Light detection device and electronic apparatus

Abstract

The present disclosure relates to a light detection device and an electronic apparatus that make it possible to improve autofocus performance. Provided is a light detection device comprising a pixel array unit in which a plurality of pixels, including a pixel in which a plurality of photoelectric conversion units are formed with respect to one microlens, are arranged in a two-dimensional array. In the pixel, the plurality of photoelectric conversion units are isolated in a first direction and / or a second direction different from the first direction in plan view. The isolation performance in the first direction is lower than the isolation performance in the second direction. The present disclosure is applicable to, for example, a solid-state imaging device such as a CMOS image sensor.

Description

Optical Detection Device and Electronic Device

[0001] The present disclosure relates to an optical detection device and an electronic device, and particularly to an optical detection device and an electronic device capable of improving autofocus performance.

[0002] As one of the autofocus methods, a so-called image-plane phase difference method is known in which a pair of pupil division signals is acquired using pixels for phase difference detection to perform phase difference type focus detection. As pixels for phase difference detection, pixels in which a plurality of photoelectric conversion units are formed for one microlens can be used (for example, see Patent Document 1). A microlens is a lens that condenses incident light of a pixel onto a photoelectric conversion unit. On the other hand, outside an optical detection device (solid-state imaging device) in which pixels are arranged in a two-dimensional array, a photographing lens that forms an image of a subject is arranged.

[0003] International Publication No. 2023 / 017838

[0004] In an optical detection device, since incident light transmitted through a photographing lens is irradiated, light from a subject enters obliquely into pixels arranged at the peripheral portion. Therefore, in pixels for phase difference detection arranged at the peripheral portion, the condensing position by the microlens is shifted from the photoelectric conversion unit, but there is a risk that autofocus performance may deteriorate due to manufacturing variations, replacement of the photographing lens, etc., and there has been a demand to improve autofocus performance.

[0005] The present disclosure has been made in view of such a situation, and aims to improve autofocus performance.

[0006] An optical detection device according to one aspect of the present disclosure includes a pixel array unit in which a plurality of pixels including pixels in which a plurality of photoelectric conversion units are formed for one microlens are arranged in a two-dimensional array, and in the pixels, the plurality of photoelectric conversion units are separated in at least one of a first direction and a second direction different from the first direction in a plan view, and the separation performance in the first direction is lower than the separation performance in the second direction.

[0007] An electronic device relating to one aspect of this disclosure comprises a photographic lens, a photodetector having a pixel array that arranges a plurality of pixels in a two-dimensional array, including pixels into which light from the photographic lens is incident and which have a plurality of photoelectric conversion units formed on a single microlens, and a control unit that performs autofocus control based on phase difference information obtained from the photodetector, wherein the plurality of photoelectric conversion units of the pixels are separated in a plan view in at least one of a first direction and a second direction different from the first direction, and the separation performance in the first direction is lower than the separation performance in the second direction.

[0008] Furthermore, the optical detection device and electronic equipment, which are aspects of this disclosure, may be independent devices or internal blocks constituting a single device.

[0009] This figure shows an example configuration of one embodiment of an imaging device to which the present disclosure is applied. This figure shows an example configuration of one embodiment of a photodetector to which the present disclosure is applied. This figure illustrates an example of the relationship between the lens pupil distance and the sensor pupil distance in a pixel for phase difference detection. This figure illustrates a first example of the configuration of a pixel capable of pupil separation in two directions. This figure shows an example of the output of the photoelectric conversion unit in left-right separation and up-down separation. This figure illustrates a second example of the configuration of a pixel capable of pupil separation in two directions. This figure illustrates a third example of the configuration of a pixel capable of pupil separation in two directions. This figure illustrates a fourth example of the configuration of a pixel capable of pupil separation in two directions. This figure illustrates a fifth example of the configuration of a pixel capable of pupil separation in two directions. This figure illustrates a sixth example of the configuration of a pixel capable of pupil separation in two directions. This figure illustrates a first example of the assignment of separation direction and separation performance. This figure illustrates a second example of the assignment of separation direction and separation performance. This figure illustrates a first example of the configuration of a pixel that achieves different separation performances. This figure shows examples of the A-A' and B-B' cross-sections in the planar configuration of Figure 13. This figure illustrates a second example of the configuration of a pixel that achieves different separation performances. This figure shows examples of the A-A' and B-B' cross-sections in the planar configuration of Figure 15. This figure illustrates another example of a pixel configuration that achieves different separation performance. This figure illustrates a third example of a pixel configuration that achieves different separation performance. This figure shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 18. This figure illustrates a fourth example of a pixel configuration that achieves different separation performance. This figure shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 20. This figure illustrates a fifth example of a pixel configuration that achieves different separation performance. This figure shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 22. This figure illustrates another example of a pixel configuration that achieves different separation performance. This figure shows an example of A-A' cross-section in the planar configuration of Figure 24. This figure illustrates another example of a pixel configuration that achieves different separation performance. This figure shows an example of A-A' cross-section in the planar configuration of Figure 26. This figure illustrates a sixth example of a pixel configuration that achieves different separation performance. This figure shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 28. This figure illustrates a seventh example of a pixel configuration that achieves different separation performance. This figure shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 30.This figure illustrates another example of a pixel configuration that achieves different separation performance. This figure illustrates an eighth example of a pixel configuration that achieves different separation performance. This figure shows examples of the A-A' and B-B' cross-sections in the planar configuration of Figure 33. This figure illustrates a first example of a pixel configuration in which the separation performance is changed according to the image height. This figure shows an example of a pixel configuration in the first region of Figure 35. This figure shows an example of a pixel configuration in the second region of Figure 35. This figure shows an example of a pixel configuration in the third region of Figure 35. This figure shows an example of a pixel configuration in the fourth region of Figure 35. This figure illustrates a second example of a pixel configuration in which the separation performance is changed according to the image height. This figure shows an example of a pixel configuration in the first region of Figure 40. This figure shows an example of a pixel configuration in the second region of Figure 40. This figure shows an example of a pixel configuration in the third region of Figure 40. This figure shows an example of a pixel configuration in the fourth region of Figure 40. This figure illustrates a third example of a pixel configuration in which the separation performance is changed according to the image height. This figure shows an example of a pixel configuration in the first region of Figure 45. This figure shows an example of a pixel configuration in the second region of Figure 45. This figure shows an example of a pixel configuration in the third region of Figure 45. This figure shows an example of a pixel configuration in the fourth region of Figure 45. This figure shows another example of a microlens configuration. This figure shows examples of pupil projection by an inner lens in the second and third regions of Figure 45. This figure illustrates an example of pupil projection using the conventional technique. This figure illustrates an example of pupil projection using the conventional technique.

[0010] <Device Configuration> Figure 1 is a diagram showing an example configuration of one embodiment of an imaging device to which the present disclosure is applied. As shown in Figure 1, the imaging device 1 consists of a photographic lens 11, a light detection unit 12, a signal processing unit 13, a compression / decompression unit 14, a control unit 15, an operation unit 16, a display unit 17, and a storage unit 18.

[0011] The photographic lens 11 focuses (forms an image) light (optical image) from the subject onto the light detection surface of the light detection unit 12. The photographic lens 11 is configured, for example, as a lens unit included in the optical lens barrel, and the optical mechanism (AF mechanism, etc.) is driven by a drive circuit under the control of the control unit 15 to realize functions such as autofocus. The light detection unit 12 detects the light incident from the photographic lens 11 and outputs a signal. The light detection unit 12 is composed of a light detection device including, for example, a solid-state imaging device such as a CMOS (Complementary Metal Oxide Semiconductor) type image sensor.

[0012] The signal processing unit 13, in accordance with the control unit 15, performs predetermined signal processing (e.g., white balance adjustment processing, color correction processing, etc.) on the signal output from the light detection unit 12 and outputs it as image data to the compression / decompression unit 14. The compression / decompression unit 14, in accordance with the control unit 15, performs compression encoding processing on the image data output from the signal processing unit 13 using a predetermined method. The compression / decompression unit 14 also performs decompression / decoding processing on the encoded image data supplied by the control unit 15 using a predetermined method, in accordance with the control unit 15. The predetermined method includes, for example, the JPEG (Joint Photographic Experts Group) method and the MPEG (Moving Picture Experts Group) method.

[0013] The control unit 15 is composed of a microcontroller having, for example, a CPU (Central Processing Unit), ROM (Read Only Memory), and RAM (Random Access Memory). The CPU controls each part of the imaging device 1 by executing a program stored in the ROM. For example, the control unit 15 performs autofocus based on phase difference information obtained from the light detection unit 12. The operation unit 16 is composed of, for example, buttons, and outputs signals to the control unit 15 according to user input operations. The display unit 17 is composed of, for example, an LCD (Liquid Crystal Display) or an organic EL display, and displays an image corresponding to the image data supplied from the control unit 15. The storage unit 18 is a recording medium such as a portable semiconductor memory, and stores image data compressed and encoded by the compression / decompression unit 14. The storage unit 18 also supplies the stored image data to the control unit 15 according to control from the control unit 15.

[0014] Figure 2 shows an example of the configuration of a photodetector unit 12 as a photodetector device to which the present disclosure is applied. As shown in Figure 2, the photodetector unit 12 consists of a pixel array unit 21, a vertical drive unit 22, a column signal processing unit 23, a horizontal drive unit 24, an output unit 25, and a control unit 26.

[0015] The pixel array section 21 has a plurality of pixels 31 arranged in a two-dimensional array on a semiconductor substrate made of silicon (Si) or the like. Each pixel 31 has a photoelectric conversion section composed of a photodiode (PD), a pixel transistor, and the like. For each row of the plurality of pixels 31 arranged in a two-dimensional array in the pixel array section 21, a pixel drive line 32 is formed and connected to a vertical drive section 22, and a vertical signal line 33 is formed for each column and connected to a column signal processing section 23.

[0016] The vertical drive unit 22 consists of a shift register, an address decoder, etc., and drives each pixel 31 arranged in the pixel array unit 21. The signals output from the pixels 31 selected and scanned by the vertical drive unit 22 are supplied to the column signal processing unit 23 through the vertical signal line 33. The column signal processing unit 23 performs predetermined signal processing (for example, AD conversion) on the signals output from each pixel 31 in the selected row through the vertical signal line 33 for each pixel column of the pixel array unit 21, and temporarily holds the signal after signal processing.

[0017] The horizontal drive unit 24 consists of a shift register, an address decoder, etc., and sequentially selects the unit circuits corresponding to the pixel sequences of the column signal processing unit 23. Through the selection scan by the horizontal drive unit 24, the signals processed by the column signal processing unit 23 are output to the output unit 25 via the horizontal signal line 34. The output unit 25 performs predetermined signal processing on the signals sequentially input from each of the column signal processing units 23 via the horizontal signal line 34, and outputs the resulting signal.

[0018] The control unit 26 consists of a timing generator and the like that generates various timing signals, and controls the drive of the vertical drive unit 22, the column signal processing unit 23, and the horizontal drive unit 24 based on the various timing signals generated by the timing generator.

[0019] <Background> Figure 3 illustrates an example of the relationship between the lens pupil distance and the sensor pupil distance in a pixel for phase difference detection. In Figure 3, the cross-sectional configuration of pixel 31-2, located in the center, and pixels 31-1 and 31-3, located at the periphery, are shown among the pixels 31 arranged in a two-dimensional array in the pixel array section 21. Pixel 31-1 is a pixel (a pixel for phase difference detection) in which multiple photoelectric conversion units 51 are formed on a single microlens 43. Pixels for phase difference detection are used in autofocus (AF) control utilizing the image plane phase difference method. The multiple photoelectric conversion units 51 are formed on a semiconductor substrate 41, and an intermediate layer 42 is provided between the semiconductor substrate 41 and the microlens 43. Pixels 31-2 and 31-3 are configured similarly to pixel 31-1. The microlens 43 is a lens (on-chip lens) that focuses the incident light from the pixel 31 onto the multiple photoelectric conversion units 51.

