Image sensor and method of manufacturing the same, electronic device

Abstract

The application describes an image sensor, comprising: a semiconductor substrate having opposite first and second surfaces; a plurality of pixel units, each pixel unit comprising at least a photoelectric conversion region located within the semiconductor substrate, the photoelectric conversion region extending from the second surface to the first surface of the semiconductor substrate; and a plurality of annular trench isolations, each annular trench isolation extending from the first surface to the second surface of the semiconductor substrate, and at least one annular trench isolation being arranged around the photoelectric conversion region to enclose the photoelectric conversion region. The application also describes a method for manufacturing the image sensor and an electronic device comprising the image sensor. In the application, the annular closed-loop trench isolation structure provides 360° full-directional continuous physical isolation for each photoelectric conversion region. Compared with the conventional linear deep trench isolation structure (BDTI) which only sets up a barrier in a limited direction, the annular trench isolation structure makes the isolation more thorough, and the blocking effect of crosstalk light is significantly enhanced.

Description

Technical Field

[0001] This invention belongs to the field of imaging, and in particular relates to an image sensor, a method for manufacturing the image sensor, and an electronic device incorporating the image sensor. Background Technology

[0002] Figure 1 The back-illuminated CMOS image sensor shown is widely used in consumer electronics, security monitoring, and automotive imaging due to its high light sensitivity. However, as pixel sizes continue to shrink, optical crosstalk between adjacent pixels becomes increasingly prominent, severely impacting the signal-to-noise ratio and color reproduction of images.

[0003] To address inter-pixel crosstalk, existing technologies commonly introduce back-side deep trench isolation (BDTI) structures at pixel boundaries. For example... Figure 2 As shown, the layout of a conventional BDTI is a linear trench, which means that straight deep trenches are set in the row and column directions of the pixel array. These straight trenches intersect each other to form a grid-like structure, and the trenches are usually filled with dielectric materials such as silicon oxide.

[0004] Traditional linear grid stray light transmission (BDTI) blocks crosstalk light from adjacent pixels to some extent by setting isolation segments along the row and column directions of pixels, including light rays incident diagonally. However, this linear isolation structure only provides a limited isolation barrier at pixel boundaries and does not completely surround the pixel. For stray light from adjacent pixels in various directions, the blocking capability of linear BDTI is still insufficient, and some light rays may still penetrate or bypass the isolation structure to enter the pixel, affecting the adequacy of the isolation effect.

[0005] Therefore, how to provide a BDTI structure with stronger isolation and more thorough blocking to more effectively block crosstalk light from all directions of surrounding pixels, so as to further improve the imaging performance of image sensors, is a technical problem that urgently needs to be solved. Summary of the Invention

[0006] In view of the above-mentioned disadvantages of the prior art, the present invention provides an image sensor, comprising: a semiconductor substrate having opposing first and second surfaces; a plurality of pixel units, each pixel unit including at least a photoelectric conversion region located within the semiconductor substrate, the photoelectric conversion region extending from the second surface of the semiconductor substrate to the first surface; a plurality of annular trench isolations extending from the first surface of the semiconductor substrate to the second surface, and at least one annular trench isolation is disposed around the photoelectric conversion region to surround the photoelectric conversion region.

[0007] Optionally, the annular trench isolation is filled with silicon nitride.

[0008] Optionally, the annular trench isolation further includes a first filling layer located within the silicon oxide, the first filling layer extending along the depth direction of the annular trench isolation; the first filling layer is composed of a first metal material layer and a first barrier layer surrounding the first metal material layer.

[0009] Optionally, the first metallic material layer includes one or more of tungsten, aluminum, copper, titanium, and tantalum; and / or, the first barrier layer includes one or more of titanium, titanium nitride, and tantalum nitride.

[0010] Optionally, the annular trench isolation further includes a second filler layer located within the silicon oxide, the second filler layer extending along the depth direction of the annular trench isolation; the second filler layer is composed of a second metal material layer, a second barrier layer surrounding the second metal material layer, and a High-K dielectric layer surrounding the second barrier layer.

[0011] Optionally, the second metallic material layer includes one or more of tungsten, aluminum, copper, titanium, and tantalum; and / or, the second barrier layer includes one or more of titanium, titanium nitride, and tantalum nitride; and / or, the High-K dielectric layer includes one or more of hafnium oxide, aluminum oxide, tantalum oxide, and zirconium oxide.

[0012] Optionally, the annular groove isolation is a closed loop, and there is a gap between adjacent annular groove isolations.

[0013] Optionally, the pixel unit further includes a color filter layer located above the first surface of the semiconductor substrate. The color filter layer includes a plurality of color filters that are disposed one-to-one with the photoelectric conversion area. The plurality of color filters include at least a first color filter and a second color filter with different filtering bands, wherein the filtering band wavelength of the first color filter is greater than the filtering band wavelength of the second color filter.

[0014] Optionally, the depth of the annular trench isolation below the first color filter is greater than the depth of the annular trench isolation below the second color filter, wherein the depth is the dimension by which the annular trench isolation extends from the first surface of the semiconductor substrate to the second surface.

