High dynamic range CMOS image sensor and preparation method thereof

Abstract

According to the high-dynamic-range CMOS image sensor and the preparation method thereof provided by the invention, the first type of pixels and the normal second type of pixels are arranged, and the second type of pixels can ensure that high-brightness image signals are not overexposed by shortening the exposure time; for the first type of pixels, the sensitivity of the first type of pixels can be effectively reduced due to the arrangement of the metal anti-reflection layer in a central symmetry pattern, and for an image signal with a relatively long stroboscopic period, the sensitivity of the first type of pixels can be effectively reduced by controlling the exposure time of the transfer gate while the first type of pixels are ensured not to be overexposed. In addition, it can be guaranteed that the exposure time covers the image signal with the relatively long stroboscopic period, so that LFM is achieved, and due to the fact that no micro-lens is arranged on the first type of pixels, the sensitivity difference between the different first type of pixels can be effectively reduced; moreover, when a part of the image light coming from the pixels adjacent to the first type of pixels is shielded by the metal anti-reflection layer, the crosstalk of the adjacent pixels on the first type of pixels with weak light sensing can be effectively reduced, and the dazzle light phenomenon during imaging is slowed down.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor manufacturing, and in particular to a high dynamic range CMOS image sensor and a preparation method thereof. Background Art

[0002] With the development of chip integration, CMOS sensors are evolving towards smaller size and lower power consumption, and are increasingly being used in monitoring, automotive, and artificial intelligence fields. At the same time, the requirements for certain performance of CMOS image sensors in different application scenarios are also becoming increasingly higher.

[0003] In the field of vehicle monitoring, some image sensors need to monitor scenes outside the vehicle, which involves traffic lights. Usually, traffic lights are a kind of stroboscopic LED, and the chip needs to have a long enough exposure time to cover the LED's stroboscopic cycle to capture the image information of the LED "bright". However, the scene outside the vehicle will contain some high-brightness signals, such as the bright light at the tunnel exit, the sun in the sky, and the headlights of nearby cars. In order to prevent some high-brightness signals from being lost in the monitoring screen, the image sensor needs a relatively short exposure time, which will conflict with the requirement of a long enough exposure time to capture the flashing LED information. Therefore, the chip needs a LFM (Light Flicker Mitigation), that is, an LED flicker suppression function, so that the high-brightness scene is not overexposed while capturing the signal of the LED with a relatively long stroboscopic cycle.

[0004] In the prior art, LFM is achieved by adding anti-reflection materials to a portion of normal pixels, reducing the transmittance while controlling the exposure time of these pixels to cover the LED strobe cycle. This method requires an additional layer of filling anti-reflection material after the optical isolation structure between pixels is completed, which increases the cost and process complexity. In addition, the anti-crosstalk ability of this type of pixel is relatively weak. Summary of the invention

[0005] In view of the above-mentioned shortcomings of the prior art, the object of the present invention is to provide a high dynamic range CMOS image sensor and a preparation method thereof, which are used to solve the problems that the CMOS image sensor in the prior art realizes LFM by adding anti-reflection materials on a part of normal pixels, reducing the transmittance and controlling the exposure time of these pixels to cover the LED stroboscopic cycle, such as complex process, high cost and weak anti-crosstalk ability.

[0006] To achieve the above-mentioned object and other related objects, the present invention provides a high dynamic range CMOS image sensor, the image sensor comprising: a first type of pixel and a second type of pixel, wherein the first type of pixel and the second type of pixel both comprise a circuit connection layer, a semiconductor structure layer and an optical structure layer;

[0007] The first type of pixel and the second type of pixel both include a color filter, a photodiode and a transfer gate from top to bottom; wherein the color filter is arranged in the optical structure layer, the photodiode is arranged in the semiconductor structure layer, and the transfer gate is arranged in the circuit connection layer;

[0008] The first type of pixel further includes a metal anti-reflection layer disposed in the optical structure layer and below the color filter, and the cross-sectional pattern of the metal anti-reflection layer is a centrally symmetrical hollow pattern, and the metal anti-reflection layer is used to block a portion of the image light from entering the corresponding photodiode;

[0009] The second type of pixel further includes a microlens disposed in the optical structure layer and located above the color filter.

[0010] Optionally, the metal anti-reflection layer is configured as a metal grid, and a grid width at the periphery of the metal grid is greater than a grid width inside the metal grid.

[0011] Optionally, the metal anti-reflection layer is configured as a metal grid, and a grid depth at the periphery of the metal grid is greater than a grid depth inside the metal grid.

[0012] Furthermore, the metal anti-reflection layer is configured as a metal grid, and the shapes of the internal divisions of the metal grid include a circle, a tic-tac-toe shape, or a square.

