminum, niobium, zirconium, hafnium, lanthanum, or magnesium fluoride can be used.

The reduced color layer 174 is originally transparent, but is a layer that develops color when reduced, and is made of, for example, tungsten oxide (WO ₃ ). Molybdenum oxide (MoO ₃ ) can also be used. When tungsten oxide is used for the reduction coloring layer 174, it becomes blue when colored.

  In the electrochromic element 170 configured as described above, when a voltage is applied to the lower electrode 171 and the upper electrode 175, the oxidation coloring layer 172 and the reduction coloring layer 174 are oxidized, reduced, colored, and opaque.

  As shown in FIGS. 28A and 28B, a composite device 180 of a polymer and a liquid crystal can be used as a light control element. The polymer liquid crystal composite device 180 is formed, for example, by disposing a polymer / liquid crystal composite layer 182 between two transparent electrodes 181a and 181b facing each other. The transparent electrodes 181a and 181b are made of, for example, ITO. The composite layer 182 is formed by filling liquid crystal molecules 184 almost continuously in the gaps of the network polymer network 183.

  Then, as shown in FIG. 28A, when a voltage is applied between the transparent electrodes 181a and 181b, the alignment directions of the liquid crystal molecules 184 are aligned, and a transparent state in which the incident light 71 is transmitted is obtained. On the other hand, as shown in FIG. 28 (B), when no voltage is applied between the transparent electrodes 181a and 18b, the liquid crystal molecules 184 are oriented almost randomly according to the surrounding polymer network 183, and the incident light 71 is changed. Scattered. That is, when no voltage is applied between the transparent electrodes 181a and 18b, the polymer liquid crystal composite device 180 becomes opaque.

  In FIG. 28, so-called PNLC (Polymer Network Liquid Crystal) is shown, but so-called PDLC (Polymer Dispersed Liquid Crystal) in which liquid crystal molecules 184 are discontinuously distributed in the polymer may be used as the light control element. it can.

  Furthermore, as shown in FIG. 29, a liquid crystal device 190 having substantially the same configuration as a popular liquid crystal display device may be used as a light control element instead of the SPD. The liquid crystal device 190 is formed by laminating a polarizing plate 191, a lower electrode 192, an alignment film 193, a liquid crystal layer 194, an alignment film 195, an upper electrode 196, a polarizing plate 197, and the like in this order. The polarizing plate 191 and the polarizing plate 197 are so-called crossed Nicols so that the transmission axes are perpendicular to each other. The alignment films 193 and 195 align the alignment direction of the liquid crystal molecules 194a of the liquid crystal layer 194 in a substantially vertical direction when no voltage is applied between the lower electrode 192 and the upper electrode 196. The lower electrode 192 and the upper electrode 196 are transparent electrodes, and are formed of, for example, ITO.

  In the liquid crystal device 190 configured as described above, when a voltage is applied between a pair of electrodes 192 and 196 to rotate the liquid crystal molecules 194a and align it in the horizontal direction as shown in FIG. Incident light that has passed through the plate 197 has a polarization plane that passes through the polarizing plate 191 due to the birefringence of the liquid crystal molecules 194a. For this reason, when a voltage is applied between the electrodes 192 and 196, the liquid crystal device 190 becomes transparent. On the other hand, as shown in FIG. 29B, when no voltage is applied between the electrodes 192 and 196, the liquid crystal molecules 194a are aligned in a substantially vertical direction by the alignment films 193 and 195. For this reason, the incident light transmitted through the polarizing plate 197 does not receive the birefringence action of the liquid crystal molecules 194a, and therefore cannot pass through the opposing polarizing plate 191. That is, when no voltage is applied between the electrodes 192 and 196, the liquid crystal device 190 is in an opaque state.

  Note that FIG. 29 shows a so-called VA type liquid crystal, but other types of liquid crystal such as a TN type, an IPS type, an MVA type, and an OCB type may be used.

  In the first to third embodiments described above, the primary color filter 41 is used, but a complementary color filter may be used. The color filter may include a colorless (transparent) color segment, and a hybrid pixel may be formed in the colorless pixel.

  In the first embodiment described above, hybrid pixels 42a and 42b that function as phase difference pixels for obtaining the parallax in the horizontal direction are provided, and in the second embodiment, functions as phase difference pixels for obtaining the parallax in the vertical and horizontal directions and in the oblique direction. Although the hybrid pixels 82a to 82d are provided, when obtaining vertical and horizontal parallaxes, the hybrid pixels 42a and 42b of the first embodiment and the hybrid pixels 42a and 42b are rotated by 90 degrees and the vertical parallax is obtained. Hybrid pixels that can be obtained may be arbitrarily arranged.

  In the first to third embodiments described above, the same dimming elements of a plurality of hybrid pixels are uniformly the same, such as applying a predetermined voltage to the dimming elements 55A of all the hybrid pixels 42a and 42b to make them transparent. Although a voltage is applied, the voltage applied to the dimming elements of all the hybrid pixels may be controlled for each pixel. In particular, when all the pixels are hybrid pixels (see FIGS. 19 and 20), switching between 2D shooting and 3D shooting is easy if the transparent electrode of each pixel can be individually controlled.