[0015] Optionally, the multiple color filters may also include a third color filter, wherein the filtering wavelength of the second color filter is greater than that of the third color filter, and the depth of the annular groove isolation below the second color filter is greater than that of the annular groove isolation below the third color filter.

[0016] Optionally, the annular groove isolation below the first color filter has a wider annular groove isolation than the annular groove isolation below the second color filter. The annular groove isolation has an inner peripheral wall and an outer peripheral wall, and the average radial distance between the inner peripheral wall and the outer peripheral wall is the annular groove isolation width.

[0017] Optionally, the radial dimensions of the area enclosed by the outer peripheral walls of the annular grooves below the first and second color filters are the same, while the radial dimensions of the area enclosed by the inner peripheral walls of the annular grooves are different.

[0018] Optionally, the radial dimensions of the area enclosed by the inner peripheral walls of the annular grooves below the first and second color filters are the same, while the radial dimensions of the area enclosed by the outer peripheral walls of the annular grooves are different.

[0019] Optionally, the multiple color filters may also include a third color filter, wherein the filtering wavelength of the second color filter is greater than that of the third color filter, and the ring width of the annular groove below the second color filter is greater than that of the annular groove below the third color filter.

[0020] Optionally, the pixel unit also includes a metal light-blocking layer located between adjacent color filters; the width of the metal light-blocking layer located between the second and third color filters is greater than the width of the metal light-blocking layer located between the second and first color filters.

[0021] Optionally, the first color filter is a red color filter, the second color filter is a green color filter, and the third color filter is a blue color filter.

[0022] The present invention also provides a method for fabricating the above-mentioned image sensor, the method comprising: providing a semiconductor substrate; performing ion implantation on the front side of the semiconductor substrate to form a plurality of photoelectric conversion regions; thinning the back side of the semiconductor substrate, wherein the thinned back side is defined as a first surface and the front side is defined as a second surface; forming a hard mask layer on the first surface of the semiconductor substrate; patterning the hard mask layer to define a plurality of annular patterns corresponding one-to-one with the plurality of photoelectric conversion regions; using the patterned hard mask layer as a mask, performing dry etching from the first surface of the semiconductor substrate to form annular trenches; filling the annular trenches with an isolation material to form a plurality of annular trench isolations, such that at least one annular trench isolation is provided around the photoelectric conversion regions to surround the photoelectric conversion regions.

[0023] Optionally, the image sensor includes a first color filter, a second color filter, and a third color filter, wherein the first, second, and third color filters have different filtering wavelengths. The annular pattern includes a first annular pattern corresponding to the first color filter, a second annular pattern corresponding to the second color filter, and a third annular pattern corresponding to the third color filter. The dry etching process includes: performing a first photolithography, opening the first annular pattern, and performing deep reactive ion etching (DRIE) with an etching depth of D1; performing a second photolithography, opening the first and second annular patterns, and performing DIE with an etching depth of D2; and performing a third photolithography, opening the first, second, and third annular patterns, and performing DIE with an etching depth of D3. The final depth of the annular trenches corresponding to the first, second, and third annular patterns decreases sequentially.

[0024] The present invention also provides an electronic device comprising the above-described image sensor.

[0025] Compared with the prior art, the present invention has at least the following outstanding advantages: In this embodiment, each photoelectric conversion region possesses an independent and complete annular trench isolation. This annular trench isolation provides continuous physical isolation for each photoelectric conversion region in all 360° directions. Regardless of the direction from which crosstalk light originates from adjacent photoelectric conversion regions, it will be blocked by this continuous annular trench isolation wall. Compared to the traditional linear deep trench isolation structure (BDTI), which only sets barriers in a limited number of directions, the annular trench isolation structure provides more thorough isolation, significantly enhancing the blocking effect of crosstalk light. Furthermore, each pixel unit's photoelectric conversion region has its own independent annular isolation wall, independent of the structure of adjacent pixel units to form the isolation boundary, resulting in consistent and predictable isolation effects. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of a partial cross-sectional structure of a back-illuminated CMOS image sensor in the prior art; Figure 2 This is a schematic diagram of the layout of a deep back trench isolation structure (BDTI) in the prior art; Figure 3 A partial cross-sectional structural diagram of an image sensor provided in this application; Figure 4 A top view of a pixel array in an image sensor provided in this application; Figure 5 This is a schematic diagram of the annular grooved isolation membrane structure provided in this application; Figure 6 A partial cross-sectional structural diagram of another image sensor provided in this application; Figure 7A top view of a pixel array in another image sensor provided in this application; Figure 8 A partial cross-sectional structural schematic diagram of another image sensor provided in this application; Figure 9 This is a top view of the annular trench isolation provided in this application; Figure 10 A top view of a pixel array in another image sensor provided in this application; Figure 11 A top view of a pixel array in another image sensor provided in this application; Figure 12 A partial cross-sectional structural schematic diagram of another image sensor provided in this application; Figure 13 This application provides a process flow diagram for fabricating an image sensor. Detailed Implementation

[0027] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0028] In the detailed description of embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged and not to scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0029] For ease of description, spatial relation terms such as "below," "below," "less than," "below," "above," and "upper" may be used herein to describe the relationship between one element or feature shown in the accompanying drawings and other elements or features. It will be understood that these spatial relation terms are intended to include directions other than those depicted in the drawings for devices in use or operation. Furthermore, when a layer is referred to as being "between" two layers, it can be the only layer between the two layers, or there may be one or more layers in between. Additionally, the term "between" as used in this invention includes both endpoint values.