[0013] Furthermore, the grid widths inside the metal grid are the same.

[0014] Optionally, a cross-sectional size of the first type of pixels is not greater than a cross-sectional size of the second type of pixels.

[0015] Further, the image sensor includes a pixel array consisting of a plurality of first Bayer array pixel unit groups and a plurality of second Bayer array pixel unit groups; each of the first Bayer array pixel unit groups includes four second-type pixels arranged in a 2×2 form; each of the second Bayer array pixel unit groups includes four first-type pixels arranged in a 2×2 form, wherein the number of the first-type pixels corresponds to the number of the second-type pixels, and the first-type pixels are formed at the center of any four mutually adjacent second-type pixels.

[0016] Further, the image sensor includes a pixel array composed of multiple quad-Bayer array pixel unit groups; each of the quad-Bayer array pixel unit groups includes four RGGB pixel units arranged in a 2×2 form; each of the pixel units includes four pixels of the same color arranged in a 2×2 form, wherein three pixels are set as the second type of pixels, one pixel is set as the first type of pixel, and all the first type of pixels are set at the same position in the pixel unit where they are located.

[0017] Further, the image sensor includes a pixel array consisting of a plurality of four Bayer array pixel unit groups and a plurality of Bayer array pixel unit groups; each of the four Bayer array pixel unit groups includes four RGGB pixel units arranged in a 2×2 form, and each of the pixel units includes four second-type pixels of the same color arranged in a 2×2 form; each of the Bayer array pixel unit groups includes four RGGB first-type pixels arranged in a 2×2 form, wherein the first-type pixel is formed at the center position of the pixel unit where the second-type pixel is located and the color of its color filter is consistent with the color of the color filter of the second-type pixel in the pixel unit where it is located.

[0018] The present invention also provides a method for preparing a high dynamic range CMOS image sensor, the method comprising:

[0019] Providing a semiconductor structure layer, and forming a photodiode in the semiconductor structure layer;

[0020] A circuit connection layer is formed on the semiconductor structure layer, wherein transfer gates corresponding to the photodiodes are formed in the circuit connection layer;

[0021] forming a first light-transmitting medium layer and a metal anti-reflection material layer in sequence on the semiconductor structure layer;

[0022] The metal anti-reflection material layer is patterned to obtain a metal anti-reflection layer and an optical isolation structure; wherein the area where the metal anti-reflection layer is located corresponds to the first type of pixels, and the area where the optical isolation structure is located corresponds to the second type of pixels, and the cross-sectional pattern of the metal anti-reflection layer is a centrally symmetrical pattern to block a part of the image light from entering the corresponding photodiode; and the optical isolation structure is used to separate the image light between two adjacent pixels;

[0023] Filling the metal anti-reflection layer with a second light-transmitting medium layer so that the surface of the metal anti-reflection layer is flush;

[0024] A color filter and a microlens are sequentially formed on the metal anti-reflection layer and the second light-transmitting medium layer, wherein the microlens is only formed on the second type of pixels.

[0025] As described above, in the high dynamic range CMOS image sensor and preparation method thereof of the present invention, the second type of pixels can be understood as normal pixels of the image sensor, the image light is focused by the microlens thereon and filtered by the color filter before entering the corresponding photodiode for exposure, the light signal is converted into an electrical signal, and the exposure time is controlled by the corresponding transfer gate. For high-brightness image signals, overexposure can be ensured by shortening the exposure time; for the first type of pixels, when the image light enters the color filter in the optical structure layer, due to the provision of a metal anti-reflection layer, and the metal anti-reflection layer is in a centrally symmetrical pattern, a part of the image light will be uniformly reflected by the metal anti-reflection layer, that is, the image light obliquely incident in all directions can be uniformly reflected, so that the light intensity entering the photodiode is uniformly weakened, thereby reducing the sensitivity of the first type of pixels. For image signals with relatively long stroboscopic periods, since the sensitivity of the first type of pixels is relatively low, the exposure of the transfer gate can be controlled. While ensuring that the first-class pixels are not overexposed, the exposure time can also ensure that the image signals with relatively long stroboscopic periods are covered by the exposure time, thereby realizing LFM. Furthermore, since no microlens is set on the first-class pixels, the sensitivity difference between different first-class pixels can be effectively reduced. Specifically, when a microlens is set, as long as the light is slightly offset between different first-class pixels, the amount of image light signal passing through the metal anti-reflection layer at the center focus will be different, resulting in sensitivity difference. If no microlens is set, even if the light is slightly offset, since the light will not be focused and the metal anti-reflection layer has a centrally symmetrical structure, the light will be uniformly reflected by the metal anti-reflection layer, so that the light enters the photodiode uniformly, thereby ensuring that the sensitivity of all first-class pixels is consistent. Furthermore, when a part of the image light coming from the adjacent pixels of the first-class pixels is blocked by the metal anti-reflection layer, the crosstalk of the first-class pixels with weak light sensitivity from the adjacent pixels can be effectively reduced, thereby alleviating the glare phenomenon during imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A schematic cross-sectional structure diagram of an example of a high dynamic range CMOS image sensor of the present invention is shown, wherein the optical path of image light in the first type of pixels and the second type of pixels is illustrated.