[0020] In the pixel array section 21, incident light transmitted through the imaging lens 11 is irradiated, so light from the subject is incident at an oblique angle on the pixels located at the periphery. For this reason, the light-collecting position by the microlens 43 is shifted from the photoelectric conversion section 51 in the pixels for phase difference detection (pixels 31-1, 31-3, etc.) located at the periphery of the pixel array section 21. In the pixel array section 21, the microlenses 43 of each pixel 31 are arranged to be continuously shifted toward the center according to the image height coordinate on a two-dimensional plane. That is, the microlenses 43 of each pixel 31 are arranged to be eccentric toward the center as the image height increases. In Figure 3, the microlenses 43 of pixels 31-1 and 31-3 located at the periphery of the pixel array section 21 are arranged to be eccentric toward the center. Furthermore, manufacturing variations (such as variations in the position of the microlenses 43) and replacement of the photographic lens 11 may cause a decrease in the autofocus performance based on phase difference information detected using the phase difference detection pixels (pixels 31-1, 31-3, etc.) located at the periphery of the pixel array 21.

[0021] Figure 3 schematically shows the relationship between the lens pupil distance and the sensor pupil distance. When the sensor pupil distance is d2, the lens pupil distance must be set to d2. In the example in Figure 3, the lens pupil distance can be the distance between the lens pupil surface of the imaging lens, such as the photographic lens 11, and the microlens 43, while the sensor pupil distance can be the distance between the sensor pupil surface and the microlens 43. The sensor pupil surface can be, for example, the plane corresponding to the position where the line representing the direction of incidence of light incident on the microlens 43 of the peripheral pixel 31-1 intersects with the line representing the angle of incidence of light for the central pixel 31-2 (the vertical line in the figure) (the plane corresponding to the dashed ellipse in the figure).

[0022] However, due to manufacturing variations or the replacement of the photographic lens 11, when the lens pupil distance becomes a different distance from d2 (d1, d3, d4), a pupil distance mismatch occurs, leading to a decrease in autofocus performance. In this case, the technology disclosed in the aforementioned Patent Document 1 makes it possible to suppress the decrease in autofocus performance even when a pupil distance mismatch occurs by intentionally weakening the light-gathering of the microlens. On the other hand, weakening the light-gathering of the microlens may suppress the autofocus performance that could have been achieved otherwise.

[0023] In other words, when the light-gathering of the microlens is strengthened, autofocus performance can be improved when the lens pupil distance matches the sensor pupil distance, but autofocus performance deteriorates when the lens pupil distance differs from the sensor pupil distance. On the other hand, when the light-gathering of the microlens is weakened, autofocus performance cannot be improved to the same extent as when the light-gathering is strengthened, when the lens pupil distance matches the sensor pupil distance, but autofocus performance does not deteriorate as much as when the light-gathering is strengthened, when the lens pupil distance differs from the sensor pupil distance.

[0024] Thus, there is a trade-off relationship between increasing and decreasing the light-gathering of the microlens and the autofocus performance. Therefore, when the lens pupil distance matches the sensor pupil distance, it is required that the autofocus performance be improved, similar to when light-gathering is increased, and when the lens pupil distance deviates from the sensor pupil distance, it is required that the autofocus performance not decrease as much as when light-gathering is increased, similar to when light-gathering is decreased. This disclosure proposes a light detection device capable of achieving such autofocus performance.

[0025] <Pixel Configuration> The configuration of pixels 31 arranged in a two-dimensional array in the pixel array section 21 of a photodetector to which this disclosure is applied will be described below.

[0026] <<Pixels capable of pupil separation in two directions>> A first example of the configuration of pixels capable of pupil separation in two directions will be explained with reference to Figures 4 and 5. Figure 4 shows an example of the planar configuration of a pixel 31. In the pixel array section 21, the color filters of the pixels 31 are arranged in a Bayer array, and a portion of this area is shown in Figure 4. As shown in Figure 4, in the pixel array section 21, pixels 31G are arranged in a checkerboard pattern, and in the remaining portion, pixels 31R and pixels 31B are arranged alternately in rows.

[0027] Pixel 31R is equipped with a color filter 61R corresponding to the red wavelength band, and is a pixel that generates an electric charge corresponding to the red component of light from the light transmitted through the color filter 61R. Pixel 31G is equipped with a color filter 61G corresponding to the green wavelength band, and is a pixel that generates an electric charge corresponding to the green component of light from the light transmitted through the color filter 61G. Pixel 31B is equipped with a color filter 61B corresponding to the blue wavelength band, and is a pixel that generates an electric charge corresponding to the blue component of light from the light transmitted through the color filter 61B.

[0028] As shown by the vertical line 71 and the horizontal line 72 in the figure, the photoelectric conversion section of the pixel 31R, which is composed of photodiodes (PDs), is divided into four sections in a 2x2 (horizontal x vertical) arrangement. That is, the pixel 31R is separated in the first direction D1 as shown by line 71 and in the second direction D2 as shown by line 72, resulting in a structure in which four photoelectric conversion sections 51 are provided for one microlens 43. In Figure 4, the first direction D1 and the second direction D2 are indicated by bidirectional arrows, and the second direction D2 is perpendicular to the first direction D1.

[0029] In pixels 31G and 31B, similar to pixel 31R, the photoelectric conversion units 51 are separated into a first direction D1 and a second direction D2, as shown by lines 71 and 72, and each microlens 43 is provided with four photoelectric conversion units 51. Thus, pixels 31R, 31G, and 31B have a structure in which each microlens 43 is provided with four 2x2 photoelectric conversion units 51, and since the photoelectric conversion units, which are composed of photodiodes (PDs), are separated into a first direction D1 and a second direction D2, it becomes possible to detect light in two directions, the first direction D1 and the second direction D2.

[0030] In pixel 31R, the pupil separation performance of the imaging lens 11 differs between the first direction D1 and the second direction D2. Specifically, the separation performance in the first direction D1 is lower than that in the second direction D2. In Figure 4, the width of the arrow and line 71 indicating the first direction D1 is narrower than the width of the arrow and line 72 indicating the second direction D2, illustrating the lower separation performance. Similarly, in pixels 31G and 31B, the separation performance in the first direction D1 is lower than that in the second direction D2. In each pixel 31, the separation performance in the first direction D1 is lower than that in the second direction D2, making it possible to select between the output of the photoelectric conversion unit separated left and right by line 71 (hereinafter also referred to as left-right separated PD) and the output of the photoelectric conversion unit separated up and down by line 72 (hereinafter also referred to as up-down separated PD).

[0031] Here, depending on the desired performance, a signal corresponding to the charge generated by the left-right separated PD or a signal corresponding to the charge generated by the up-down separated PD can be selected. For example, if prioritizing the effect of pupil distance deviation, the output of the left-right separated PD separated in the first direction D1, which has lower separation performance, can be used. If prioritizing peak performance, the output of the up-down separated PD separated in the second direction D2, which has higher separation performance, can be used.

[0032] Figure 5 shows examples of the output of the photoelectric conversion unit in left-right separation and up-down separation. In Figure 5, the vertical axis represents the output of the photoelectric conversion unit, and the horizontal axis represents the angle of incidence of light, showing the output of the left-right separated PD and up-down separated PD according to the angle of incidence of light.

[0033] In Figure 5, in a left-right separated photoelectric generator (PD), the output of one photoelectric converter (e.g., the left PD) is represented by waveform A, and the output of the other photoelectric converter (e.g., the right PD) is represented by waveform B. Waveforms A and B have a relationship such that when the output value of one increases, the output value of the other decreases. Similarly, in a top-bottom separated PD, the output of one photoelectric converter (e.g., the upper PD) is represented by waveform C, and the output of the other photoelectric converter (e.g., the lower PD) is represented by waveform D. Waveforms C and D have a relationship such that when the output value of one increases, the output value of the other decreases. For example, waveforms A and B are more robust and correspond to pupillary distance deviations, while waveforms C and D are more vertical and correspond to peak performance. Waveforms A and B have different slopes than waveforms C and D.

[0034] As shown by waveforms A and B, and waveforms C and D, the left-right separated PD and the up-down separated PD produce different outputs (signals) depending on their separation performance, allowing the output to be selected according to the performance to be prioritized. For example, when the lens pupil distance differs from the sensor pupil distance, and the effect of the pupil distance difference is prioritized, the output of the left-right separated PD can be selected to use the output corresponding to waveforms A and B. Conversely, when the lens pupil distance matches the sensor pupil distance (or when the pupil distance difference is small), and peak performance is prioritized, the output of the up-down separated PD can be selected to use the output corresponding to waveforms C and D. This makes it possible to suppress the decrease in autofocus performance when the pupil distance difference is large, while maintaining high autofocus performance when the pupil distances match (or when the pupil distance difference is small). Thus, autofocus performance can be improved.

[0035] Referring to Figure 6, a second example of a pixel configuration capable of pupil separation in two directions will be described. Figure 6 shows an example of a planar configuration of pixel 31. In Figure 6, parts corresponding to those in Figure 4 are given the same reference numerals, and their explanations will be omitted as appropriate.

[0036] In Figure 6, the pixel 31R is divided into four 2x2 sections, as shown by the vertical and horizontal lines 71 in the figure. Specifically, the pixel 31R is provided with two horizontally elongated elliptical microlenses 43, separated in the first direction D1 as shown by the vertical line 71, and each microlens 43 is provided with two 2x1 photoelectric conversion sections 51.

[0037] Pixel 31G has four circular microlenses 43, with one photoelectric conversion unit 51 provided for each microlens 43. Pixel 31B has two vertically elongated elliptical microlenses 43, separated in a second direction D2 as shown by the horizontal line 72, with two 1x2 photoelectric conversion units 51 provided for each microlens 43.

[0038] In Figure 6, pixels 31R and 31B of the pixels 31 have a structure in which two photoelectric conversion units 51 are provided for one microlens 43. In pixel 31R, the photoelectric conversion unit is separated in the first direction D1, and in pixel 31B, the photoelectric conversion unit is separated in the second direction D2. Therefore, pixels 31R and 31B can detect light in two directions, the first direction D1 and the second direction D2. Furthermore, in pixels 31R and 31B, the separation performance in the first direction D1 is made lower than the separation performance in the second direction D2. This makes it possible to select between the output of the left and right separated photoelectric conversion units (left and right separated PD) in pixel 31R and the output of the up and down separated photoelectric conversion units (up and down separated PD) in pixel 31B.

[0039] Referring to Figure 7, a third example of a pixel configuration capable of pupil separation in two directions will be described. Figure 7 shows an example of a planar configuration of pixel 31. In Figure 7, parts corresponding to those in Figure 4 are given the same reference numerals, and their explanations will be omitted as appropriate.

[0040] In Figure 7, the photoelectric conversion unit of pixel 31R is divided into two sections, left and right, as shown by the vertical line 71 in the figure. That is, pixel 31R is separated in the first direction D1, and has a structure in which two 2x1 photoelectric conversion units 51 are provided for one microlens 43. Similarly, pixel 31G is separated in the first direction D1 as shown by the vertical line 71, and has a structure in which two 2x1 photoelectric conversion units 51 are provided for one microlens 43.

[0041] As shown by the vertical line 71 in the figure, the photoelectric conversion unit of pixel 31B is divided into two parts, left and right. Also, in pixel 31B, as shown by the horizontal line 72, the area corresponding to the upper 2x1 photoelectric conversion unit 51 of the vertically separated photoelectric conversion unit is shielded by the light-shielding unit 81, resulting in a structure where the lower 2x1 photoelectric conversion unit 51 is provided for one microlens 43. By providing the light-shielding unit 81, for example, a pair of pixels with the upper half shielded and the lower half shielded, which are placed in close proximity, can be used as pixels for phase difference detection. In this example, the pixel array unit 21 is arranged with pixels that output a signal according to the amount of light from the incident subject and pixels from which the incident subject light is shielded.

[0042] In FIG. 7, among the pixels 31, the pixel 31R and the pixel 31G are structured such that two photoelectric conversion units 51 are provided for one microlens 43, and the respective photoelectric conversion units are separated in the first direction D1. Also, the pixel 31B has its photoelectric conversion unit shielded by a light-shielding unit 81, and in the pixel 31B (a set of pixels with the upper half and the lower half shielded), the photoelectric conversion unit is separated in the first direction D1 and the second direction D2. Therefore, it becomes possible to detect light in two directions, namely, the first direction D1 by the pixels 31R, 31G, 31B and the second direction D2 by the pixel 31B.