[0030] In the context of this application, the structure described above the first feature may include embodiments in which the first and second features are formed in direct contact, or embodiments in which additional features are formed between the first and second features, such that the first and second features may not be in direct contact.

[0031] It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0032] like Figure 3 and Figure 4 As shown, the image sensor 10 of this application includes: Semiconductor substrate 100 has opposing first and second surfaces; Multiple pixel units, each pixel unit including at least a photoelectric conversion region 200 located within a semiconductor substrate 100, the photoelectric conversion region 200 extending from a second surface of the semiconductor substrate 100 to a first surface; Understandably, for a clear purpose, Figure 3 Only six pixel units are shown. Figure 4 Only nine pixel units are shown in the image. However, in actual applications of the image sensor 10, the image sensor 10 may include thousands of rows and / or columns of pixel units; similarly, in some embodiments, the pixel units may have arrangements other than rows and / or columns.

[0033] Furthermore, Figure 3 In this embodiment, the pixel unit also includes a color filter layer located above the first surface of the semiconductor substrate 100. The color filter layer includes multiple color filters 300, and each color filter 300 is correspondingly arranged with a photoelectric conversion region 200 to selectively transmit or absorb incident light of a specific wavelength to the photoelectric conversion region 200, thereby achieving color reproduction of the image sensor. The color filters 300 can be made of optical resin, and by adjusting the proportion of materials inside the optical resin, light of different wavelengths can be allowed to pass through, thereby achieving the color filtering effect. Generally speaking, the color filters 300 include three basic colors: red (R), green (G), and blue (B). The most common color filter arrangement is the Bayer array RGGB, where each basic repeating unit has one R color, one B color, and two G color filters. Of course, this embodiment is applicable to color image sensors containing multiple color pixels, including but not limited to the RGGB Bayer array, and can also be applied to other color filter arrays such as RGBW, RCCB, and RGB-IR. Optionally, the pixel unit also includes a microlens layer disposed on the side of the color filter layer away from the photoelectric conversion area 200. The microlens layer includes a plurality of microlenses 400, each microlens corresponding to a color filter and a photoelectric conversion area to form a pixel unit. The purpose of making the microlenses is to gather more light into the photoelectric conversion area 200 to improve the photosensitivity of the photoelectric conversion area 200.

[0034] Multiple annular trench isolations 500 extend from a first surface of the semiconductor substrate 100 to a second surface, and at least one annular trench isolation 500 is disposed around the photoelectric conversion region 200.

[0035] Understandable, for reference Figure 4 The annular groove isolation 500 is a closed loop, and there is a gap between adjacent annular groove isolations 500. That is, each photoelectric conversion area is surrounded by an independent, top-view, ring-shaped, end-to-end annular groove isolation 500. The annular groove isolation 500 is composed of four deep groove edges connected end to end, completely and seamlessly enclosing the photosensitive area of ​​the corresponding PD. Furthermore, in other embodiments, two or more annular groove isolations 500 can be arranged around the photoelectric conversion area 200, with multiple annular groove isolations 500 surrounding the photoelectric conversion area 200 in layers.

[0036] In this embodiment, each of the two adjacent photoelectric conversion regions has an independent and complete annular trench isolation, forming a "double-walled" parallel relationship, rather than... Figure 1 , 2 In the existing scheme shown, adjacent photoelectric conversion areas share the same "wall." The annular closed-loop trench isolation structure provides continuous physical isolation for each photoelectric conversion area in all 360° directions. Regardless of the direction of crosstalk light originating from adjacent photoelectric conversion areas, it will be blocked by this continuous annular trench isolation wall. Compared to the traditional linear deep trench isolation structure (BDTI), which only sets barriers in a limited number of directions, the annular trench isolation structure provides more thorough isolation, significantly enhancing the blocking effect of crosstalk light. Furthermore, each pixel unit's photoelectric conversion area has its own independent annular isolation wall, independent of the structure of adjacent pixel units to form the isolation boundary, resulting in consistent and predictable isolation effects.