[0027] Figure 2 Display as Figure 1 Optical path diagram of the first type of pixel in a high dynamic range CMOS image sensor that mitigates crosstalk from adjacent pixels.

[0028] Figure 3 Display as Figure 4 Schematic diagram of the cross-sectional structure along the AA direction.

[0029] Figures 4 to 9 Schematic diagrams showing several different geometric structures of the metal grid in the high dynamic range CMOS image sensor of the present invention.

[0030] Figures 10 to 12 Shown are schematic diagrams of three different pixel arrays in the high dynamic range CMOS image sensor of the present invention.

[0031] Figures 13 to 17 It is a schematic diagram of the cross-sectional structure after each step in the method for preparing a high dynamic range CMOS image sensor of the present invention is completed.

[0032] Component number description

[0033] 10 Circuit connection layer

[0034] 100 Transfer Gate

[0035] 11 Semiconductor structure layer

[0036] 110 Photodiode

[0037] 111 Isolation trench

[0038] 12 Optical structure layer

[0039] 120 Color Filters

[0040] 121 Metal anti-reflection layer

[0041] 122 Metal Grid

[0042] 123 Microlens

[0043] 125 first light-transmitting medium layer

[0044] 126 Metal anti-reflection material layer

[0045] 127 Optical Isolation Structure

[0046] 128 second light-transmitting medium layer

[0047] 13 First-class pixels

[0048] 14 Second type of pixels

[0049] 15 Pixel Array

[0050] 150 First Bayer array pixel unit group

[0051] 151 Second Bayer array pixel unit group

[0052] 152 four-bayer array pixel unit group

[0053] 153 pixel unit

[0054] 154 Bayer array pixel unit group

[0055] 16 Adhesive layer DETAILED DESCRIPTION

[0056] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0057] See also Figures 1 to 17 It should be noted that the illustrations provided in this embodiment are only used to illustrate the basic concept of the present invention in a schematic manner, and the illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0058] Embodiment 1

[0059] like Figures 1 to 3 As shown, this embodiment provides a high dynamic range CMOS image sensor, including: a first type of pixel 13 and a second type of pixel 14, and the first type of pixel 13 and the second type of pixel 14 both include a circuit connection layer 10, a semiconductor structure layer 11 and an optical structure layer 12;

[0060] The first type of pixel 13 and the second type of pixel 14 both include a color filter 120, a photodiode 110 and a transfer gate 100 from top to bottom; wherein the color filter 120 is disposed in the optical structure layer 12, the photodiode 110 is disposed in the semiconductor structure layer 11, and the transfer gate 100 is disposed in the circuit connection layer 10; it should be noted here that the order from top to bottom refers to the direction from the optical structure layer 12 to the semiconductor structure layer 11 and then to the circuit connection layer 10;

[0061] The first type of pixel 13 further includes a metal anti-reflection layer 121 disposed in the optical structure layer 12 and located below the color filter 120, and the cross-sectional pattern of the metal anti-reflection layer 121 is a centrally symmetrical hollow pattern, and the metal anti-reflection layer 121 is used to block a portion of the image light from entering the corresponding photodiode 110; wherein the cross-sectional pattern refers to a pattern looking downward from the optical structure layer 12 to the semiconductor structure layer 11 and then to the circuit connection layer 10;

[0062] The second type of pixel 14 further includes a microlens 123 disposed in the optical structure layer 12 and located above the color filter 120 .