[0043] Also, the separation performance in the first direction D1 in the pixels 31R, 31G, 31B is made lower than the separation performance in the second direction D2 in the pixel 31B. Thereby, it becomes possible to select the outputs of the left and right separated photoelectric conversion units (left and right separated PD) in the pixels 31R, 31G, 31B and the outputs of the upper and lower separated photoelectric conversion units (upper and lower separated PD) in the pixel 31B. Note that, as a pixel for phase difference detection, a configuration using a set of a pixel with the left half shielded and a pixel with the right half shielded arranged adjacent to each other may be adopted.

[0044] While referring to FIG. 8, a fourth example of the configuration of a pixel capable of pupil separation in two directions will be described. FIG. 8 shows an example of the planar configuration of the pixel 31. In FIG. 8, the parts corresponding to FIGS. 4 and 7 are given the same reference numerals, and the description thereof will be omitted as appropriate.

[0045] In Figure 8, as shown by the vertical line 71 in the figure, pixel 31R has a structure in which the left region of the photoelectric conversion unit, which is separated into left and right sections, is shielded by the light-shielding section 81, and a right-side photoelectric conversion unit 51 is provided for one microlens 43. Pixel 31G has a structure in which one photoelectric conversion unit 51 is provided for one microlens 43. Pixel 31B has a structure in which the upper region of the photoelectric conversion unit, which is separated into upper and lower sections, is shielded by the light-shielding section 81, as shown by the horizontal line 72, and a lower-side photoelectric conversion unit 51 is provided for one microlens 43. In this example, a pair of pixels with the left half shielded and a pair of pixels with the right half shielded, and a pair of pixels with the upper half shielded and a pair of pixels with the lower half shielded, can be used as pixels for phase difference detection. In this example as well, the pixel array 21 has pixels that output a signal according to the amount of light from the incident subject and pixels from which the incident subject light is shielded.

[0046] In Figure 8, light-shielding sections 81 are provided for pixels 31R and 31B, respectively, so that the photoelectric conversion section is separated in a first direction D1 for pixel 31R and in a second direction D2 for pixel 31B. As a result, it is possible to detect light in two directions: the first direction D1 by pixel 31R (a pair of pixels with the left and right halves shielded) and the second direction D2 by pixel 31B (a pair of pixels with the upper and lower halves shielded). Furthermore, the separation performance in the first direction D1 for pixel 31R is made lower than the separation performance in the second direction D2 for pixel 31B. This makes it possible to select between the output of the left and right separated photoelectric conversion section (left and right separated PD) in pixel 31R and the output of the up and down separated photoelectric conversion section (up and down separated PD) in pixel 31B.

[0047] A fifth example of a pixel configuration capable of pupil separation in two directions will be described with reference to Figure 9. Figure 9 shows an example of a planar configuration of pixel 31. In Figure 9, parts corresponding to those in Figure 4 are given the same reference numerals, and their explanations will be omitted as appropriate.

[0048] In FIG. 9, as shown by the vertical line 71 in the figure, the photoelectric conversion section of pixel 31R is divided into two parts in a 2×1 configuration. Pixel 31R is divided in the first direction D1 and has a structure in which two photoelectric conversion sections 51 are provided on the left and right with respect to one microlens 43. Similarly, pixel 31B is separated in the first direction D1 as shown by the line 71 and has a structure in which two photoelectric conversion sections 51 are provided on the left and right with respect to one microlens 43.

[0049] Of the two pixels 31G, one pixel 31G-1 is separated in the first direction D1 as shown by the line 71 and has a structure in which two photoelectric conversion sections 51 are provided on the left and right with respect to one microlens 43. The other pixel 31G-2 is separated in the second direction D2 as shown by the line 72 and has a structure in which two photoelectric conversion sections 51 are provided above and below with respect to one microlens 43.

[0050] In FIG. 9, each pixel 31 has a structure in which two photoelectric conversion sections 51 are provided with respect to one microlens 43. For pixels 31R, 31G-1, and 31B, the photoelectric conversion sections are separated in the first direction D1, and for pixel 31G-2, the photoelectric conversion sections are separated in the second direction D2. Therefore, it is possible to detect light in two directions: the first direction D1 by pixels 31R, 31G-1, and 31B, and the second direction D2 by pixel 31G-2. Also, the separation performance in the first direction D1 for pixels 31R, 31G-1, and 31B is made lower than the separation performance in the second direction D2 for pixel 31G-2. As a result, it becomes possible to select the output of the photoelectric conversion section separated in the first direction D1 (left-right separated PD) and the output of the photoelectric conversion section separated in the second direction D2 (up-down separated PD).

[0051] Referring to FIG. 10, a sixth example of a pixel configuration capable of pupil separation in two directions will be described. FIG. 10 shows an example of the planar configuration of pixel 31. In FIG. 10, parts corresponding to FIGS. 4 and 9 are given the same reference numerals, and their descriptions will be omitted as appropriate.

[0052] In FIG. 10, pixels 31R and 31B are separated in the first direction D1 as shown by the line 71 in the figure, similar to FIG. 9, and have a structure in which two photoelectric conversion sections 51 separated on the left and right are provided with respect to one microlens 43.

[0053] Of the two pixels 31G, one pixel 31G-1 is separated in a second direction D2-1 as shown by line 72, which is a diagonal line sloping downwards to the left from the top edge to the bottom edge, and has a structure in which two diagonally separated photoelectric conversion units 51 are provided with one microlens 43. The other pixel 31G-2 is divided in a second direction D2-2 as shown by line 72, which is a diagonal line sloping downwards to the right from the top edge to the bottom edge, and has a structure in which two diagonally separated photoelectric conversion units 51 are provided with one microlens 43. Note that the second directions D2-1 and D2-2 may be considered together as the second direction D2, since they are different directions from the first direction D1, or they may be considered as separate directions, with the pixels separated in three directions including the first direction D1.

[0054] In Figure 10, each pixel 31 has a structure in which two photoelectric conversion units 51 are provided for one microlens 43. In pixels 31R and 31B, the photoelectric conversion units are separated in the first direction D1, and in pixels 31G-1 and 31G-2, the photoelectric conversion units 51 are separated in the second direction D2 (D2-1, D2-2). Therefore, it is possible to detect light in the first direction D1 by pixels 31R and 31B and in the second direction D2 (D2-1, D2-2) by pixels 31G-1 and 31G-2. Furthermore, the separation performance in the first direction D1 in pixels 31R and 31B is made lower than the separation performance in the second direction D2 (D2-1, D2-2) in pixels 31G-1 and 31G-2. This makes it possible to select between the output of the photoelectric conversion unit (left and right separated PD) separated in the first direction D1 and the output of the photoelectric conversion unit 51 (diagonally separated PD) separated in the second direction D2 (D2-1, D2-2).

[0055] As described above, in the photodetector to which this disclosure is applied, the plurality of pixels 31 arranged in a two-dimensional array in the pixel array section 21 include pixels 31 in which a plurality of photoelectric conversion sections 51 are formed on a single microlens 43, and the pixels 31 are separated in at least one of the first direction D1 and the second direction D2 in a plan view. Note that the two-directional pupil separation of pixels shown in Figures 4 to 10 is just one example, and other configurations may be adopted as long as pupil separation in two directions is possible.

[0056] <<Example of Separation Direction and Separation Performance Allocation>> Figure 11 is a diagram illustrating a first example of the allocation of separation direction and separation performance. Figure 11 shows the external configuration of a photodetector 12 as a photodetector to which the present disclosure is applied. As shown in Figure 11, in the photodetector 12, light focused by the photographic lens 11 is incident on the photodetector surface 12A. The photodetector surface 12A is provided with a pixel array section 21 in which a plurality of pixels 31 are arranged in a two-dimensional array. In the pixel array section 21, a plurality of pixels 31 are arranged so that the aspect ratio of the photodetector surface 12A is horizontally elongated.

[0057] In this case, the higher the image height, the more susceptible it is to the effect of pupillary distance deviation. Focusing on the horizontal direction, the pixels 31 located at the periphery have a higher image height. Therefore, as indicated by the arrow on the light detection surface 12A, the horizontal direction is designated as the first direction D1, and the separation performance is lowered (the separation is weakened). This output is used when prioritizing the effect of pupillary distance deviation. Furthermore, focusing on the vertical direction, the image height is not as high as in the horizontal direction. Therefore, as indicated by the arrow on the light detection surface 12A, the vertical direction is designated as the second direction D2, and the separation performance is higher (the separation is strengthened). This output is used when prioritizing peak performance.

[0058] In this way, the first direction D1 and the second direction D2 can be determined according to the aspect ratio of the photodetector surface 12A, and the separation direction and separation performance (separation strength) can be assigned. In the example in Figure 11, the separation performance is lower when the first direction D1 separates the multiple photoelectric conversion units 51 left and right, and the separation performance is higher when the second direction D2 separates the multiple photoelectric conversion units 51 up and down.

[0059] Figure 12 illustrates a second example of the assignment of separation direction and separation performance. Figure 12 shows an enlarged view of some pixels 31 in a specific row among a plurality of pixels 31 arranged in a two-dimensional array in the pixel array unit 21. In the pixel array unit 21, a plurality of pixels 31 are arranged so that the aspect ratio of the light detection surface is horizontally elongated. In the pixel array unit 21, phase difference information can be obtained by discretely including pairs of pixels with the left side shaded and pixels with the right side shaded as pixels for phase difference detection in a specific row.

[0060] In Figure 12, among the enlarged 4x2 pixels 31, the pixels 31B located in the lower row are configured as pixels with the right side shielded by the light-shielding portion 81 (right-shielded pixels) and pixels with the left side shielded by the light-shielding portion 81 (left-shielded pixels). Here, since the lens pupil distance can be adjusted by design for the pixels shielded by the light-shielding portion 81 (left-shielded pixels, right-shielded pixels), it is possible to change the lens pupil distance corresponding to the left-shielded pixels and right-shielded pixels for each row. For example, the design values ​​can be changed so that the left-shielded pixels and right-shielded pixels located in one row correspond to the lens pupil distance d11, and the left-shielded pixels and right-shielded pixels located in other rows correspond to the lens pupil distance d12.

[0061] In Figure 12, when using shaded pixels (left shaded pixel, right shaded pixel) as pixels for phase difference detection, it is possible to adjust the pupil distance during the design. Therefore, the shaded pixels (left shaded pixel, right shaded pixel) are assigned to the lateral direction, and the lateral direction is designated as the second direction D2 to improve separation performance. That is, since the pupil distance can be controlled in the lateral direction, which is the second direction D2, the separation performance is increased (separation is strengthened), and its output is used when peak performance is prioritized. On the other hand, since the pupil distance cannot be controlled in the vertical direction, the separation performance is reduced (separation is weakened), and its output is used when the effect of pupil distance deviation is prioritized.

[0062] In this way, the first direction D1 and the second direction D2 can be determined depending on whether or not the pupil distance is controlled, and the separation direction and separation performance (separation strength) can be assigned. In the example in Figure 12, when a pixel for phase difference detection is included, which is configured to include a light-shielding portion 81 that shiels a predetermined area adjusted to correspond to the second direction D2, the separation performance decreases in the direction where the first direction D1 separates the multiple photoelectric conversion units 51 vertically, and the separation performance increases in the direction where the second direction D2 separates the multiple photoelectric conversion units 51 horizontally. In the example in Figure 12, an example of adjusting the lens pupil distance by design for pixels that are shielded on the left and right sides (left shielded pixel, right shielded pixel) has been described, but the present disclosure can also be applied similarly to pixels that are shielded on the upper and lower sides (upper shielded pixel, lower shielded pixel).

[0063] <<Means for Achieving Different Separation Performance>> A first example of a pixel configuration that achieves different separation performance will be explained with reference to Figures 13 and 14. Figure 13 shows an example of a planar configuration of a pixel 31. Figure 14 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 13. In Figure 14, when the side of the main surface of the semiconductor substrate 41 on which the wiring layer is provided is considered the front surface and the side opposite to the front surface is considered the back surface, the back surface, which is the incident surface of light, is shown on the upper side. In Figure 14, a cross-section of a pixel 31 located at the periphery of a plurality of pixels 31 arranged in a two-dimensional array in the pixel array section 21 is shown, so the microlens 43 is arranged to be eccentric toward the center.

