A bionic vision sensor with pulse coding capability and a manufacturing method thereof

Abstract

The application discloses a kind of bionic visual sensor devices with pulse coding capability.Light quantum reaches device, and photoelectron-hole pair will be generated.Light-generated carriers continue to collide with lattice atoms by the strong electric field force between cathode and anode, and new photoelectron-hole pair is generated.The process will be repeated in turn, and then macroscopic current is generated.Through external resistance voltage division, the electric field between cathode and anode is reduced, so that macroscopic current is quenched, thereby realizing the pulse current lasting several nanoseconds.After tens of nanoseconds, the device quickly recovers to the initial state, preparing for detecting the next light quantum.In addition, the inversion layer formed below the polysilicon gate can effectively regulate the wavelength response of the device, enabling it to have a biologically reasonable visual color perception ability.Meanwhile, the device is fully compatible with standard microelectronic technology, enabling large-scale integrated bionic visual sensors to meet the development needs of fields such as bionic robots.

Description

Technical Field

[0001] This invention relates to the field of bionic vision technology, and in particular to a bionic vision sensor with pulse coding capability and its manufacturing method. Background Technology

[0002] The visual system of humans and animals provides the core function for organisms to acquire environmental information and is an important foundation for perception, decision-making, and reaction. Through efficient visual perception systems, organisms in nature can not only perceive light, color, depth, and motion, but also react quickly in complex environments. Bionics, as a discipline that imitates the structure and function of organisms in nature, has been widely applied in various fields, with research on bionic visual sensors receiving significant attention in recent years. Bionic visual sensors provide new technological approaches to intelligent perception and environmental interaction by simulating the perception mechanisms of biological visual systems. Taking the human eye as an example, it has extremely strong light sensitivity and dynamic adjustment capabilities, enabling it to obtain clear visual images under different lighting conditions. However, existing artificial visual systems cannot fully simulate the functions of the human eye, especially in low light, dynamic range, fast response, and energy consumption. To overcome these problems, the design of bionic visual sensors not only requires the characteristics of traditional optical sensors such as high sensitivity, wide field of view, and high resolution, but also the advantages of biological visual systems such as strong adaptability and low energy consumption. Summary of the Invention

[0003] To achieve the above objectives, this invention provides a biomimetic visual sensor with pulse coding capability and its fabrication method, which can simultaneously realize both color selectivity and pulse coding functions. Furthermore, this invention has a simple structure, is compatible with standard microelectronic processes, and is easy to integrate and manufacture on a large scale, thus possessing broad application prospects.

[0004] To achieve the above objectives, the technical solution of the present invention is: a biomimetic vision sensor with pulse coding capability; an N-type buried layer region (102) is provided on a P-type substrate (101); an N-type deep well region (103) is provided on the N-type buried layer region (102).

[0005] The N-type deep well region (103) is provided with the following regions from left to right: a first shallow trench isolation region (201), a first N+ heavily doped region (105), a second shallow trench isolation region (202), a first polysilicon gate region (205), a first P+ heavily doped region (106), a second polysilicon gate region (206), a second P+ heavily doped region (107), a third polysilicon gate region (207), a third P+ heavily doped region (108), an N-well region (104), a fourth polysilicon gate region (208), a fourth P+ heavily doped region (109), a fifth polysilicon gate region (209), a fifth P+ heavily doped region (110), a sixth polysilicon gate region (210), a third shallow trench isolation region (203), a second N+ heavily doped region (111), and a fourth shallow trench isolation region (204).

[0006] The first N+ heavily doped region (105) and the second N+ heavily doped region (111) are connected to the cathode through metal leads, and the first to sixth polysilicon gate regions (205, 206, 207, 208, 209, 210) and the first to fifth P+ heavily doped regions (106, 107, 108, 109, 110) are connected to the anode through metal leads.

[0007] The N-well region (104) and the third P+ heavily doped region (108) form a deep PN junction, the first P+ heavily doped region (106), the second P+ heavily doped region (107), the fourth P+ heavily doped region (109), the fifth P+ heavily doped region (110), and the N-type deep well region (103) form a middle PN junction, and the first to sixth polysilicon gate regions (205, 206, 207, 208, 209, 210) and the N-type deep well region (103) form a shallow PN junction; the three-layer PN junction achieves photon detection of different wavelengths by controlling the bias voltage of the anode and cathode of the control device.

[0008] When photons are absorbed by the three-layer PN junction, a macroscopic current pulse is output from the anode under the action of the avalanche multiplication region formed by the N-well region (104) and the third P+ heavily doped region (108). Within the same time and light intensity, the current pulse frequencies caused by photons of different wavelengths have huge differences, and the device can achieve color-selective perception.

[0009] A method for fabricating a biomimetic vision sensor with pulse coding capability includes the following steps:

[0010] Step 1: A thin silicon oxide (SiO2) layer is grown on a P-type substrate using dry oxidation with oxygen to protect the silicon wafer surface and control the doping depth. The temperature is approximately 900℃.