[0063] like Figure 1As shown, the second-type pixel 14 can be understood as a normal pixel of the image sensor. The image light is focused by the microlens 123 thereon and filtered by the color filter 120 before entering the corresponding photodiode 110 for exposure, converting the optical signal into an electrical signal, and controlling its exposure time through the corresponding transfer gate 100. For a high-brightness image signal, overexposure can be ensured by shortening the exposure time. For the first-type pixel 13, after the image light enters the color filter 120 in the optical structure layer 12, due to the provision of a metal anti-reflection layer 121, and the metal anti-reflection layer 121 is in a centrally symmetrical pattern, a portion of the image light will be uniformly reflected by the metal anti-reflection layer 121, that is, the image light obliquely incident in all directions can be uniformly reflected, so that the light intensity entering the photodiode 110 is uniformly weakened, thereby reducing the sensitivity of the first-type pixel 13. For an image signal with a relatively long stroboscopic period, For example, since the sensitivity of the first-category pixel 13 is relatively low, the exposure time of the transfer gate 100 can be controlled to ensure that the first-category pixel 13 is not overexposed while ensuring that the exposure time covers the image signal with a relatively long stroboscopic period, thereby realizing LFM. Moreover, since no microlens is provided on the first-category pixel 13, the sensitivity difference between different first-category pixels 13 can be effectively reduced. Specifically, when the microlens is provided, as long as the light is slightly offset between different first-category pixels 13, the amount of image light signal passing through the metal anti-reflection layer at the center focus will be different, resulting in sensitivity difference. If the microlens is not provided, even if the light is slightly offset, since the light will not be focused and the metal anti-reflection layer is a centrally symmetrical structure, the light will be uniformly reflected by the metal anti-reflection layer, so that the light enters the photodiode uniformly, thereby ensuring that the sensitivity of all first-category pixels 13 is consistent. Furthermore, Figure 2 As shown, when the image light coming from the pixels adjacent to the first type of pixel 13, part of it will be blocked by the metal anti-reflection layer, thereby effectively reducing the crosstalk of the first type of pixel 13 with weak photosensitivity from the adjacent pixels and alleviating the glare phenomenon during imaging.

[0064] It should be noted that the imaging process of the CMOS image sensor is that the external image light is adjusted by the optical structure layer and then enters the photodiode of the semiconductor structure layer, converting the optical signal into an electrical signal, and then reading the electrical signal through the circuit connection layer and imaging it. Therefore, according to different application scenarios and different requirements, under the conditions based on the imaging process, the structures in the circuit connection layer, the semiconductor structure layer and the optical structure layer may be different, but this is a well-known technology in the field and will not be elaborated here.

[0065] like Figures 3 to 9 As shown, as an example, the metal anti-reflection layer 121 is preferably in the form of a metal grid 122. The metal grid 122 reflects light, and the hollow lattices surrounded by the grid allow light to pass normally.

[0066] like Figure 3 As shown, further, the grid depth D1 of the outer periphery of the metal grid 122 can be set greater than the grid depth D2 inside the metal grid to improve the shielding of light from adjacent pixels, further reduce the crosstalk of the weakly sensitive first-type pixels 13 from adjacent pixels, and alleviate the glare phenomenon during imaging.

[0067] like Figures 5 to 9 As shown, further, the grid width L1 of the outer periphery of the metal grid 122 can be set to be greater than the grid width L2 of the inner periphery of the metal grid 122, so as to improve the shielding of light from adjacent pixels, further reduce the crosstalk of the first type of pixels 13 with weak light sensitivity to adjacent pixels, and reduce the glare phenomenon during imaging. It should be noted here that the grid width L1 of the outer periphery of the metal grid 122 can be set to be greater than the grid width L2 of the inner periphery of the metal grid 122, but the size of the grid width L2 of the inner periphery of the metal grid 122 is not excessively limited. Its main function is to adjust the amount of light entering, so it can be set according to actual needs, such as Figures 4 to 7 As shown, for example, the width of L2 can be the same as L1 or smaller than L1. In addition, the internal grid width of the metal grid 122 in each first-type pixel 13 can be the same or different. Preferably, Figures 4 to 7 As shown, the internal grid width L2 of the metal grid 122 in each first-type pixel 13 is selected to be the same.

[0068] It should also be noted that, from the perspective of reducing glare during imaging, the grid depth D1 of the outer periphery of the metal grid 122 can be set greater than the grid depth D2 of the inner periphery of the metal grid, or the grid width L1 of the outer periphery of the metal grid 122 can be set greater than the grid width L2 of the inner periphery of the metal grid 122; or the grid depth D1 of the outer periphery of the metal grid 122 can be made greater than the grid depth D2 of the inner periphery of the metal grid and the grid width L1 of the outer periphery of the metal grid 122 can be made greater than the grid width L2 of the inner periphery of the metal grid 122. The specific selection is made according to actual needs.

[0069] As an example, the shape of the internal division of the metal grid 122 is not too limited, as long as its geometric shape layer is centrally symmetrical, for example, it can be a circle (such as Fig. 9 As shown), Tic-Tac-toe (as shown Figure 8 as shown) or square (as Figure 7 As shown), etc., wherein the shape of the internal division of the tic-tac-toe pattern is obtained by rotating the shape of the internal division of the square by 45°.