[0064] As shown in Figure 13, the semiconductor substrate 41 has a grid of inter-pixel isolation sections 111 formed around each of the photoelectric conversion sections of each pixel 31. The inter-pixel isolation sections 111 are formed, for example, by embedding a substance (material) such as tungsten or an oxide film in trenches formed from the back side of the semiconductor substrate 41. These trenches do not penetrate to the front side of the semiconductor substrate 41 and are non-penetrating trenches.

[0065] As shown in Figure 14, a waveguide 121 and an inner lens 131 can be formed in the intermediate layer 42 provided between the microlens 43, color filter 61, and semiconductor substrate 41, for example, an oxide film. The light L focused by the microlens 43 is incident on the photoelectric conversion unit 51 by the inner lens 131. Inter-pixel light-shielding sections 112 and 113 are formed in the intermediate layer 42, corresponding to the inter-pixel separation section 111. The inter-pixel light-shielding sections 112 and 113 are made of a material such as tungsten, and can block light.

[0066] In the pixels 31 shown in Figures 13 and 14, a structure is realized in which four photoelectric conversion units 51 (2x2) are provided for one microlens 43 by forming an inter-element isolation unit 111A that separates the photoelectric conversion unit, which is composed of a photodiode (PD), in a first direction D1, and an inter-element isolation unit 111B that separates it in a second direction D2. The inter-element isolation units 111A and 111B are constructed by embedding a material such as tungsten or an oxide film in a non-penetrating trench formed from the back side of the semiconductor substrate 41, similar to the inter-pixel isolation unit 111. In other words, the inter-element isolation units 111A and 111B have the same configuration as the inter-pixel isolation unit 111 and serve as the same isolation means as the inter-pixel isolation unit 111. The inter-element isolation units 111A and 111B may also be configured as part of the inter-pixel isolation unit 111.

[0067] The element-to-element separation section 111A has a narrower line width compared to the element-to-element separation section 111B, resulting in lower separation performance in the first direction D1 than in the second direction D2. On the other hand, the element-to-element separation section 111B has a wider line width compared to the element-to-element separation section 111A, resulting in higher separation performance in the second direction D2 than in the first direction D1. Thus, as a means of realizing different separation performances in a pixel 31 capable of pupil separation in two directions, the first direction D1 and the second direction D2, an element-to-element separation section 111A that separates the photoelectric conversion section left and right, and an element-to-element separation section 111B that separates it up and down are provided, and their line widths are made different so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0068] A second example of a pixel configuration that achieves different separation performance will be described with reference to Figures 15 and 16. Figure 15 shows an example of a planar configuration of pixel 31. Figure 16 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 15. In Figures 15 and 16, the same reference numerals are used for parts corresponding to those in Figures 13 and 14, and their explanations will be omitted as appropriate.

[0069] In the pixels 31 shown in Figures 15 and 16, compared to the pixels 31 shown in Figures 13 and 14, the inter-element separation section 111A is omitted. This means that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2, depending on the presence or absence of the inter-element separation section 111B. In other words, the inter-element separation section is not formed in the first direction D1, but is formed in the second direction D2. Thus, as a means of realizing different separation performances in a pixel 31 capable of pupil separation in two directions, the first direction D1 and the second direction D2, by providing only the inter-element separation section 111B that separates the photoelectric conversion section vertically, the separation performance in the first direction D1 can be made lower than the separation performance in the second direction D2.

[0070] As shown in Figure 17, the inter-element separation portion 111A and the inter-element separation portion 111B may protrude toward the center of the pixel 31 to form a projection. In this case, the inter-element separation portion 111A and the inter-element separation portion 111B formed as projections will have different separation performances due to the difference in their protrusion amounts. In Figure 17, the protrusion amount of the inter-element separation portion 111A is smaller than that of the inter-element separation portion 111B, so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2. The line widths of the inter-element separation portion 111A and the inter-element separation portion 111B may also be the same. The A-A' and B-B' cross sections in the planar configuration of Figure 17 are the same as in Figure 16.

[0071] As described above, by embedding material in trenches formed in the semiconductor substrate 41 and forming inter-element isolation sections 111A and 111B as physical isolation sections, different isolation performance can be achieved in the first direction D1 and the second direction D2.

[0072] Next, a third example of a pixel configuration that achieves different separation performance will be described with reference to Figures 18 and 19. Figure 18 shows an example of a planar configuration of pixel 31. Figure 19 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 18. In Figures 18 and 19, the same reference numerals are used for parts corresponding to those in Figures 13 and 14, and their explanations will be omitted as appropriate.

[0073] As shown in Figures 18 and 19, the inner lens 131 formed in the intermediate layer 42 between the microlens 43 and the semiconductor substrate 41 directs the light incident from the microlens 43 into photoelectric conversion sections separated into a first direction D1 and a second direction D2 via the waveguide 121. In the pixel 31 of Figures 18 and 19, the inner lens 131 separates the photoelectric conversion section, which is composed of photodiodes (PDs), into a first direction D1 and a second direction D2, thereby realizing a structure in which four 2x2 photoelectric conversion sections 51 are provided for one microlens 43.

[0074] Here, the inner lens 131 is provided with an elliptical shape, and its cross-sectional shape differs between the A-A' and B-B' sections. As shown in Figure 19, the inner lens 131 is a plano-convex lens having a flat surface and a convex surface, but the convex portion of the convex surface is larger in the B-B' section than in the A-A' section. Therefore, in the inner lens 131 of Figure 19, the B-B' section can collect more light from the microlens 43 than the A-A' section.

[0075] When the photoelectric conversion unit shown in the A-A' cross section is separated left and right, the inner lens 131 concentrates light more weakly than when it is separated up and down, and when the photoelectric conversion unit shown in the B-B' cross section is separated up and down, the inner lens 131 concentrates light more strongly than when it is separated left and right. Therefore, depending on the strength of the light concentration, the separation performance in the first direction D1 can be made lower than the separation performance in the second direction D2. In this way, as a means of realizing different separation performances in a pixel 31 that can separate pupils in two directions, the first direction D1 and the second direction D2, an elliptical inner lens 131 is provided so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0076] A fourth example of a pixel configuration that achieves different separation performance will be described with reference to Figures 20 and 21. Figure 20 shows an example of a planar configuration of pixel 31. Figure 21 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 20. In Figures 20 and 21, parts corresponding to those in Figures 18 and 19 are denoted by the same reference numerals, and their explanations will be omitted as appropriate.

[0077] As shown in Figure 20, the inner lens 131 is provided with a cylindrical shape, and its cross-sectional shape differs between the A-A' and B-B' sections. As shown in Figure 21, the inner lens 131 is a cylindrical lens with a shape obtained by cutting off a part of a cylinder, and the B-B' section has a curved shape, while the A-A' section does not have a curved shape. Therefore, in the inner lens 131 of Figure 21, the B-B' section can collect more light from the microlens 43 than the A-A' section.

[0078] As shown in Figures 20 and 21, a cylindrical inner lens 131 is provided as a means to achieve different separation performance in a pixel 31 capable of pupil separation in two directions, a first direction D1 and a second direction D2, so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0079] A fifth example of a pixel configuration that achieves different separation performance will be described with reference to Figures 22 and 23. Figure 22 shows an example of a planar configuration of pixel 31. Figure 23 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 22. In Figures 22 and 23, the same reference numerals are used for parts corresponding to those in Figures 20 and 21, and their explanations will be omitted as appropriate.

[0080] As shown in Figures 22 and 23, the inner lens 131 formed in the intermediate layer 42 between the microlens 43 and the semiconductor substrate 41 directs the light incident from the microlens 43 into photoelectric conversion sections separated into a first direction D1 and a second direction D2. In the pixel 31 of Figures 22 and 23, the inner lens 131, composed of lenses 131A to 131D, separates the photoelectric conversion section, composed of photodiodes (PDs), into a first direction D1 and a second direction D2, thereby realizing a structure with four 2x2 photoelectric conversion sections 51.

[0081] In the inner lens 131, the four 2x2 lenses 131A to 131D are configured such that the gap between them differs. Specifically, as shown in Figure 22, when the gap between the left and right lenses 131A, 131B (131C, 131D) of the four 2x2 lenses is defined as gap g1, and the gap between the upper and lower lenses 131A, 131C (131B, 131D) is defined as gap g2, gaps g1 and g2 are different. That is, gap g2 is larger than gap g1. Therefore, in the inner lens 131 shown in Figure 23, the B-B' cross section can collect more light from the microlens 43 than the A-A' cross section.

[0082] When the photoelectric conversion unit shown in the A-A' cross section is separated left and right, the inner lens 131 concentrates light more weakly than when it is separated up and down, and when the photoelectric conversion unit shown in the B-B' cross section is separated up and down, the inner lens concentrates light more strongly than when it is separated left and right. Therefore, depending on the strength of the light concentration, the separation performance in the first direction D1 can be made lower than the separation performance in the second direction D2. In this way, as a means of realizing different separation performances in a pixel 31 that can separate pupils in two directions, the first direction D1 and the second direction D2, an inner lens 131 composed of multiple lenses with different gap amounts between the lenses is provided so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0083] As shown in Figures 24 and 25, the inner lens 131 may be composed of two elliptical lenses 131E and 131F. In Figures 24 and 25, the elliptical lenses 131E and 131F constituting the inner lens 131 are provided in correspondence with two 2x1 photoelectric conversion units 51, thereby realizing a structure in which four 2x2 photoelectric conversion units 51 are provided for one microlens 43. Here as well, the A-A' cross section shown in Figure 25 has weaker light focusing than the B-B' cross section, and the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0084] Furthermore, as shown in Figures 26 and 27, the inner lens 131 may be composed of two cylindrical lenses 131G and 131H. In Figures 26 and 27, the two cylindrical lenses 131G and 131H constituting the inner lens 131 are provided in correspondence with two 2x1 photoelectric conversion units 51, thereby realizing a structure in which four 2x2 photoelectric conversion units 51 are provided for one microlens 43. Here as well, the A-A' cross section shown in Figure 27 has weaker light focusing than the B-B' cross section, and the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0085] As described above, the inner lens 131 formed in the intermediate layer 42 between the microlens 43 and the semiconductor substrate 41 makes it possible to achieve different separation performance in the first direction D1 and the second direction D2. Furthermore, when different separation performance is achieved by the inner lens 131, a physical separation section is unnecessary, eliminating the possibility of reduced sensitivity due to light-gathering components in the separation section and color mixing due to scattered light. Therefore, it is possible to improve autofocus performance while suppressing the impact on image quality.

[0086] Next, a sixth example of a pixel configuration that achieves different separation performance will be described with reference to Figures 28 and 29. Figure 28 shows an example of a planar configuration of pixel 31. Figure 29 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 28. In Figures 28 and 29, the same reference numerals are used for parts corresponding to those in Figures 13 and 14, and their explanations will be omitted as appropriate.

[0087] In the pixels 31 of Figures 28 and 29, compared to the pixels 31 of Figures 13 and 14, there is no intermediate layer 42 between the microlens 43, color filter 61 and semiconductor substrate 41. Therefore, there is no inner lens 131. In addition, element-to-element isolation sections 112A and 112B are provided as separation means in two directions, a first direction D1 and a second direction D2. The element-to-element isolation sections 112A and 112B have a different configuration from the pixel-to-pixel isolation section 111 and are different separation means from the pixel-to-pixel isolation section 111. For example, the trench depth of the element-to-element isolation sections 112A and 112B is different from that of the pixel-to-pixel isolation section 111. That is, the trenches of the element-to-element isolation sections 112A and 112B can be shallower than the trenches of the pixel-to-pixel isolation section 111. Also, the material embedded in the trenches of the element-to-element isolation sections 112A and 112B and the pixel-to-pixel isolation section 111 may be different.