[0037] In some embodiments, such as Figure 5 As shown in (a), the annular trench isolation 500 is filled with silicon oxide (SiO2). In other embodiments, in addition to being filled with silicon oxide, the annular trench isolation 500 further includes a first filling layer or a second filling layer located within the silicon oxide, the first filling layer or the second filling layer extending along the depth direction of the annular trench isolation. Optionally, as... Figure 5As shown in (b), the first filling layer consists of a first metal material layer Metal_1 and a first barrier layer Block_1 surrounding the metal material layer. The first metal material layer may include one or more of tungsten (W), aluminum (Al), copper (Cu), titanium (Ti), and tantalum (Ta), and the first barrier layer may include one or more of titanium (Ti), titanium nitride (TiN), and tantalum nitride (TaN). The metal material has high reflectivity to visible and near-infrared light. After being filled in the annular trench, it can effectively reflect obliquely incident light and prevent it from leaking from one photoelectric conversion area to an adjacent photoelectric conversion area, thereby suppressing optical crosstalk. However, the adhesion between metal materials (especially tungsten) and silicon oxide is usually poor. The barrier layer is used to connect the metal filling layer and the insulating medium, and can form good bonding force with both the metal and silicon oxide at the same time, preventing metal layer peeling or detachment in subsequent processes and preventing metal atom diffusion contamination of the semiconductor substrate. Optionally, such as Figure 5 As shown in (c), the second filling layer consists of a second metal material layer Metal_2, a second barrier layer Block_2 surrounding the second metal material layer, and a High-K dielectric layer surrounding the second barrier layer Block_2. Similarly, the second metal material layer may include one or more of tungsten (W), aluminum (Al), copper (Cu), titanium (Ti), and tantalum (Ta), the second barrier layer may include one or more of titanium (Ti), titanium nitride (TiN), and tantalum nitride (TaN), and the High-K dielectric layer may include one or more of hafnium oxide (HfO2), aluminum oxide (Al2O3), tantalum oxide (Ta2O5), and zirconium oxide (ZrO2). The High-K dielectric layer has a high dielectric constant, which can provide greater capacitive coupling at the same thickness. It can enhance the isolation effect without increasing the trench area, and further reduce electrical crosstalk and leakage between adjacent photoelectric conversion regions.

[0038] In this embodiment, the annular trench isolation is filled with a metal material or a metal composite material layer. Metal has extremely strong absorption and reflection capabilities for visible and near-infrared light, making it difficult for stray light to penetrate the metal wall and enter adjacent pixels. Compared to the traditional silicon oxide filling mechanism that relies on total internal reflection, the blocking effect is more reliable and thorough. The "double-wall" structure between two independent annular trench isolations of adjacent pixels forms a "metal-thin silicon-metal" dual isolation. Even if there are minor defects on one side of the interface, the other side can still provide effective blocking, further enhancing the reliability of the isolation. In this application, by changing BDTI from "linear" to "annular full enclosure" and combining it with metal filling, the isolation between pixel units is made more comprehensive and thorough. Compared to traditional solutions, it has a stronger ability to block crosstalk, significantly improving the imaging quality of image sensors with small pixel sizes.

[0039] Building upon the aforementioned basic scheme of independent annular trench isolation for pixel units, this invention further provides an embodiment: multiple color filters include a first color filter and a second color filter with different filtering wavelengths, wherein the wavelength of the filtering band of the first color filter is greater than that of the filtering band of the second color filter. That is, based on the spectral characteristics of different color pixel units, the annular trench isolation corresponding to pixel units with different spectral responses is designed differently, particularly in terms of the depth of the annular trench isolation.

[0040] In some embodiments, the depth of the annular trench isolation below the first color filter is greater than the depth of the annular trench isolation below the second color filter, wherein the depth is the dimension of the annular trench isolation extending from the first surface to the second surface of the semiconductor substrate. In this embodiment, since different color pixels have significantly different penetration depths in silicon due to their corresponding incident light wavelengths, if all pixel units use annular trench isolation of the exact same depth, this depth must be designed according to the pixel unit with the deepest penetration to ensure effective isolation for it. However, this "one-size-fits-all" design means that for other pixels with shallower penetration depths, the depth of the annular trench isolation far exceeds their actual optical isolation requirements. Deep trench etching processes can damage the silicon substrate lattice, introducing interface defects and stress. The deeper the annular trench isolation, the larger the damaged area, resulting in higher levels of dark current and noise. Therefore, for pixel units with shallower penetration depths, excessively deep annular trench isolation unnecessarily increases the damaged area of ​​the silicon substrate in depth areas where isolation is not required, leading to increased dark current, decreased full-well capacity, and consequently, dragging down the overall electrical performance of the image sensor. Furthermore, similarly, when multiple color filters include a third color filter, and the filtering wavelength of the second color filter is greater than the filtering wavelength of the third color filter, then the depth of the annular groove isolation below the second color filter is greater than the depth of the annular groove isolation below the third color filter.

[0041] Taking the common Bayer array RGGB as an example, combined with Figure 6 , Figure 7 As shown, the color filters contain three colors: red (R), green (G), and blue (B). The wavelength of red is greater than that of green, which is greater than that of blue. Therefore, the first color filter is a red filter, the second is a green filter, and the third is a blue filter. The depth of the annular groove isolation 500 below the red filter is greater than the depth of the annular groove isolation 500 below the green filter, and the depth of the annular groove isolation 500 below the green filter is greater than the depth of the annular groove isolation 500 below the blue filter.