[0070] As another example, the cross-sectional size of the first-type pixel 13 is set to be smaller than or equal to the cross-sectional size of the second-type pixel 14 .

[0071] As is known, pixels in an image sensor are generally arranged in an array, and the following illustrates several ways of arranging the first type of pixels and the second type of pixels in the pixel array of the image sensor in this embodiment. However, the arrangement of the first type of pixels and the second type of pixels in the pixel array can be performed in other known arrangements in the art according to actual needs, so the following examples cannot limit the protection scope of this application.

[0072] The first example, such as Fig.10 As shown, the image sensor includes a pixel array 15 composed of a plurality of first Bayer array pixel unit groups 150 and a plurality of second Bayer array pixel unit groups 151; each of the first Bayer array pixel unit groups 150 includes four second-type pixels 14 arranged in a 2×2 form; each of the second Bayer array pixel unit groups 151 includes four first-type pixels 13 arranged in a 2×2 form, wherein the number of the first-type pixels 13 corresponds to the number of the second-type pixels 14, and the first-type pixels 13 are formed at the center of any four mutually adjacent second-type pixels 14, and in this example, the cross-sectional size of the first-type pixels 13 is set to be smaller than the cross-sectional size of the second-type pixels 14.

[0073] It should be noted that, for ease of understanding, Fig.10 FIG. 1 shows only the pixel array 15 formed by four first Bayer array pixel unit groups 150 arranged in a 2×2 form and four second Bayer array pixel unit groups 151 arranged in a 2×2 form. However, those skilled in the art know that the scale of the pixel array 15 can be set according to actual conditions. Fig.10 The scale of the pixel array 15 limits the pixel array of the present application.

[0074] As an example, the four second-type pixels 14 in the first Bayer array pixel unit group 150 may be in an RGGB array mode, that is, the four second-type pixels 14 in the first Bayer array pixel unit group 150 are respectively red pixels, green pixels, green pixels and blue pixels, and may also be replaced by color filters for other colors, such as RYYB, RGBW, etc., or some pixels may not be provided with color filters to transmit all spectra of light, or may be replaced with infrared enhancement or infrared cutoff filters. Correspondingly, the four first-type pixels 13 in the second Bayer array pixel unit group 151 may also be in an RGGB array mode, that is, the four first-type pixels 13 in the second Bayer array pixel unit group 151 are respectively red pixels, green pixels, green pixels and blue pixels, and may also be replaced by color filters for other colors, such as RYYB, RGBW, etc., or some pixels may not be provided with color filters to transmit all spectra of light, or may be replaced with infrared enhancement or infrared cutoff filters.

[0075] As an example, the shape of the first type of pixel 13 when formed at the center of any four mutually adjacent second type of pixels 14 can be set arbitrarily, for example, Fig.10 The rhombus, rectangle, circle, trapezoid, arc, etc. can be set according to actual needs.

[0076] The second example, such as Fig.11 As shown, the image sensor includes a pixel array 15 composed of a plurality of four-Bayer array pixel unit groups 152; each of the four-Bayer array pixel unit groups 152 includes four RGGB pixel units 153 arranged in a 2×2 form; each of the pixel units 153 includes four pixels of the same color arranged in a 2×2 form, wherein three pixels are set as the second-type pixels 14, one pixel is set as the first-type pixel 13, and all the first-type pixels 13 are set at the same position in the pixel unit 153 where they are located, and in this example, the cross-sectional size of the first-type pixels 13 is set to be consistent with the cross-sectional size of the second-type pixels 14 and the shape is also the same. However, in practice, the cross-sectional size of the first-type pixels 13 may not be consistent with the cross-sectional size of the second-type pixels 14, and the shape of the first-type pixels 13 may not be the same as the shape of the second-type pixels 14, and the selection is made according to actual needs.

[0077] It should be noted that, for ease of understanding, Fig.11 FIG. 1 shows only a pixel array 15 composed of four quad-Bayer array pixel unit groups 152 arranged in a 2×2 form. However, those skilled in the art know that the size of the pixel array 15 can be set according to actual conditions. Fig.11 The scale of the pixel array 15 limits the pixel array of the present application.

[0078] The four pixel units 153 in the four-Bayer array pixel unit group 152 are in an RGGB array mode, that is, the four pixel units 153 in the four-Bayer array pixel unit group 152 are respectively a red pixel unit, a green pixel unit, a green pixel unit and a blue pixel unit, and accordingly, the red pixel unit corresponds to four red pixels, the green pixel unit corresponds to four green pixels, and the blue pixel unit corresponds to four blue pixels; in addition, the four pixel units 153 in the four-Bayer array pixel unit group 152 can also be replaced by color filters for other colors, such as RYYB, RGBW, etc., or some pixels may not be provided with color filters to transmit light of all spectra, or may be replaced with infrared enhancement or infrared cutoff filters.