[0088] The element-to-element separation section 112A has a narrower line width compared to the element-to-element separation section 112B, resulting in lower separation performance in the first direction D1 than in the second direction D2. On the other hand, the element-to-element separation section 112B has a wider line width compared to the element-to-element separation section 112A, resulting in higher separation performance in the second direction D2 than in the first direction D1. Thus, as a means of realizing different separation performances in a pixel 31 capable of pupil separation in two directions, the first direction D1 and the second direction D2, an element-to-element separation section 112A that separates the photoelectric conversion section left and right, and an element-to-element separation section 112B that separates it up and down are provided, and their line widths are made different so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2.

[0089] A seventh example of a pixel configuration that achieves different separation performance will be described with reference to Figures 30 and 31. Figure 30 shows an example of a planar configuration of a pixel 31. Figure 31 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 30. In Figures 30 and 31, parts corresponding to those in Figures 28 and 29 are denoted by the same reference numerals, and their explanations will be omitted as appropriate.

[0090] In the pixels 31 of Figures 30 and 31, compared to the pixels 31 of Figures 28 and 29, the element-to-element separation section 112A is omitted, so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2, depending on the presence or absence of the element-to-element separation section 112B. In this way, as a means of realizing different separation performance in a pixel 31 capable of pupil separation in two directions, the first direction D1 and the second direction D2, by providing only the element-to-element separation section 112B that separates the photoelectric conversion section vertically, the separation performance in the first direction D1 can be made lower than the separation performance in the second direction D2.

[0091] As shown in Figure 32, the inter-element isolation section 112A and the inter-element isolation section 112B may be configured as protrusions that extend toward the center of the pixel 31. In this case, the inter-element isolation section 112A and the inter-element isolation section 112B, which are formed as protrusions, will have different separation performances due to the difference in their protrusion amounts. In Figure 32, the protrusion amount of the inter-element isolation section 112A is smaller than that of the inter-element isolation section 112B, so that the separation performance in the first direction D1 is lower than the separation performance in the second direction D2. The line widths of the inter-element isolation section 112A and the inter-element isolation section 112B may also be the same. The A-A' and B-B' cross sections in the planar configuration of Figure 32 are the same as in Figure 31.

[0092] As described above, even in a configuration without an inner lens 131, by embedding material in trenches formed in the semiconductor substrate 41 to form physical separation sections, such as the inter-pixel separation section 111 and the inter-element separation sections 112A and 112B, different separation performance can be achieved in the first direction D1 and the second direction D2. Furthermore, since there is no need to provide an inner lens 131, it is possible to reliably improve autofocus performance regardless of the application.

[0093] Next, an eighth example of a pixel configuration that achieves different separation performance will be described with reference to Figures 33 and 34. Figure 33 shows an example of a planar configuration of pixel 31. Figure 34 shows examples of A-A' and B-B' cross-sections in the planar configuration of Figure 33. In Figures 33 and 34, the same reference numerals are used for parts corresponding to those in Figures 28 and 29, and their explanations will be omitted as appropriate.

[0094] In the pixels 31 of Figures 33 and 34, similar to the pixels 31 of Figures 28 and 29, there is no intermediate layer 42 between the microlens 43 and the semiconductor substrate 41, and therefore there is no inner lens 131.

[0095] As shown in Figure 33, inter-pixel isolation regions 212 are formed in a grid pattern on the semiconductor substrate 41, surrounding each photoelectric conversion region of each pixel 31. The photoelectric conversion region is composed of a photodiode (PD) which is an n-type region. The inter-pixel isolation regions 212 are p-type impurity regions formed on the semiconductor substrate 41. In other words, in Figures 33 and 34, the pixels are separated by impurities.

[0096] In the pixels 31 shown in Figures 33 and 34, a structure is realized in which four 2x2 photoelectric conversion units 211 are provided for one microlens 43 by forming an inter-element isolation unit 212A that separates the photoelectric conversion unit, which is composed of a photodiode (PD), in a first direction D1, and an inter-element isolation unit 212B that separates it in a second direction D2. The inter-element isolation units 212A and 212B are p-type impurity regions formed on the semiconductor substrate 41, similar to the inter-pixel isolation unit 212. In other words, in Figures 33 and 34, the photoelectric conversion units 211 are separated by impurities, and the inter-element isolation units 212A and 212B are the same isolation means as the inter-pixel isolation unit 212.

[0097] The inter-element isolation section 212A has a wider line width compared to the inter-element isolation section 212B, and its separation performance in the first direction D1 is lower than that in the second direction D2. On the other hand, the inter-element isolation section 212B has a narrower line width compared to the inter-element isolation section 212A, and its separation performance in the second direction D2 is higher than that in the first direction D1. In physical separation (optical separation) by the inter-element isolation sections 111A and 111B, the wider the line width, the stronger the separation. However, in impurity separation by the inter-element isolation sections 212A and 212B, since the separation is based on electronic behavior, the wider the line width, the weaker the separation. For example, the inter-element isolation section 212A can be formed by narrowing the n-type region or widening the p-type region during ion implantation for PD formation, which is implanted at least in a region close to the light incident surface (back surface).

[0098] As described above, by injecting impurities into the semiconductor substrate 41 and forming inter-element isolation sections 212A and 212B as impurity isolation sections, different separation performance can be achieved in the first direction D1 and the second direction D2. Furthermore, since there is no need to provide an inner lens 131, it is possible to reliably improve autofocus performance regardless of the application.

[0099] <<Separation Performance According to Image Height>> Figure 35 illustrates a first example of a pixel configuration in which the separation performance is changed according to the image height. Figure 35 shows a plurality of pixels 31 arranged in a two-dimensional array in the pixel array section 21. In the pixel array section 21, pixels 31R, 31B, and 31G are arranged in a Bayer array so that the aspect ratio of the light detection surface 12A is horizontally elongated. In the pixel array section 21, the image height is higher for pixels 31 located towards the periphery, so as indicated by the arrows on the light detection surface 12A, the horizontal direction is designated as the first direction D11, and the separation performance can be lowered according to the image height. Also, as indicated by the arrows on the light detection surface 12A, the vertical direction is designated as the second direction D12, and the separation performance can be lowered according to the image height.

[0100] Here, in the light detection surface 12A, the central region is designated as the first region 301, the peripheral region at the right edge of the center in the figure is designated as the second region 302, the peripheral region at the upper edge of the center in the figure is designated as the third region 303, and the peripheral region at the upper right edge of the center in the figure is designated as the fourth region 304. The configuration of the pixels 31 in each region is shown in Figures 36 to 39.

[0101] Figure 36 shows an example of the configuration of pixels 31 in the first region 301 of Figure 35. In Figure 36, inter-pixel separation sections 311 are formed in a grid pattern so as to surround each photoelectric conversion section of each pixel 31. In pixel 31R, protrusions from the inter-pixel separation section 311 extend toward the center of the pixel 31R, forming inter-element separation sections 311A ​​to 311D, thereby realizing a structure in which four 2x2 photoelectric conversion sections 51 are provided for one microlens 43. The amount of protrusion of the inter-element separation sections 311A ​​to 311D is the same. The inter-element separation sections 311A ​​to 311D are provided as physical separation sections by embedding material in trenches formed in the semiconductor substrate 41, similar to the inter-pixel separation section 311. The inter-element separation sections 311A ​​to 311D may also be configured as part of the inter-pixel separation section 311. In Figure 36, pixels 31G-1, 31G-2 and pixel 31B are provided with an inter-element isolation section as a physical separation section, similar to pixel 31R.

[0102] Figure 37 shows an example of the configuration of a pixel 31 in the second region 302 of Figure 35. Since the second region 302 is a peripheral region, each pixel 31 in the second region 302 is arranged such that the microlens 43 is eccentrically positioned toward the center (left side in the figure). As shown in Figure 37, in the pixel 31R, a projection from the inter-pixel separation portion 311 extends toward the center of the pixel 31R, thereby providing inter-element separation portions 321A to 321D. The inter-element separation portions 321A to 321D are provided as physical separation portions by embedding material in trenches formed in the semiconductor substrate 41.

[0103] Here, comparing the inter-element isolation sections 321A to 321D of pixel 31R in Figure 37 with the inter-element isolation sections 311A ​​to 311D of pixel 31R in Figure 36, the inter-element isolation sections 321B and 321D have the same amount of protrusion as the inter-element isolation sections 311B and 311D, but the inter-element isolation sections 321A and 321C have a smaller amount of protrusion than the inter-element isolation sections 311A ​​and 311C. In other words, by making the amount of protrusion of the inter-element isolation sections 321A and 321C smaller than the amount of protrusion of the inter-element isolation sections 321B and 321D, the separation performance in the first direction D11 is made lower than the separation performance in the second direction D12. The line width of the inter-element isolation sections 321A and 321C is the same as the line width of the inter-element isolation sections 321B and 321D. In Figure 37, pixels 31G-1, 31G-2 and pixel 31B are provided with an inter-element isolation section as a physical separation section, similar to pixel 31R.

[0104] Figure 38 shows an example of the configuration of a pixel 31 in the third region 303 of Figure 35. Since the third region 303 is a peripheral region, each pixel 31 in the third region 303 is arranged such that the microlens 43 is eccentrically positioned toward the center (downward in the figure). As shown in Figure 38, in the pixel 31R, the protrusions from the inter-pixel separation portion 311 extend toward the center of the pixel 31R, thereby providing inter-element separation portions 331A to 331D. The inter-element separation portions 331A to 331D are provided as physical separation portions by embedding material in trenches formed in the semiconductor substrate 41.

[0105] Here, comparing the inter-element separation sections 331A to 331D of pixel 31R in Figure 38 with the inter-element separation sections 311A ​​to 311D of pixel 31R in Figure 36, the inter-element separation sections 331A and 331C have the same amount of protrusion as the inter-element separation sections 311A ​​and 311C, but the inter-element separation sections 331B and 331D have a smaller amount of protrusion than the inter-element separation sections 311B and 311D. In other words, by making the amount of protrusion of the inter-element separation sections 331B and 331D smaller than the amount of protrusion of the inter-element separation sections 311A ​​and 311C, the separation performance in the second direction D12 is made lower than the separation performance in the first direction D11. The line width of the inter-element separation sections 331B and 331D is the same as the line width of the inter-element separation sections 311A ​​and 311C. In Figure 38, pixels 31G-1, 31G-2 and pixel 31B are provided with an inter-element isolation section as a physical separation section, similar to pixel 31R.

[0106] Figure 39 shows an example of the configuration of a pixel 31 in the fourth region 304 of Figure 35. Since the fourth region 304 is a peripheral region, each pixel 31 in the fourth region 304 has a structure in which the microlens 43 is eccentrically positioned toward the center (lower left side in the figure). As shown in Figure 39, in the pixel 31R, a projection from the inter-pixel separation portion 311 extends toward the center of the pixel 31R, thereby providing inter-element separation portions 341A to 341D. The inter-element separation portions 341A to 341D are provided as physical separation portions by embedding material in trenches formed in the semiconductor substrate 41.

[0107] Here, comparing the inter-element isolation sections 341A to 341D of pixel 31R in Figure 39 with the inter-element isolation sections 311A ​​to 311D of pixel 31R in Figure 36, the amount of protrusion of the inter-element isolation sections 341A to 341D is smaller than that of the inter-element isolation sections 311A ​​to 311D. In other words, by making the amount of protrusion of the inter-element isolation sections 341A to 341D smaller than that of the inter-element isolation sections 311A ​​to 311D, the separation performance in both the first direction D11 and the second direction D12 is reduced. The line width of the inter-element isolation sections 341A to 341D is the same as the line width of the inter-element isolation sections 311A ​​to 311D. In Figure 39, pixels 31G-1, 31G-2 and pixel 31B are provided with inter-element isolation sections as physical separation sections, similar to pixel 31R.

[0108] Although the configuration of the pixels 31 located in the region between the first region 301 and the second region 302 in Figure 35 is not shown, the pixels 31 in this region have a configuration in which the amount of protrusion of the inter-element separation portion changes according to the image height, thereby changing the separation performance in the first direction D11. Specifically, in the direction of the arrow indicated by the first direction D11 in Figure 35 (horizontal direction), the amount of protrusion of the vertical inter-element separation portion of the pixels 31 in this region gradually decreases, until the amount of protrusion of the vertical inter-element separation portion of the pixels 31 in the second region 302 is smallest.