[0042] Because red light has a longer wavelength, approximately 620-750nm, it penetrates deeply into silicon and easily propagates laterally in deeper substrate layers, causing crosstalk to adjacent pixels. Blue light has a shorter wavelength, approximately 450-495nm, and penetrates shallowly, mainly being absorbed near the silicon surface. Green light has a wavelength between red and blue, approximately 495-570nm. Of course, for other color combinations, the spectral response and penetration depth of each channel also differ. Specific differentiated structures for red, green, and blue pixels are shown below. Figure 6 As shown: the red pixel R (long wavelength) has the deepest annular trench isolation to match the long absorption depth of red light, effectively blocking stray light propagating laterally from deep layers; the blue pixel B (short wavelength) has the shallowest annular trench isolation because blue light is absorbed only in the near-surface region of silicon, and a shallower annular trench isolation is sufficient to meet optical isolation requirements, while significantly reducing etching damage to the silicon substrate, lowering dark current, and improving signal-to-noise ratio and full-well capacity; the green pixel G (medium wavelength) has an annular trench isolation depth between that of the red and blue pixels. In the color pixel array, each type of pixel is equipped with an annular trench isolation depth that matches its own spectral characteristics, i.e., the long wavelength pixels have the deepest annular trench isolation, the short wavelength pixels have the shallowest annular trench isolation, and the medium wavelength pixels have an intermediate annular trench isolation depth. Similarly, this differentiated design principle also applies to other color filter arrays such as RGBW, RCCB, and RGB-IR, and the depth of the annular trench isolation can be flexibly adjusted according to the actual spectral response of each channel.

[0043] In this embodiment, the goal is not to achieve uniform depth of the annular trench isolation for all pixels. Instead, effective isolation is achieved at a depth that is "just enough" based on the actual optical isolation requirements of each color pixel, thereby minimizing unnecessary damage to the silicon substrate and optimizing electrical performance.

[0044] In other embodiments, to further enhance the isolation effect, the annular trench isolation widths of pixels with different filtering wavelengths can differ. Multiple color filters include a first color filter and a second color filter with different filtering wavelengths. The annular trench isolation width below the first color filter is greater than the annular trench isolation width below the second color filter. The annular trench isolation has an inner peripheral wall and an outer peripheral wall, and the average radial distance between the inner and outer peripheral walls is the annular trench isolation width. It is understood that the inner peripheral wall is the sidewall of the annular trench isolation facing its central axis, and the outer peripheral wall is the sidewall of the annular trench isolation away from its central axis. Furthermore, when the multiple color filters also include a third color filter, and the filtering wavelength of the second color filter is greater than the filtering wavelength of the third color filter, then the annular trench isolation width below the second color filter is greater than the annular trench isolation width below the third color filter.

[0045] Taking the common Bayer RGGB array as an example, such as Figure 8 and 9 As shown, the color filters include three colors: red, green, and blue. The wavelength of red is greater than that of green, which is greater than that of blue. Therefore, the first color filter is red, the second is green, and the third is blue. The annular groove isolation 500 located below the blue (B) filter has a width of Width_B, the annular groove isolation 500 located below the green (G) filter has a width of Width_G, and the annular groove isolation 500 located below the red (R) filter has a width of Width_R, where Width_B < Width_G < Width_R.

[0046] Among them, long-wavelength pixels (such as red) have deeper light penetration and longer lateral propagation distance, so their annular trench isolation uses a wider width to provide stronger lateral blocking capability; short-wavelength pixels (such as blue) have shallower light penetration, so they can still use a narrower annular trench isolation width. The annular trench isolation width of medium-wavelength pixels (such as green) is between that of red pixels and blue pixels.

[0047] Based on the above embodiments, such as Figure 10 As shown, in this embodiment, the radial dimensions of the region enclosed by the outer peripheral walls of the annular groove isolation below the color filters with different wavelengths are consistent, while the radial dimensions of the region enclosed by the inner peripheral walls of the annular groove isolation are inconsistent. Taking the Bayer array RGGB as an example, the inconsistent radial dimensions of the region enclosed by the inner peripheral walls mean that the photoelectric conversion area below the blue B color filter is larger than the photoelectric conversion area below the green G color filter, and the photoelectric conversion area below the green G color filter is larger than the photoelectric conversion area below the red R color filter. In this embodiment, the annular groove isolation of each pixel is moved outward relative to the pixel center. This adjustment expands the area of ​​the photoelectric conversion area enclosed by the annular groove isolation, thereby improving the photoelectric conversion efficiency of the pixel. In this layout, the position of the annular groove isolation of each pixel unit is uniformly expanded outward, while the gap width of the annular groove isolation between adjacent pixel units remains uniform. In addition, a metal light-blocking layer (Metal Grid, MG) is often provided in the image sensor between the color filters 300. Figure 8(Not shown in the image), its core function is to prevent optical signals from interfering with other pixel units. In this embodiment, the metal light-blocking layer between pixel units only needs to maintain a uniform width, ensuring that its vertical projection completely covers the gap area isolated by the annular groove below. Since the width of the annular groove isolation gap is uniform, the metal light-blocking layer does not need to be differentiated. As long as it is uniformly widened to just cover the gap, it can effectively block oblique incident light from leaking through the gap, avoiding crosstalk, while reducing the complexity of mask design and process, and maintaining good manufacturability.