[0079] The third example, such as Fig.12As shown, the image sensor includes a pixel array 15 composed of a plurality of four Bayer array pixel unit groups 152 and a plurality of Bayer array pixel unit groups 154; each of the four Bayer array pixel unit groups 152 includes four RGGB pixel units 153 arranged in a 2×2 form, and each of the pixel units 153 includes four second-type pixels 14 of the same color arranged in a 2×2 form; each of the Bayer array pixel unit groups 154 includes four RGGB first-type pixels 13 arranged in a 2×2 form, wherein the first-type pixel 13 is formed at the center position of the pixel unit 153 where the second-type pixel 14 is located and the color of its color filter is consistent with the color of the color filter of the second-type pixel 14 in the pixel unit 153 where it is located, and in this example, the cross-sectional size of the first-type pixel 13 is set to be smaller than the cross-sectional size of the second-type pixel 14.

[0080] It should be noted that, for ease of understanding, Fig.12 FIG. 1 shows only the pixel array 15 composed of four four-Bayer array pixel unit groups 152 arranged in a 2×2 form and four Bayer array pixel unit groups 154 arranged in a 2×2 form. However, those skilled in the art know that the scale of the pixel array 15 can be set according to actual conditions. Fig.12 The scale of the pixel array 15 limits the pixel array of the present application.

[0081] The four pixel units 153 in the four-Bayer array pixel unit group 152 are in an RGGB array mode, that is, the four pixel units 153 in the four-Bayer array pixel unit group 152 are respectively a red pixel unit, a green pixel unit, a green pixel unit and a blue pixel unit, and accordingly, the red pixel unit corresponds to four red pixels, the green pixel unit corresponds to four green pixels, and the blue pixel unit corresponds to four blue pixels; in addition, the four pixel units 153 in the four-Bayer array pixel unit group 152 can also be replaced by color filters for other colors, such as RYYB, RGBW, etc., or some pixels may not be provided with color filters to transmit light of all spectra, or may be replaced with infrared enhancement or infrared cutoff filters. The four first-type pixels 13 in the Bayer array pixel unit group 154 ​​are in the RGGB array mode, that is, the four first-type pixels 13 in the Bayer array pixel unit group 154 ​​are red pixel, green pixel, green pixel and blue pixel respectively, and can also be replaced by other colors by color filters, such as RYYB, RGBW, etc., and some pixels can also be not set with color filters and transmit light of all spectra, and can also be replaced with infrared enhancement or infrared cutoff filters.

[0082] As an example, the shape of the first type pixel 13 when formed at the center position of the pixel unit 153 where the second type pixel 14 is located can be set arbitrarily, for example, Fig.12 The rhombus, rectangle, circle, trapezoid, arc, etc. can be set according to actual needs.

[0083] Embodiment 2

[0084] This embodiment provides a method for preparing a high dynamic range CMOS image sensor, which can be used to prepare the high dynamic range CMOS image sensor of Embodiment 1, but is not limited thereto, and other preparation methods that can realize the image sensor structure of Embodiment 1 are also acceptable.

[0085] The preparation method of this embodiment is described in detail below with reference to the specific drawings.

[0086] like Fig.13 As shown, step S1 is first performed to provide a semiconductor structure layer 11 and form a photodiode 110 in the semiconductor structure layer 11 .

[0087] It should be noted here that the photodiodes 110 formed in the semiconductor structure layer 11 are arranged in an array. The photodiode 110 can be formed by an ion implantation process, which generally includes: providing a lightly doped semiconductor substrate, which can be a silicon substrate, or germanium, silicon germanium, silicon carbide, gallium arsenide, indium gallium, or a silicon substrate on an insulator surface, a germanium substrate on an insulator surface, or a substrate with an epitaxial layer grown thereon, and other suitable materials for image sensors; then forming a photodiode by implanting ions into the semiconductor substrate, and controlling the energy and concentration of the ion implantation to control the depth and range of the ion implantation. The photodiode 110 can also be an organic photodiode, which is formed by depositing on a substrate using an existing conventional deposition process, and the material of the organic photodiode uses an existing conventional organic material suitable for image sensors.

[0088] like Fig.13 As shown, step S2 is then performed to form a circuit connection layer 10 on the semiconductor structure layer 11 , and transfer gates 100 corresponding to the photodiodes 110 are formed in the circuit connection layer 10 .