[0109] Furthermore, in the pixels 31 located in the region between the first region 301 and the third region 303 in Figure 35, the amount of protrusion of the lateral element separation portion gradually decreases toward the direction of the arrow indicated by the second direction D12 (vertical direction), until the amount of protrusion of the lateral element separation portion of the pixels 31 in the third region 303 is smallest. In the pixels 31 located in the region between the first region 301 and the fourth region 304 in Figure 35, the amount of protrusion of the vertical and lateral element separation portions gradually decreases toward the upper right direction in the figure, until the amount of protrusion of the vertical and lateral element separation portions of the pixels 31 in the fourth region 304 is smallest.

[0110] Furthermore, in the light detection surface 12A of Figure 35, the pixels 31 in the peripheral region from the center to the left edge in the figure have a vertical element separation portion protrusion that is smaller than the horizontal element separation portion protrusion, similar to the pixels 31 in the second region 302. The pixels 31 in the peripheral region from the center to the bottom edge in the figure have a horizontal element separation portion protrusion that is smaller than the vertical element separation portion protrusion, similar to the pixels 31 in the third region 303. The pixels 31 in the peripheral regions (the four corner regions) from the center to the upper left edge, lower right edge, or lower left edge in the figure have vertical and horizontal element separation portion protrusions that are smaller than the element separation portion protrusions of the pixels 31 in the first region 301, similar to the pixels 31 in the fourth region 304. In Figures 36 to 39, the fact that four 2x2 photoelectric conversion units 51 are provided for one microlens 43 is indicated by dashed lines in the vertical and horizontal directions. In Figures 36 to 39, for explanatory purposes, the shape of the color filter 61 is shown as a recessed shape corresponding to the element separation section, but the actual shape is rectangular.

[0111] In this way, in the pixels 31 arranged in a two-dimensional array in the pixel array section 21, the amount of protrusion of the inter-element separation section is set to an amount corresponding to a predetermined image height in the first direction D11 and the second direction D12 from the optical axis of the pixel array section 21, so that the separation performance in the first direction D11 and the second direction D12 is independently reduced according to each image height. As a result, when a structure is adopted in which multiple photoelectric conversion sections are provided for one microlens, it becomes possible to individually optimize the degree and direction of pupil distance shift, making it possible to achieve both robustness against pupil distance shift and high autofocus performance.

[0112] Figure 40 illustrates a second example of a pixel configuration in which the separation performance is changed according to the image height. In Figure 40, parts corresponding to those in Figure 35 are denoted by the same reference numerals, and their explanations are omitted as appropriate.

[0113] In the light detection surface 12A of Figure 40, the central region is designated as the first region 401, the peripheral region at the right edge of the figure from the center is designated as the second region 402, the peripheral region at the upper edge of the figure from the center is designated as the third region 403, and the peripheral region at the upper right edge of the figure from the center is designated as the fourth region 404. The configuration of the pixels 31 in each region is shown in Figures 41 to 44.

[0114] Figure 41 shows an example of the configuration of a pixel 31 in the first region 401 of Figure 40. In Figure 41, inter-pixel separation sections 411 are formed in a grid pattern so as to surround each photoelectric conversion section of each pixel 31. In the pixel 31R, the photoelectric conversion section is separated into an inter-element separation section 411A that separates it in a first direction D11 and an inter-element separation section 411B that separates it in a second direction D12, thereby realizing a structure in which four 2x2 photoelectric conversion sections 51 are provided for one microlens 43. The line widths of the inter-element separation sections 411A and 411B are the same. The inter-element separation sections 411A and 411B are provided as physical separation sections by embedding material in trenches formed in the semiconductor substrate 41, similar to the inter-pixel separation section 411. The inter-element separation sections 411A and 411B may also be configured as part of the inter-pixel separation section 411. In Figure 41, pixels 31G-1, 31G-2 and pixel 31B are provided with an inter-element isolation section as a physical separation section, similar to pixel 31R.

[0115] Figure 42 shows an example of the configuration of pixels 31 in the second region 402 of Figure 40. Since the second region 402 is a peripheral region, each pixel 31 in the second region 402 is arranged such that the microlens 43 is eccentrically positioned toward the center (left side in the figure). As shown in Figure 42, in the pixel 31R, element isolation sections 421A and 421B are provided as physical isolation sections by embedding material in trenches formed in the semiconductor substrate 41.

[0116] Comparing the element-to-element isolation sections 421A and 421B in Figure 42 with the element-to-element isolation sections 411A and 411B in Figure 41, the element-to-element isolation section 421B has the same line width as the element-to-element isolation section 411B, but the element-to-element isolation section 421A has a narrower line width than the element-to-element isolation section 411A. In other words, by making the line width of the element-to-element isolation section 421A narrower than that of the element-to-element isolation section 421B, the separation performance in the first direction D11 is made lower than the separation performance in the second direction D12. In Figure 42, pixels 31G-1, 31G-2 and pixel 31B are provided with element-to-element isolation sections as physical separation sections, similar to pixel 31R.

[0117] Figure 43 shows an example of the configuration of a pixel 31 in the third region 403 of Figure 40. Since the third region 403 is a peripheral region, each pixel 31 in the third region 403 has a structure in which the microlens 43 is eccentrically positioned toward the center (downward in the figure). As shown in Figure 43, in the pixel 31R, element isolation sections 431A and 431B are provided as physical isolation sections by embedding material in trenches formed in the semiconductor substrate 41.

[0118] Comparing the inter-element separation sections 431A and 431B of pixel 31R in Figure 43 with the inter-element separation sections 411A and 411B in Figure 41, inter-element separation section 431A has the same line width as inter-element separation section 411A, but inter-element separation section 431B has a narrower line width than inter-element separation section 411B. In other words, by making the line width of inter-element separation section 431B narrower than that of inter-element separation section 431A, the separation performance in the second direction D12 is made lower than the separation performance in the first direction D11. In Figure 43, pixels 31G-1, 31G-2 and pixel 31B are provided with inter-element separation sections as physical separation sections, similar to pixel 31R.

[0119] Figure 44 shows an example of the configuration of pixels 31 in the fourth region 404 of Figure 40. Since the fourth region 404 is a peripheral region, each pixel 31 in the fourth region 404 has a structure in which the microlens 43 is eccentrically positioned toward the center (lower left side in the figure). As shown in Figure 44, in the pixel 31R, element isolation sections 441A and 441B are provided as physical isolation sections by embedding material in trenches formed in the semiconductor substrate 41.

[0120] Here, comparing the inter-element isolation sections 441A and 441B of pixel 31R in Figure 44 with the inter-element isolation sections 411A and 411B in Figure 41, the line widths of the inter-element isolation sections 441A and 441B are narrower than those of the inter-element isolation sections 411A and 411B. In other words, by making the line widths of the inter-element isolation sections 441A and 441B narrower than those of the inter-element isolation sections 411A and 411B, the separation performance in both the first direction D11 and the second direction D12 is reduced. In Figure 44, pixels 31G-1, 31G-2 and pixel 31B are provided with inter-element isolation sections as physical separation sections, similar to pixel 31R.

[0121] In Figure 40, for pixels 31 located in the region between the first region 401 and the second region 402, the line width of the vertical inter-element separation gradually narrows in the direction of the arrow indicated by the first direction D11 (horizontal direction), with the line width of the vertical inter-element separation of pixels 31 in the second region 402 being the narrowest. In Figure 40, for pixels 31 located in the region between the first region 401 and the third region 403, the line width of the horizontal inter-element separation gradually narrows in the direction of the arrow indicated by the second direction D12 (vertical direction), with the line width of the horizontal inter-element separation of pixels 31 in the third region 403 being the narrowest. In Figure 40, for pixels 31 located in the region between the first region 401 and the fourth region 404, the line widths of the vertical and horizontal inter-element separations gradually narrow towards the upper right direction in the figure, with the line widths of the vertical and horizontal inter-element separations of pixels 31 in the fourth region 404 being the narrowest.

[0122] Furthermore, in the light detection surface 12A of Figure 40, the pixels 31 in the peripheral region from the center to the left edge in the figure have a vertical element separation line width that is thinner than the horizontal element separation line width, similar to the pixels 31 in the second region 402. The pixels 31 in the peripheral region from the center to the bottom edge in the figure have a horizontal element separation line width that is thinner than the vertical element separation line width, similar to the pixels 31 in the third region 403. The pixels 31 in the peripheral regions (the four corner regions) from the center to the upper left edge, lower right edge, or lower left edge in the figure have vertical and horizontal element separation line widths that are thinner than the element separation line widths of the pixels 31 in the first region 401, similar to the pixels 31 in the fourth region 404. In Figures 41 to 44, the fact that four 2x2 photoelectric conversion units 51 are provided for one microlens 43 is indicated by dashed lines in the vertical and horizontal directions. In Figures 41 to 44, for explanatory purposes, the shape of the color filter 61 is shown divided into four sections according to the inter-element isolation section, but in reality, it is an undivided rectangle.

[0123] In this way, in the pixels 31 arranged in a two-dimensional array in the pixel array section 21, the line width of the element separation section is set to a line width corresponding to a predetermined image height in the first direction D11 and the second direction D12 from the optical axis of the pixel array section 21, so that the separation performance in the first direction D11 and the second direction D12 is independently reduced according to each image height. As a result, when a structure is adopted in which multiple photoelectric conversion sections are provided for one microlens, it becomes possible to individually optimize the degree and direction of pupil distance shift, making it possible to achieve both robustness against pupil distance shift and high autofocus performance.

[0124] Figure 45 illustrates a third example of a pixel configuration in which the separation performance is changed according to the image height. In Figure 45, parts corresponding to those in Figure 35 are denoted by the same reference numerals, and their explanations are omitted as appropriate.

[0125] In the light detection surface 12A of Figure 45, the central region is designated as the first region 501, the peripheral region at the right edge of the figure from the center is designated as the second region 502, the peripheral region at the upper edge of the figure from the center is designated as the third region 503, and the peripheral region at the upper right edge of the figure from the center is designated as the fourth region 504. The configuration of the pixels 31 in each region is shown in Figures 46 to 49.

[0126] Figure 46 shows an example of the configuration of a pixel 31 in the first region 501 of Figure 45. In Figure 46, in a cross-sectional view, the inner lens 131 formed in the intermediate layer between the microlens 43 and the semiconductor substrate 41 directs the light incident from the microlens 43 to the photoelectric conversion section separated into a first direction D11 and a second direction D12 via a waveguide. In pixel 31R, the inner lens 131 separates the photoelectric conversion section, which is composed of photodiodes (PDs), into a first direction D1 and a second direction D2, thereby realizing a structure in which four 2x2 photoelectric conversion sections 51 are provided for one microlens 43. In a plan view, the inner lens 131 is a circular lens, and its diameter is slightly smaller (almost the same size) than that of the microlens 43, and its center position coincides with the center position of the microlens 43. In Figure 46, pixels 31G-1, 31G-2 and pixel 31B are provided with a circular inner lens 131, similar to pixel 31R.

[0127] Figure 47 shows an example of the configuration of a pixel 31 in the second region 502 of Figure 45. Since the second region 502 is a peripheral region, each pixel 31 in the second region 502 has a structure in which the microlens 43 is eccentrically positioned toward the center (left side in the figure). As shown in Figure 47, in pixel 31R, the inner lens 131 consists of a vertically elongated elliptical lens in plan view and is eccentrically positioned toward the center (left side in the figure), but its center position is different from the center position of the microlens 43. Here, as explained with reference to Figures 18 and 19 above, by making the shape of the inner lens 131 a vertically elongated ellipse, the separation performance in the first direction D11 is made lower than the separation performance in the second direction D12. In Figure 47, pixels 31G-1, 31G-2 and pixel 31B are provided with a vertically elongated elliptical inner lens 131, similar to pixel 31R.