[0048] In other embodiments, such as Figure 11 As shown, the radial dimensions of the region enclosed by the inner peripheral walls of the annular groove isolation below color filters with different wavelengths are consistent, while the radial dimensions of the region enclosed by the outer peripheral walls of the annular groove isolation are inconsistent. Again, taking the Bayer RGGB array as an example, the outer ring width of the annular groove isolation 500 below the blue B color filter is smaller than the outer ring width of the annular groove isolation 500 below the green color filter G, and the outer ring width of the annular groove isolation 500 below the green color filter G is smaller than the outer ring width of the annular groove isolation 500 below the red R color filter. Furthermore, the size of the photoelectric conversion area of ​​each pixel unit remains the same to maintain the consistency of the photosensitivity of each pixel.

[0049] In this situation, because the annular trench isolation widths of each pixel unit are different, while the photoelectric conversion area size is the same, the silicon gap widths between the annular trench isolation sidewalls of adjacent pixels are inconsistent. The existence of this gap can become a path for optical crosstalk—tilted incident light converged by microlenses may pass through this gap and enter adjacent pixels.

[0050] To address this issue, this embodiment employs a differentiated design on the inter-pixel metal light-blocking layer (MG), specifically a differentiated width design for the MG at the boundaries of different colors. This results in the width of the metal light-blocking layer located between the second and third color filters being greater than the width of the metal light-blocking layer located between the second and first color filters. For example... Figure 11 and 12As shown, the width of the metal light-blocking layer between the blue and green color filters is MG_B&G, and the width of the metal light-blocking layer between the red and green color filters is MG_R&G, with MG_B&G > MG_R&G. Because the spacing of the annular groove isolation between the blue and green pixels is relatively wide, a wider MG is used at the BG junction to more effectively block light from leaking through the annular groove isolation gap, reducing crosstalk. Conversely, the spacing of the annular groove isolation between the red and green pixels is relatively narrow, so a narrower MG is used at the RG junction to prioritize pixel sensitivity and avoid unnecessary sensitivity loss due to excessive occlusion. This design, which adjusts the MG width specifically for different color pixel combinations, achieves a fine balance between crosstalk suppression and sensitivity maintenance.

[0051] This application also provides a method for fabricating an image sensor, such as... Figure 13 As shown, the method for forming the image sensor described above includes: S11: Provides a semiconductor substrate; Before forming the annular trench isolation, the FEOL (front-end process) + BEOL (back-end interconnect) process must be completed. The front-end process before forming the photoelectric conversion region may include defining the active area and forming shallow trench isolation (STI). It also forms well regions (P-well / N-well) to provide the basic doping background for the transistors and photodiodes in the subsequent pixel circuit.

[0052] S12: Ion implantation from the front to the back of the semiconductor substrate to form multiple photoelectric conversion regions; This is the core step in the FEOL process for forming the photoelectric conversion region. The PN junction of the photodiode and a p+ pinning layer on the surface are formed in the semiconductor substrate through ion implantation to suppress dark current. Subsequent processes include gate formation.

[0053] S13: Thin the back side of the semiconductor substrate, and use the thinned new surface as the first surface and the front side of the semiconductor substrate as the second surface; After completing the FEOL (front-end process) + BEOL (back-end interconnect) processes, the back side of the semiconductor substrate is polished and CMP thinned to the target thickness. As the subsequent light-gathering surface, the remaining silicon thickness on the back side must be greater than the maximum etching depth of the annular trench isolation, and the photoelectric conversion region must have a sufficient absorption layer.

[0054] S14: Form a hard mask layer on the first surface of the semiconductor substrate; A hard mask layer is formed by depositing a hard mask material, such as a SiO2 / SiN stack.

[0055] S15: Pattern the hard mask layer and define multiple ring patterns that correspond one-to-one with multiple photoelectric conversion regions; Then photolithography is used to define each individual closed-loop pattern.

[0056] S16: Using the patterned hard mask layer as a mask, dry etching is performed from the first surface of the semiconductor substrate to the second surface to form annular trenches. Dry etching is used to transfer the pattern to a hard mask. After resist removal, deep reactive ion etching (DRIE) is performed using the hard mask as a shield to form continuous, uninterrupted annular trenches. DRIE employs alternating etching and passivation cycles (e.g., Bosch process), enabling the achievement of trench morphologies with vertical sidewalls under high aspect ratio conditions. Etching gases can include SF6, C4F8, O2, etc., where SF6 is used to etch silicon, and C4F8 is used to form a polymer passivation layer on the sidewalls to prevent lateral etching.

[0057] S17: Fill the annular groove with insulating material to form multiple annular groove isolations, so that at least one annular groove isolation is provided around the photoelectric conversion area to surround the photoelectric conversion area.