[0089] The circuit connection layer 10 also includes other circuit structures, such as a source follower transistor, a row selection transistor, a floating diffusion, a reset transistor, etc., which are well-known structures in the art and will not be described in detail here.

[0090] like Fig.14As shown, step S3 is then performed to sequentially form a first light-transmitting medium layer 125 and a metal anti-reflective material layer 126 on the semiconductor structure layer 11 .

[0091] It should be noted that the circuit connection layer 10 and the first light-transmitting medium layer 125 are respectively formed on two opposite sides of the semiconductor structure layer 11. Before forming the first light-transmitting medium layer 125 and the metal anti-reflection material layer 126, an isolation groove 111 may be formed in the semiconductor structure layer 11. The isolation groove 111 is formed inside the semiconductor structure layer 11 on a side close to the first light-transmitting medium layer 125 (such as Fig.14 As shown), it may also be formed inside the side close to the circuit connection layer 10 (as shown Figures 1 to 3 shown).

[0092] like Fig.15 As shown, step S4 is then performed to pattern the metal anti-reflection material layer 126 to obtain a metal anti-reflection layer 121 and an optical isolation structure 127; wherein the area where the metal anti-reflection layer 121 is located corresponds to the first type of pixel 13, and the area where the optical isolation structure 127 is located corresponds to the second type of pixel 14, and the cross-sectional shape of the metal anti-reflection layer 121 is a centrally symmetrical shape to block a portion of the image light from entering the corresponding photodiode 110; the optical isolation structure 127 is used to separate the image light between two adjacent pixels.

[0093] As an example, the metal anti-reflective material layer 126 may be patterned by photolithography and etching processes.

[0094] like Fig.16 As shown, the next step S5 is to fill the metal anti-reflection layer 121 with the second light-transmitting medium layer 128 to make the surface of the metal anti-reflection layer 121 flush. Filling the surface of the metal anti-reflection layer 121 flat can improve the bonding quality of the subsequent bonding of the color filter.

[0095] Preferably, the first light-transmitting medium layer 125 and the second light-transmitting medium layer 128 are made of the same material, and the material is a light-transmitting material.

[0096] like Fig.17 As shown, step S6 is finally performed to sequentially form a color filter 120 and a microlens 123 on the metal anti-reflection layer 121 and the second light-transmitting medium layer 128 , wherein the microlens 123 is only formed on the second type of pixel 14 .

[0097] As an example, the microlens 123 may be bonded to the color filter 120 using a bonding layer 16 .

[0098] In summary, the present invention provides a high dynamic range CMOS image sensor and a preparation method thereof. The second type of pixels can be understood as normal pixels of the image sensor. The image light is focused by the microlens thereon and filtered by the color filter before entering the corresponding photodiode for exposure. The light signal is converted into an electrical signal, and its exposure time is controlled by the corresponding transfer gate. For high-brightness image signals, overexposure can be ensured by shortening the exposure time. For the first type of pixels, after the image light enters the color filter in the optical structure layer, a part of the image light will be uniformly reflected by the metal anti-reflection layer because a metal anti-reflection layer is provided and the metal anti-reflection layer is in a centrally symmetrical pattern. That is, the image light obliquely incident in all directions can be uniformly reflected, so that the light intensity entering the photodiode is uniformly weakened, thereby reducing the sensitivity of the first type of pixels. For image signals with relatively long stroboscopic periods, since the sensitivity of the first type of pixels is relatively low, the exposure time of the transfer gate can be controlled. While ensuring that the first-class pixels are not overexposed, the light time can also ensure that the exposure time covers the image signal with a relatively long stroboscopic period, thereby realizing LFM. Since no microlens is set on the first-class pixels, the sensitivity difference between different first-class pixels can be effectively reduced. Specifically, when the microlens is set, as long as the light is slightly offset between different first-class pixels, the amount of image light signal passing through the metal anti-reflection layer at the center focus will be different, resulting in sensitivity difference. If the microlens is not set, even if the light is slightly offset, since the light will not be focused and the metal anti-reflection layer is a central symmetrical structure, the light will be uniformly reflected by the metal anti-reflection layer, so that the light enters the photodiode uniformly, thereby ensuring that the sensitivity of all first-class pixels is consistent; furthermore, when the image light from the adjacent pixels of the first-class pixels is partially blocked by the metal anti-reflection layer, the crosstalk of the first-class pixels with weak light sensitivity to the adjacent pixels can be effectively reduced, and the glare phenomenon during imaging can be slowed down. Therefore, the present invention effectively overcomes various shortcomings in the prior art and has a high industrial utilization value.