[0128] Figure 48 shows an example of the configuration of a pixel 31 in the third region 503 of Figure 45. Since the third region 503 is a peripheral region, each pixel 31 in the third region 503 has a structure in which the microlens 43 is eccentrically positioned toward the center (lower side in the figure). As shown in Figure 48, in pixel 31R, the inner lens 131 consists of a horizontally elongated elliptical lens in plan view and is eccentrically positioned toward the center (lower side in the figure), but its center position is different from the center position of the microlens 43. Here, by making the shape of the inner lens 131 a horizontally elongated ellipse, the separation performance in the second direction D12 is made lower than the separation performance in the first direction D11. In Figure 48, pixels 31G-1, 31G-2 and pixel 31B are provided with horizontally elongated elliptical inner lenses 131, similar to pixel 31R.

[0129] Figure 49 shows an example of the configuration of a pixel 31 in the fourth region 504 of Figure 45. Since the fourth region 504 is a peripheral region, each pixel 31 in the fourth region 504 has a structure in which the microlens 43 is eccentrically positioned toward the center (lower left side in the figure). As shown in Figure 49, in the pixel 31R, the inner lens 131 is a circular lens in plan view and is eccentrically positioned toward the center (lower left side in the figure), but its center position is different from the center position of the microlens 43. The diameter of the inner lens 131 is smaller than the diameter of the microlens 43 (in the example of Figure 49, the diameter of the inner lens 131 is set to be about 0.6 to 0.7 times the diameter of the microlens 43). Here, by making the shape of the inner lens 131 circular and making its diameter smaller than the diameter of the microlens 43, the separation performance in both the first direction D11 and the second direction D12 is reduced. In Figure 49, pixels 31G-1, 31G-2, and pixel 31B are provided with a circular inner lens 131 that is smaller in diameter than the microlens 43, similar to pixel 31R.

[0130] In Figure 45, for the pixels 31 located in the region between the first region 501 and the second region 502, the shape of the inner lens 131 gradually deforms from a circular shape to a vertically elongated ellipse in the direction of the arrow indicated by the first direction D11 (horizontal direction). In Figure 45, for the pixels 31 located in the region between the first region 501 and the third region 503, the shape of the inner lens 131 gradually deforms from a circular shape to a horizontally elongated ellipse in the direction of the arrow indicated by the second direction D12 (vertical direction). In Figure 45, for the pixels 31 located in the region between the first region 501 and the fourth region 504, the diameter of the circular inner lens 131 gradually decreases towards the upper right direction in the figure.

[0131] Furthermore, in the light detection surface 12A of Figure 45, the pixels 31 in the peripheral region from the center to the left edge in the figure are provided with vertically elongated elliptical inner lenses 131, similar to the pixels 31 in the second region 502. The pixels 31 in the peripheral region from the center to the bottom edge in the figure are provided with horizontally elongated elliptical inner lenses 131, similar to the pixels 31 in the third region 503. The pixels 31 in the peripheral regions (the four corner regions) from the center to the upper left edge, lower right edge, or lower left edge in the figure are provided with circular inner lenses 131 smaller than the diameter of the microlens 43, similar to the pixels 31 in the fourth region 504.

[0132] Although the explanation is omitted to avoid repetition, the planar configuration in Figure 47 corresponds to the planar configuration in Figure 18, and the cross-section in the planar configuration of Figure 47 corresponds to the cross-section in Figure 19. The planar configurations in Figures 46, 48, and 49 differ from the planar configuration in Figure 18 in the shape and size of the inner lens 131, but their cross-sections basically correspond to the cross-section in Figure 19. In the above explanation, the microlens 43 was described as a plano-convex lens having a flat surface and a convex surface in cross-sectional view, but a microlens 601 consisting of a plano-concave lens having a flat surface and a concave surface, as shown in Figure 50, may also be provided. Specifically, in Figures 46 to 49, a microlens 601 may be placed instead of the microlens 43 to intentionally blur the focus. In Figures 46 to 49, the fact that four 2x2 photoelectric conversion units 51 are provided for one microlens 43 is indicated by dashed lines in the vertical and horizontal directions.

[0133] In this way, in the pixels 31 arranged in a two-dimensional array in the pixel array section 21, the shape and size of the inner lens 131 are set to correspond to predetermined image heights in the first direction D11 and the second direction D12 from the optical axis of the pixel array section 21, so that the separation performance in the first direction D11 and the second direction D12 is independently reduced according to each image height. As a result, when a structure is adopted in which multiple photoelectric conversion units are provided for one microlens, it becomes possible to individually optimize the degree and direction of pupil distance shift, making it possible to achieve both robustness against pupil distance shift and high autofocus performance.

[0134] Figure 51 shows an example of pupil projection by the inner lens 131 in the second region 502 and the third region 503 in Figure 45. Figure 51A shows an example of pupil projection by the inner lens 131 provided on a pixel 31 located in the second region 502, that is, the peripheral region which is the right edge relative to the center. In Figure 51A, region A11 corresponds to region a11, region A12 corresponds to region a12, region A13 corresponds to region a13, and region A14 corresponds to region a14, and regions A11 to A14 are projected as regions a11 to a14 by the inner lens 131 which has a vertically elongated elliptical shape. In the second region 502, when the photoelectric conversion unit is separated left and right, the light focusing is weaker than when it is separated up and down, and when the photoelectric conversion unit is separated up and down, the light focusing is stronger than when it is separated left and right. Therefore, depending on the strength of the light focusing, the separation performance in the first direction D11 can be made lower than the separation performance in the second direction D12. In this way, the focusing is smoothed out according to the image height in the first direction D11 and the second direction D12. As shown by the width A1 corresponding to the horizontal direction, when the pupil projection is misaligned between the left and right regions, such as between regions a11 and a12, or between regions a13 and a14, performance degradation can be suppressed and performance can be improved. Also, as shown by the width A2 corresponding to the vertical direction, when the pupil projection is not misaligned between the upper and lower regions, such as between regions a11 and a13, or between regions a12 and a14, performance degradation can be suppressed.

[0135] Figure 51B shows an example of pupil projection by an inner lens 131 provided on a pixel 31 located in the third region 503, that is, the peripheral region which is the upper edge relative to the center. In Figure 51B, regions B11 and B11, B12 and B12, B13 and B13, and B14 and B14 correspond to each other, and regions B11 to B14 are projected as regions b11 to b14 by the horizontally elongated elliptical inner lens 131. In the third region 503, when the photoelectric conversion unit is separated vertically, the light focusing is weaker than when it is separated horizontally, and when the photoelectric conversion unit is separated horizontally, the light focusing is stronger than when it is separated vertically. Therefore, depending on the strength of the light focusing, the separation performance in the second direction D12 can be made lower than the separation performance in the first direction D11. In this way, the focusing is smoothed out according to the image height in the first direction D11 and the second direction D12. As shown by the width B2 corresponding to the vertical direction, when the pupil projection is misaligned in the upper and lower regions, such as between regions b11 and b13, or between regions b12 and b14, performance degradation can be suppressed and performance can be improved. Also, as shown by the width B1 corresponding to the horizontal direction, when the pupil projection is not misaligned in the left and right regions, such as between regions b11 and b12, or between regions b13 and b14, performance degradation can be suppressed.

[0136] Here, for comparison, an example of pupil projection using the conventional technology will be explained with reference to Figures 52 and 53. Figure 52 shows the shape of the inner lens 711 in each region of the light detection surface 12A, where the central region is designated as the first region 701 and the peripheral region at the right edge from the center in the figure is designated as the second region 702. In other words, in the conventional technology, the diameter of the inner lens 711 is made smaller as the image height increases, intentionally weakening the light focusing to suppress performance degradation when the pupil distance is mismatched. However, while performance improves when there is a mismatch, there is a risk that performance will decrease when there is a match. Furthermore, in a structure in which multiple (for example, four in a 2x2 arrangement) photoelectric conversion units (PDs) are provided for a single microlens 43, there is also the problem that optimization has not been achieved in the image height direction and the PD division direction.

[0137] Figure 53A shows an example of pupil projection by an inner lens 711 provided on a pixel located in the first region 701 of Figure 52. In Figure 53A, regions A21 and a21, A22 and a22, A23 and a23, and A24 and a24 correspond to each other, and regions A21 to A24 are projected as regions a21 to a24 by the inner lens 711. In the first region 701, the light focusing is not weakened, and as shown by the width A3 corresponding to the horizontal direction and the width A4 corresponding to the vertical direction, there is no performance degradation when the pupil projection is not misaligned between the left and right regions and the top and bottom regions.

[0138] Figure 53B shows an example of pupil projection by an inner lens 711 provided on a pixel located in the second region 702 of Figure 52. In Figure 53B, regions B21 and B21, B22 and B22, B23 and B23, and B24 and B24 correspond to each other, and regions B21 to B24 are projected as regions b21 to b24 by the inner lens 711. In the second region 702, the diameter of the inner lens 711 is reduced to intentionally weaken the light focusing, so that performance degradation is suppressed and performance is improved when pupil projection is misaligned in the left and right regions, as shown by the width B3 corresponding to the horizontal direction. On the other hand, as shown by the width B4 corresponding to the vertical direction, when pupil projection is not misaligned in the upper and lower regions, performance degrades because the light focusing is intentionally weakened. In contrast, in this disclosure, as shown in Figure 51A, performance degradation can be suppressed when pupil projection is not misaligned in the upper and lower regions.

[0139] In this specification, the pupil separation performance of the photographic lens 11 can also be interpreted as the pupil division performance of the photographic lens 11, so separation performance (pupil separation performance) may be read as division performance (pupil division performance). Also, a pixel in which multiple photoelectric conversion units are formed on a single microlens may be interpreted as a pixel unit in which multiple pixels are formed on a single microlens. For example, a pixel composed of four photoelectric conversion units separated in a first direction and a second direction on a single microlens can also be said to be a pixel unit composed of four pixels separated in a first direction and a second direction on a single microlens.

[0140] <Examples of Use of the Light Detection Device> Figure 1 shows a configuration in which a light detection unit 12, as a light detection device to which the present disclosure is applied, is mounted on an imaging device 1. However, the light detection device to which the present disclosure is applied is not limited to imaging devices such as digital cameras, surveillance cameras, and in-vehicle cameras, but can also be mounted on electronic devices including, for example, smartphones, tablet terminals, and mobile phones. Note that imaging devices such as digital cameras are included in electronic devices. Furthermore, the light detection device to which the present disclosure is applied can be used in various cases where light such as visible light, infrared light, ultraviolet light, and X-rays is sensed. In other words, the light detection device to which the present disclosure is applied can be used not only in the field of appreciation, where images for appreciation are taken, but also in devices used in fields such as transportation, home appliances, medical and healthcare, security, beauty, sports, or agriculture.

[0141] Specifically, in the field of appreciation, for example, a light detection device to which this disclosure is applied can be used in devices for capturing images for appreciation, such as digital cameras, smartphones, and mobile phones with camera functions. In the field of traffic, for example, a light detection device to which this disclosure is applied can be used in devices used for traffic, such as in-vehicle sensors that capture images of the front, rear, surroundings, and interior of a vehicle for safe driving such as automatic stopping, or for recognizing the driver's condition, surveillance cameras that monitor moving vehicles and roads, and distance measuring sensors that measure distances between vehicles.

[0142] In the field of home appliances, for example, an optical detection device to which this disclosure is applied can be used in devices used in home appliances such as television sets, refrigerators, and air conditioners to capture user gestures and operate the device according to those gestures. In the field of medical and healthcare, for example, an optical detection device to which this disclosure is applied can be used in devices used for medical and healthcare purposes, such as endoscopes and devices that perform angiography by receiving infrared light. In the field of security, for example, an optical detection device to which this disclosure is applied can be used in devices used for security purposes, such as surveillance cameras for crime prevention and cameras for person authentication.

[0143] In the field of beauty, for example, a photodetector applying this disclosure can be used in devices used for beauty purposes, such as skin measuring devices for photographing skin or microscopes for photographing the scalp. In the field of sports, for example, a photodetector applying this disclosure can be used in devices used for sports purposes, such as action cameras and wearable cameras for sports use. In the field of agriculture, for example, a photodetector applying this disclosure can be used in devices used for agricultural purposes, such as cameras for monitoring the condition of fields and crops.

[0144] The embodiments described herein are not limited to those described above, and various modifications are possible without departing from the spirit of this disclosure. For example, the embodiments described above may be implemented individually or in combination with other embodiments. Furthermore, the effects described herein are merely illustrative and not limiting, and other effects may also occur.