  In the first to third embodiments described above, two or more dimming elements are provided in one hybrid pixel. However, one electrode may be provided in the hybrid pixel. For example, in the case of the hybrid pixels 42a and 42b of the first embodiment, the dimming element 55A that is always kept transparent may be eliminated, and only the dimming element 55B may be provided. In this case, if the dimming element 55B is set to zero voltage to make it opaque, the incident light 71 can be transmitted only through the portion where the dimming device is not provided (the portion where the dimming device 55A is provided in the first embodiment). . For this reason, when it is determined in advance that the hybrid pixels 42a and 42b function as a right light beam phase difference pixel and a left light beam phase difference pixel, respectively, by setting the dimming element 55B to zero and making it opaque. As described above, it is possible to provide only a light control element that is controlled to be opaque. The hybrid pixels 42a and 42b function as a right beam phase difference pixel and a left light beam phase difference pixel according to the subject or shooting conditions, and conversely, the hybrid pixels 42a and 42b respectively function as a left light beam phase difference pixel and a right beam phase difference pixel. When functioning as a light beam phase difference pixel, it is preferable to provide two or more light control elements as in the first to third embodiments.

  In the first to third embodiments described above, a plurality of hybrid pixels are provided uniformly in the light receiving region 11, but at least two hybrid pixels are sufficient for performing phase difference AF. When three or more hybrid pixels are provided, the number of hybrid pixels and the distribution of the arrangement are arbitrary.

  The solid-state imaging device 10 of the present invention can be used for any imaging device that performs phase difference AF. In particular, the area of the dimming element 55A of the third embodiment enlarged at the peripheral portion 11b is the principal ray angle at the peripheral portion, such as a thin digital camera, a camera unit mounted on a mobile phone, PDA, smart photon, etc. It is particularly suitable for thin and small products that tend to be large.

DESCRIPTION OF SYMBOLS 10,80 Solid-state imaging device 11 Light-receiving area 21 Pixel 42a, 42b, 82a-82d Hybrid pixel 43 Normal pixel 54 Light control layer 55A, 55B, 85A-85D Light control element

Claims (9)

A first dimming element that covers a part of an incident surface of a photodiode that photoelectrically converts incident light from a subject and changes a transmitted light amount according to an applied voltage, and a remaining portion of the incident surface of the photodiode A first hybrid pixel having a second dimming element that covers and changes the amount of transmitted light according to the applied voltage;
A second hybrid pixel in which positions of the first dimming element and the second dimming element are interchanged with respect to the first hybrid pixel;
By applying a voltage to at least one of the first dimming element and the second dimming element, making one of the first dimming element or the second dimming element transparent and making the other opaque, A dimming control unit that causes the first hybrid pixel and the second hybrid pixel to function as a pair of phase difference pixels;
With
The first hybrid pixel and the second hybrid pixel are provided at least in the central part and the peripheral part of a light receiving region in which a plurality of pixels are arranged,
The first hybrid pixel disposed in the peripheral portion in a predetermined first direction with respect to the central portion has an area of the first dimming element larger than an area of the second dimming element. On the other hand, in the second hybrid pixel disposed in the peripheral portion in the first direction, the area of the first dimming element is smaller than the area of the second dimming element . The first dimming element and the second dimming element include a pair of transparent electrodes and a layer provided between the pair of transparent electrodes and in which oriented particles are dispersedly arranged in a transparent predetermined substrate. ,
The dimming control unit adjusts the amount of transmitted light of the first dimming element and the second dimming element by adjusting the voltage applied between the pair of transparent electrodes to change the orientation direction of the oriented particles. The solid-state imaging device according to claim 1. The first dimming element and the second dimming element each include a first transparent electrode, an oxidized color layer that is colored by being oxidized, and insulating ions that pass ions related to the coloring of the oxidized color layer. An electrochromic element formed by laminating a conductive layer, a reduced coloring layer colored by being reduced through the ion conductive layer, and a second transparent electrode in this order;
The dimming control unit adjusts the voltage applied between the first transparent electrode and the second transparent electrode to control the coloring of the oxidation coloring layer and the reduction coloring layer, The solid-state imaging device according to claim 1, wherein a transmitted light amount of the first dimming element and the second dimming element is adjusted. The first dimming element and the second dimming element include a pair of transparent electrodes and a polymer liquid crystal composite layer having a structure in which liquid crystal molecules are filled in a gap between polymer networks,
The dimming control unit adjusts the amount of light transmitted through the first dimming element and the second dimming element by adjusting a voltage applied between the pair of transparent electrodes to change the alignment direction of the liquid crystal molecules. The solid-state imaging device according to claim 1. The first dimming element and the second dimming element include a first polarizing plate, a first transparent electrode, a liquid crystal layer made of liquid crystal molecules, a second transparent electrode, and a second polarizing plate, With
The dimming control unit adjusts a voltage applied to the first transparent electrode and the second transparent electrode, and controls an alignment direction of the liquid crystal molecules, thereby controlling the first dimming element and the second dimming element. The solid-state imaging device according to claim 1, wherein the amount of light transmitted through the light control element is adjusted. The first hybrid pixel disposed in a peripheral portion in a direction opposite to the first direction with respect to the central portion has an area of the first dimming element smaller than an area of the second dimming element, wherein said second hybrid pixels arranged in the peripheral portion of the direction of the first direction and the reverse is greater claim than the area of the area of the first light adjusting device said second light adjusting device with respect to the central portion The solid-state imaging device according to any one of 1 to 5 .   The solid-state imaging device according to claim 1, wherein the first hybrid pixel and the second hybrid pixel are included in at least two colors of red, green, and blue. The solid-state imaging device according to any one of claims 1 to 7 All the pixels in the light receiving region in which a plurality of pixels are arranged is the first hybrid pixel or the second hybrid pixel. The first hybrid pixel and the second hybrid pixel have a photoelectric conversion region divided into a plurality of element isolation regions as the photodiode,
The first dimming element and the second dimming element are provided corresponding to the photoelectric conversion regions,
One wherein the said isolation region included in the first hybrid pixel or the second hybrid pixel, either high amount claims 1-8 additives than the element isolation region for isolating the between other pixels The solid-state imaging device according to claim 1.

Images