[0058] In one embodiment, the isolation material includes silicon oxide, which can be deposited using HDP-CVD or TEOS processes to achieve good trench filling. In another embodiment, the filling step involves first filling the annular trench with silicon nitride, and then forming a filling layer within the silicon nitride. The filling layer consists of a metal material layer; or the filling layer consists of a metal material layer and a barrier layer surrounding the metal material layer; or the filling layer consists of a metal material layer, a barrier layer surrounding the metal material layer, and a High-K dielectric layer surrounding the barrier layer. After filling, excess isolation material on the first surface can be removed using a CMP process to planarize the first surface and form the final annular trench isolation. The annular trench isolations corresponding to adjacent photoelectric conversion regions are independent, with thin silicon spacers retained in between to provide a structural basis for subsequent double-wall isolation.

[0059] After the final annular trench isolation is formed, a back passivation layer is deposited, and MG, color filter layer and microlens are fabricated, all according to the conventional BSI process.

[0060] Based on the above-mentioned method for preparing an image sensor, the image sensor includes a first color filter, a second color filter, and a third color filter. The first color filter, the second color filter, and the third color filter have different filtering wavelengths. The annular pattern includes a first annular pattern corresponding to the first color filter, a second annular pattern corresponding to the second color filter, and a third annular pattern corresponding to the third color filter. The etching depth of the annular trench isolation is modulated according to the color filter type. In one embodiment, depth modulation is achieved using a step-by-step photolithography-etching process, wherein dry etching includes: First photolithography: Open the first annular pattern corresponding to the first color filter and perform deep reactive ion etching with an etching depth of D1; Second photolithography: Open the first and second ring patterns and perform deep reactive ion etching with an etching depth of D2; Third photolithography: Open the first, second, and third ring patterns and perform deep reactive ion etching to a depth of D3.

[0061] Since the first annular pattern has already been etched in the previous etching step, it continues to be etched in subsequent etching steps; similarly, the second annular pattern continues to be etched in the third etching step after the second etching step, so the final depth is: The depth of the first annular pattern corresponding to the first color filter is: D1+D2+D3; The depth of the second annular pattern corresponding to the second color filter is: D2+D3; The depth of the third annular pattern corresponding to the third color filter is D3.

[0062] This results in the final depth of the annular groove corresponding to the first color filter being greater than the final depth of the annular groove corresponding to the second color filter, and the final depth of the annular groove corresponding to the second color filter being greater than the final depth of the annular groove corresponding to the third color filter. In one embodiment, the first color filter is a red color filter, the second color filter is a green color filter, and the third color filter is a blue color filter; that is, the depth of the annular groove below the red color filter is greater than the depth of the annular groove below the green color filter, and the depth of the annular groove below the green color filter is greater than the depth of the annular groove below the blue color filter.

[0063] This application also provides an electronic device including the aforementioned image sensor. In this electronic device, each photoelectric conversion area possesses an independent and complete annular trench isolation. The annular trench isolation provides continuous physical isolation for each photoelectric conversion area in all 360° directions. Regardless of the direction from which crosstalk light originates from adjacent photoelectric conversion areas, it will be blocked by this continuous annular trench isolation wall. Compared to the traditional linear deep trench isolation structure (BDTI), which only sets barriers in a limited number of directions, the annular trench isolation structure provides more thorough isolation, significantly enhancing the blocking effect of crosstalk light. Furthermore, each pixel unit's photoelectric conversion area has its own independent annular isolation wall, independent of the structure of adjacent pixel units to form the isolation boundary, resulting in consistent and predictable isolation effects.

[0064] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. An image sensor, characterized in that, include: A semiconductor substrate having opposing first and second surfaces; A plurality of pixel units, wherein each pixel unit includes at least a photoelectric conversion region located within the semiconductor substrate, the photoelectric conversion region extending from a second surface of the semiconductor substrate toward a first surface; Multiple annular trench isolations are provided, the annular trench isolations extending from a first surface of the semiconductor substrate to a second surface, and at least one of the annular trench isolations is provided around the photoelectric conversion region to surround the photoelectric conversion region.

2. The image sensor as described in claim 1, characterized in that, The annular trench is filled with silicon oxide.

3. The image sensor as described in claim 2, characterized in that, The annular trench isolation further includes a first filling layer located within the silicon oxide, the first filling layer extending along the depth direction of the annular trench isolation; the first filling layer is composed of a first metal material layer and a first barrier layer surrounding the first metal material layer.

4. The image sensor as described in claim 3, characterized in that, The first metallic material layer comprises one or more of tungsten, aluminum, copper, titanium, and tantalum; and / or, The first barrier layer comprises one or more of titanium, titanium nitride, and tantalum nitride.

5. The image sensor as described in claim 2, characterized in that, The annular trench isolation further includes a second filling layer located within the silicon oxide, the second filling layer extending along the depth direction of the annular trench isolation; the second filling layer is composed of a second metal material layer, a second barrier layer surrounding the second metal material layer, and a High-K dielectric layer surrounding the second barrier layer.