[0145] Furthermore, this disclosure can take the following form.

[0146] (1) A photodetector comprising a pixel array portion in which a plurality of pixels, including a pixel having a plurality of photoelectric conversion units formed on a single microlens, are arranged in a two-dimensional array, wherein the plurality of photoelectric conversion units of the pixels are separated in a plan view in at least one of a first direction and a second direction different from the first direction, and the separation performance in the first direction is lower than the separation performance in the second direction. (2) The photodetector according to (1), wherein the second direction is a direction orthogonal to the first direction. (3) The photodetector according to (1) or (2), wherein the separation direction and separation performance of the first direction and the second direction are determined according to the aspect ratio of the photodetector surface of the pixel array portion. (4) The photodetector according to (3), wherein the plurality of pixels in the pixel array portion are arranged such that the photodetector surface is horizontally elongated, the first direction is a direction that separates the plurality of photoelectric conversion units left and right, and the second direction is a direction that separates the plurality of photoelectric conversion units up and down. (5) The photodetector according to (3), wherein the pixel array portion is configured such that the photodetector surface is horizontally elongated, and the pixels include a light-shielding portion that shields a predetermined area adjusted to correspond to the second direction, the first direction is the direction that separates the plurality of photoelectric conversion units vertically, and the second direction is the direction that separates the plurality of photoelectric conversion units horizontally. (6) The photodetector according to any one of (2) to (4), wherein the pixels include a pixel configured with four photoelectric conversion units separated in the first direction and the second direction for one microlens. (7) The photodetector according to any one of (2) to (4), wherein the pixels include a pixel configured with two photoelectric conversion units separated in the first direction for one microlens, and a pixel configured with two photoelectric conversion units separated in the second direction for one microlens. (8) The light detection device according to any one of (2) to (4), wherein the pixel includes a light-shielding portion that shields a region corresponding to the first direction or the second direction.(9) An optical detection device according to any one of (2) to (8), wherein an inter-element isolation section is formed on a semiconductor substrate on which the plurality of photoelectric conversion sections are formed, and the inter-element isolation section separates the plurality of photoelectric conversion sections in a first direction and a second direction. (10) An optical detection device according to (9), wherein the inter-element isolation section is constructed by embedding a material in a trench formed in the semiconductor substrate. (11) An optical detection device according to (10), wherein the inter-element isolation section has a different width in the first direction and the second direction in a plan view. (12) An optical detection device according to (10), wherein the inter-element isolation section is not formed in the first direction but is formed in the second direction. (13) An optical detection device according to (10), wherein the inter-element isolation section has a shape that protrudes toward the center of the pixel, and the amount of protrusion in the first direction is smaller than the amount of protrusion in the second direction. (14) The photodetector according to any one of (10) to (13), wherein an inter-pixel separation section for separating the plurality of pixels is formed on the semiconductor substrate, and the inter-element separation section has the same configuration as the inter-pixel separation section or a different configuration from the inter-pixel separation section. (15) The photodetector according to any one of (9) to (14), wherein an inner lens is formed in the intermediate layer between the microlens and the semiconductor substrate. (16) The photodetector according to (1) or (2), wherein an inner lens is formed in the intermediate layer between the microlens and the semiconductor substrate on which the plurality of photoelectric conversion sections are formed, and the inner lens separates the plurality of photoelectric conversion sections in the first direction and the second direction. (17) The photodetector according to (16), wherein the inner lens is an elliptical lens or a cylindrical lens. (18) The photodetector according to (9), wherein the inter-element separation section is formed by injecting impurities into the semiconductor substrate. (19) The photodetector according to (18), wherein the element separation portion has different widths in the first direction and the second direction when viewed from above.(20) An electronic device comprising: a photographic lens; a pixel array section having a plurality of pixels arranged in a two-dimensional array, to which light from the photographic lens is incident and which includes a plurality of pixels in which a plurality of photoelectric conversion units are formed for one microlens; and a control section that performs autofocus control based on phase difference information obtained from the photodetector section, wherein the plurality of photoelectric conversion units of the pixels are separated in a plan view into at least one of a first direction and a second direction different from the first direction, and the separation performance in the first direction is lower than the separation performance in the second direction. (21) An electronic device comprising: a pixel array section having a plurality of pixels arranged in a two-dimensional array, to which a plurality of pixels in which a plurality of photoelectric conversion units are formed for one microlens, wherein the plurality of photoelectric conversion units of the pixels are separated in a plan view into a first direction and a second direction different from the first direction, and the separation performance in the first direction and the second direction of the pixels arranged at a predetermined image height in the first direction and the second direction from the center of the pixel array section is lower independently according to the respective image height. (22) The photodetector according to (21), wherein the second direction is perpendicular to the first direction, and the pixels arranged at a predetermined image height in the first direction from the center have a separation performance in the first direction that is lower than the separation performance in the second direction, depending on the image height of the pixels. (23) The photodetector according to (22), wherein an inter-element isolation section is formed on a semiconductor substrate on which the plurality of photoelectric conversion sections are formed, and the inter-element isolation section separates the plurality of photoelectric conversion sections in a first direction and a second direction according to the image height of the pixels. (24) The photodetector according to (23), wherein the inter-element isolation section is formed by embedding material in a trench formed in the semiconductor substrate. (25) The photodetector according to (22), wherein an inner lens is formed in an intermediate layer between the microlens and the semiconductor substrate on which the plurality of photoelectric conversion units are formed, and the inner lens separates the plurality of photoelectric conversion units in a first direction and a second direction according to the image height of the pixels.

[0147] 1 Imaging device, 11 Imaging lens, 12 Light detection unit (light detection device), 12A Light detection surface, 13 Signal processing unit, 14 Compression / decompression unit, 15 Control unit, 16 Operation unit, 17 Display unit, 18 Storage unit, 21 Pixel array unit, 22 Vertical drive unit, 23 Column signal processing unit, 24 Horizontal drive unit, 25 Output unit, 26 Control unit, 31, 31R, 31G, 31B Pixel, 32 Pixel drive line, 33 Vertical signal line, 34 Horizontal signal line, 41 Semiconductor substrate, 42 Intermediate layer, 43 Microlens, 51 Photoelectric conversion unit, 61R, 61G, 61B Color filter, 81 Light shielding unit, 111 Inter-pixel separation unit, 111A, 111B Inter-element separation unit, 112A, 112B Inter-element isolation section, 112, 113 Inter-pixel light shielding section, 121 Waveguide, 131 Inner lens, 211 Photoelectric conversion section, 212 Inter-pixel isolation section, 212A, 212B Inter-element isolation section, 311 Inter-pixel isolation section, 311A-311D, 321A-321D, 331A-331D, 341A-341D Inter-element isolation section, 411 Inter-pixel isolation section, 411A, 411B, 421A, 421B, 431A, 431B, 441A, 441B Inter-element isolation section, 601 Microlens

Claims

1. A photodetector comprising a pixel array section in which a plurality of pixels, each containing a pixel with a plurality of photoelectric conversion sections formed on a single microlens, are arranged in a two-dimensional array, wherein the plurality of photoelectric conversion sections of the pixels are separated in a plan view in at least one of a first direction and a second direction different from the first direction, and the separation performance in the first direction is lower than the separation performance in the second direction.

2. The light detection device according to claim 1, wherein the second direction is a direction perpendicular to the first direction.

3. The photodetector according to claim 1, wherein the separation direction and separation performance of the first and second directions are determined according to the aspect ratio of the photodetector surface of the pixel array.

4. The photodetector according to claim 3, wherein the pixel array portion is arranged such that the photodetector surface is horizontally elongated, the first direction is the direction that separates the plurality of photoelectric conversion units to the left and right, and the second direction is the direction that separates the plurality of photoelectric conversion units to the up and down.

5. The photodetector according to claim 3, wherein the pixel array portion is configured such that the plurality of pixels are arranged so that the light detection surface is horizontally elongated, and the pixels include a light-shielding portion that shields a predetermined area adjusted to correspond to the second direction, the first direction is the direction that separates the plurality of photoelectric conversion portions vertically, and the second direction is the direction that separates the plurality of photoelectric conversion portions horizontally.

6. The photodetector according to claim 2, wherein the pixel comprises four photoelectric conversion units separated in the first and second directions with respect to one microlens.

7. The photodetector according to claim 2, wherein the pixel comprises a pixel comprising two photoelectric conversion units separated in the first direction with respect to one microlens, and a pixel comprising two photoelectric conversion units separated in the second direction with respect to one microlens.

8. The light detection device according to claim 2, wherein the pixel includes a light-shielding portion that shields a region corresponding to the first direction or the second direction.

9. The photodetector according to claim 2, wherein an inter-element isolation section is formed on a semiconductor substrate on which the plurality of photoelectric conversion sections are formed, and the inter-element isolation section separates the plurality of photoelectric conversion sections in the first direction and the second direction.

10. The photodetector according to claim 9, wherein the element isolation section is configured by embedding a material in a trench formed in the semiconductor substrate.

11. The light detection device according to claim 10, wherein the element separation portion has different widths in the first direction and the second direction when viewed in plan.

12. The photodetector according to claim 10, wherein the element separation portion is not formed in the first direction but is formed in the second direction.

13. The optical detection device according to claim 10, wherein the element separation portion has a shape that protrudes toward the center of the pixel, and the amount of protrusion in the first direction is smaller than the amount of protrusion in the second direction.

14. The photodetector according to claim 10, wherein an inter-pixel separation section for separating the plurality of pixels is formed on the semiconductor substrate, and the inter-element separation section has the same configuration as the inter-pixel separation section or a different configuration from the inter-pixel separation section.

15. The photodetector according to claim 9, wherein an inner lens is formed in the intermediate layer between the microlens and the semiconductor substrate.

16. The photodetector according to claim 2, wherein an inner lens is formed in an intermediate layer between the microlens and the semiconductor substrate on which the plurality of photoelectric conversion units are formed, and the inner lens separates the plurality of photoelectric conversion units in the first direction and the second direction.

17. The light detection device according to claim 16, wherein the inner lens is an elliptical lens or a cylindrical lens.

18. The photodetector according to claim 9, wherein the element isolation portion is formed by injecting impurities into the semiconductor substrate.

19. The light detection device according to claim 18, wherein the element separation portion has different widths in the first direction and the second direction when viewed in plan.

20. An electronic device comprising: a photographic lens; a pixel array section having a plurality of pixels arranged in a two-dimensional array, including pixels into which light from the photographic lens is incident and into which a plurality of photoelectric conversion units are formed for a single microlens; and a control section that performs autofocus control based on phase difference information obtained from the photoelectric detection section, wherein the plurality of photoelectric conversion units of the pixels are separated in a plan view in at least one of a first direction and a second direction different from the first direction, and the separation performance in the first direction is lower than the separation performance in the second direction.

21. A light detection device comprising a pixel array section in which a plurality of pixels, each containing a pixel with a plurality of photoelectric conversion units formed on a single microlens, are arranged in a two-dimensional array, wherein the plurality of photoelectric conversion units in the pixels are separated in a first direction and a second direction different from the first direction in a plan view, and the pixels, which are positioned at predetermined image heights in the first and second directions from the center of the pixel array section, have separation performance in the first and second directions that decreases independently according to their respective image heights.

22. The photodetector according to claim 21, wherein the second direction is perpendicular to the first direction, and the pixels arranged at a predetermined image height in the first direction from the center have a separation performance in the first direction that is lower than the separation performance in the second direction, depending on the image height of the pixels.

23. The photodetector according to claim 22, wherein an inter-element isolation section is formed on a semiconductor substrate on which the plurality of photoelectric conversion sections are formed, and the inter-element isolation section separates the plurality of photoelectric conversion sections in a first direction and a second direction according to the image height of the pixels.

24. The photodetector according to claim 23, wherein the element isolation portion is configured by embedding a material in a trench formed in the semiconductor substrate.

25. The photodetector according to claim 22, wherein an inner lens is formed in an intermediate layer between the microlens and the semiconductor substrate on which the plurality of photoelectric conversion units are formed, and the inner lens separates the plurality of photoelectric conversion units in a first direction and a second direction according to the image height of the pixels.

Images