[0099] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A high dynamic range CMOS image sensor, characterized in that: The image sensor comprises: a first type of pixel and a second type of pixel, and the first type of pixel and the second type of pixel both comprise a circuit connection layer, a semiconductor structure layer and an optical structure layer; The first type of pixel and the second type of pixel both include a color filter, a photodiode and a transfer gate from top to bottom; wherein the color filter is arranged in the optical structure layer, the photodiode is arranged in the semiconductor structure layer, and the transfer gate is arranged in the circuit connection layer; The first type of pixel further includes a metal anti-reflection layer disposed in the optical structure layer and below the color filter, and the cross-sectional pattern of the metal anti-reflection layer is a centrally symmetrical hollow pattern, and the metal anti-reflection layer is used to block a portion of the image light from entering the corresponding photodiode; The second type of pixel further includes a microlens disposed in the optical structure layer and located above the color filter.

2. The high dynamic range CMOS image sensor according to claim 1, characterized in that: The metal anti-reflection layer is configured as a metal grid, and the grid width at the periphery of the metal grid is greater than the grid width inside the metal grid.

3. The high dynamic range CMOS image sensor according to claim 1, characterized in that: The metal anti-reflection layer is configured as a metal grid, and the grid depth at the periphery of the metal grid is greater than the grid depth inside the metal grid.

4. The high dynamic range CMOS image sensor according to any one of claims 1 to 3, characterized in that: The metal anti-reflection layer is configured as a metal grid, and the shapes of the internal divisions of the metal grid include a circle, a tic-tac-toe shape, or a square.

5. The high dynamic range CMOS image sensor according to claim 4, characterized in that: The grid width inside the metal grid is the same.

6. The high dynamic range CMOS image sensor according to claim 1, characterized in that: The cross-sectional size of the first type of pixels is not greater than the cross-sectional size of the second type of pixels.

7. The high dynamic range CMOS image sensor according to claim 6, characterized in that: The image sensor includes a pixel array consisting of a plurality of first Bayer array pixel unit groups and a plurality of second Bayer array pixel unit groups; each of the first Bayer array pixel unit groups includes four second-type pixels arranged in a 2×2 form; each of the second Bayer array pixel unit groups includes four first-type pixels arranged in a 2×2 form, wherein the number of the first-type pixels corresponds to the number of the second-type pixels, and the first-type pixels are formed at the center of any four mutually adjacent second-type pixels.

8. The high dynamic range CMOS image sensor according to claim 6, characterized in that: The image sensor includes a pixel array composed of multiple quad-Bayer array pixel unit groups; each of the quad-Bayer array pixel unit groups includes four RGGB pixel units arranged in a 2×2 form; each of the pixel units includes four pixels of the same color arranged in a 2×2 form, wherein three pixels are set as the second type of pixels, one pixel is set as the first type of pixel, and all the first type of pixels are set at the same position in the pixel unit where they are located.

9. The high dynamic range CMOS image sensor according to claim 6, characterized in that: The image sensor includes a pixel array consisting of a plurality of four Bayer array pixel unit groups and a plurality of Bayer array pixel unit groups; each of the four Bayer array pixel unit groups includes four RGGB pixel units arranged in a 2×2 form, and each of the pixel units includes four second-type pixels of the same color arranged in a 2×2 form; each of the Bayer array pixel unit groups includes four RGGB first-type pixels arranged in a 2×2 form, wherein the first-type pixel is formed at the center position of the pixel unit where the second-type pixel is located and the color of its color filter is consistent with the color of the color filter of the second-type pixel in the pixel unit where it is located.

10. A method for preparing a high dynamic range CMOS image sensor, characterized in that: The preparation method comprises: Providing a semiconductor structure layer, and forming a photodiode in the semiconductor structure layer; A circuit connection layer is formed on the semiconductor structure layer, wherein transfer gates corresponding to the photodiodes are formed in the circuit connection layer; forming a first light-transmitting medium layer and a metal anti-reflection material layer in sequence on the semiconductor structure layer; The metal anti-reflection material layer is patterned to obtain a metal anti-reflection layer and an optical isolation structure; wherein the area where the metal anti-reflection layer is located corresponds to the first type of pixels, and the area where the optical isolation structure is located corresponds to the second type of pixels, and the cross-sectional pattern of the metal anti-reflection layer is a centrally symmetrical pattern to block a portion of the image light from entering the corresponding photodiode; and the optical isolation structure is used to separate the image light between two adjacent pixels; Filling the metal anti-reflection layer with a second light-transmitting medium layer so that the surface of the metal anti-reflection layer is flush; A color filter and a microlens are sequentially formed on the metal anti-reflection layer and the second light-transmitting medium layer, wherein the microlens is only formed on the second type of pixels.

Images