6. The image sensor as described in claim 5, characterized in that, The second metallic material layer includes one or more of tungsten, aluminum, copper, titanium, and tantalum; and / or, The second barrier layer comprises one or more of titanium, titanium nitride, and tantalum nitride; and / or, The High-K dielectric layer includes one or more of hafnium oxide, aluminum oxide, tantalum oxide, and zirconium oxide.

7. The image sensor as claimed in claim 1, characterized in that, The annular groove isolation is a closed loop, and there is a gap between adjacent annular groove isolations.

8. The image sensor as claimed in claim 1, characterized in that, The pixel unit further includes a color filter layer located above the first surface of the semiconductor substrate, the color filter layer including a plurality of color filters disposed in one-to-one correspondence with the photoelectric conversion region; The plurality of color filters includes at least a first color filter and a second color filter with different filtering bands, wherein the wavelength of the filtering band of the first color filter is greater than the wavelength of the filtering band of the second color filter.

9. The image sensor as claimed in claim 8, characterized in that, The depth of the annular trench isolation below the first color filter is greater than the depth of the annular trench isolation below the second color filter, wherein the depth is the dimension by which the annular trench isolation extends from the first surface of the semiconductor substrate to the second surface.

10. The image sensor as claimed in claim 9, characterized in that, The plurality of color filters also includes a third color filter, wherein the wavelength of the filtering band of the second color filter is greater than the wavelength of the filtering band of the third color filter, and the depth of the annular groove isolation below the second color filter is greater than the depth of the annular groove isolation below the third color filter.

11. The image sensor as claimed in claim 8, characterized in that, The annular groove isolation below the first color filter has a wider annular groove isolation than the annular groove isolation below the second color filter. The annular groove isolation has an inner peripheral wall and an outer peripheral wall, and the average radial distance between the inner peripheral wall and the outer peripheral wall is the annular groove isolation width.

12. The image sensor as claimed in claim 11, characterized in that, The radial dimensions of the regions enclosed by the outer peripheral walls of the annular grooves below the first and second color filters are the same, while the radial dimensions of the regions enclosed by the inner peripheral walls of the annular grooves are different.

13. The image sensor as claimed in claim 11, characterized in that, The radial dimensions of the regions enclosed by the inner peripheral walls of the annular grooves below the first and second color filters are the same, while the radial dimensions of the regions enclosed by the outer peripheral walls of the annular grooves are different.

14. The image sensor as claimed in claim 11, characterized in that, The plurality of color filters also includes a third color filter, wherein the wavelength of the filtering band of the second color filter is greater than the wavelength of the filtering band of the third color filter, and the ring width of the annular groove below the second color filter is greater than the ring width of the annular groove below the third color filter.

15. The image sensor as claimed in claim 14, characterized in that, The pixel unit also includes a metal light-blocking layer located between adjacent color filters; The width of the metal light-blocking layer located between the second color filter and the third color filter is greater than the width of the metal light-blocking layer located between the second color filter and the first color filter.

16. The image sensor as claimed in any one of claims 10, 14, or 15, characterized in that, The first color filter is a red color filter, the second color filter is a green color filter, and the third color filter is a blue color filter.

17. A method for fabricating an image sensor, used to form an image sensor as described in any one of claims 1 to 16, characterized in that, The method includes: Provide semiconductor substrates; Ion implantation is performed on the front side of the semiconductor substrate to form multiple photoelectric conversion regions; The back side of the semiconductor substrate is thinned, and the thinned back side is defined as the first surface, and the front side is defined as the second surface; A hard mask layer is formed on the first surface of the semiconductor substrate; The hard mask layer is patterned to define multiple annular patterns that correspond one-to-one with the multiple photoelectric conversion regions; Using the patterned hard mask layer as a mask, dry etching is performed from the first surface of the semiconductor substrate to form annular trenches; The annular groove is filled with insulating material to form multiple annular groove isolations, so that at least one of the annular groove isolations is provided around the photoelectric conversion area to surround the photoelectric conversion area.

18. The method for fabricating an image sensor as described in claim 17, characterized in that, The image sensor includes a first color filter, a second color filter, and a third color filter. The first color filter, the second color filter, and the third color filter have different filtering wavelengths. The annular pattern includes a first annular pattern corresponding to the first color filter, a second annular pattern corresponding to the second color filter, and a third annular pattern corresponding to the third color filter. The dry etching includes: The first photolithography is performed to open the first annular pattern, and deep reactive ion etching is performed with an etching depth of D1. A second photolithography process is performed to open the first and second annular patterns, followed by deep reactive ion etching to a depth of D2. A third photolithography step is performed to open the first ring pattern, the second ring pattern and the third ring pattern, and deep reactive ion etching is performed with an etching depth of D3. The final depth of the annular grooves corresponding to the first annular pattern, the second annular pattern, and the third annular pattern decreases sequentially.

19. An electronic device, characterized in that, Including the image sensor as described in any one of claims 1 to 16.

Images