[0011] Step 2: The pattern on the mask of the N-type buried layer region is transferred to the wafer using micro-development technology to form the photoresist pattern of the N-type buried layer region, while the photoresist is retained in the non-N-type buried layer region; high-energy phosphorus ions are implanted to form the N-type buried layer region, and the photoresist is removed.

[0012] Step 3: The pattern on the N-type deep well region mask is transferred to the wafer using micro-development technology to form the photoresist pattern of the N-type deep well region, while the photoresist is retained in the non-N-type deep well regions; high-energy phosphorus ions are implanted to form the N-type deep well region, and the photoresist is removed.

[0013] Step 4: Apply photoresist onto the silicon wafer to form a uniform photoresist layer. Define the first to fourth shallow trench isolation regions, and expose the photoresist using a mask and ultraviolet light to form a pattern. Use dry etching to remove the silicon oxide layer and some silicon, forming a shallow trench structure with an etching depth of approximately 0.4 μm and an etching angle of approximately 10° with the sidewall of the shallow trench. Use low-pressure chemical vapor deposition (LPCVD) to fill the trench with approximately 1.0 μm thick silicon oxide. Remove excess silicon oxide material using chemical mechanical polishing (CMP) to smooth the silicon wafer surface.

[0014] Step 5: Transfer the pattern on the N-well mask to the wafer using micro-development technology to form the N-well photoresist pattern, while retaining the photoresist in non-N-well areas; implant high-energy phosphorus ions to form local N-wells, and remove the photoresist.

[0015] Step 6: Anneal the ion implantation region. Use rapid thermal annealing (RTA) technology to repair ion implantation damage and activate dopants, effectively reducing dopant diffusion and thus precisely controlling the doping concentration and doping distribution in the shallow region. The temperature is controlled at 1000℃-1300℃ for about 10-60 seconds.

[0016] Step 7: Growth of a thick gate oxide layer. A thick silicon dioxide layer is grown using dry oxidation at a temperature of approximately 900°C. Polysilicon with a thickness of approximately 200 nm is deposited using LPCVD. The gate is formed through photolithography and etching, and the photoresist is removed.

[0017] Step 8: Coat the silicon wafer with photoresist, defining the first and second N+ heavily doped regions. Irradiate the photoresist with a mask and ultraviolet light to form a pattern. Implant heavily doped arsenic ions to form N+ heavily doped regions, then remove the photoresist.

[0018] Step Nine: Coat the silicon wafer with photoresist, defining the first to fifth P+ heavily doped regions. Irradiate the photoresist with a mask and ultraviolet light to form a pattern. Implant heavily doped boron ions to form P+ heavily doped regions, then remove the photoresist. Annealing is then performed to repair crystal damage on the silicon surface caused by ion implantation and to eliminate impurity migration in the heavily doped regions.

[0019] The beneficial effects of this invention are as follows:

[0020] 1. The N-well region, P+ injection region, N-type deep well region, P+ injection region, N-type deep well region, and polysilicon gate region in this invention form three PN junctions of different depths, which improves the photon detection probability of the device and effectively modulates the optical wavelength response of the device, thereby meeting the requirements of biomimetic vision for color perception. Furthermore, by using an external resistor to divide the voltage, the electric field between the cathode and anode is reduced, quenching the macroscopic current and achieving pulse coding.

[0021] 2. The manufacturing method of this invention is simple and easy to operate. The resulting bionic vision sensor with pulse coding capability neither violates the basic design rules of the layout nor requires steps other than the standard BCD process. It effectively controls the light wavelength response of the device, meeting the requirements of bionic vision for color perception. Furthermore, its compatibility with standard microelectronic processes facilitates large-scale integration and production, offering broad application prospects. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the embodiments will be briefly described below.

[0023] Figure 1 This is a cross-sectional view and circuit connection diagram of the bionic vision sensor with pulse coding capability in an embodiment of the present invention;

[0024] Figure 2 This is a schematic diagram of the parasitic structure and working principle of the bionic vision sensor with pulse coding capability in an embodiment of the present invention.

[0025] Figure 3 This is a top view of the bionic vision sensor with pulse coding capability in an embodiment of the present invention;

[0026] Figure 4 The waveform diagrams of the pulse excitation frequencies generated by the bionic vision sensor with pulse coding capability in this embodiment of the invention for four colors are shown.

[0027] The figure includes: 1. P-type substrate, 2. N-type buried layer region, 3. N-type deep well region, 4. N-well region, 5. First N+ heavily doped region, 6. Second N+ heavily doped region, 7. First P+ heavily doped region, 8. Second P+ heavily doped region, 9. Third P+ heavily doped region, 10. Fourth P+ heavily doped region, 11. Fifth P+ heavily doped region, 12. First polysilicon gate region, 13. Second polysilicon gate region, 14. Third polysilicon gate region, 15. Fourth polysilicon gate region, 16. Fifth polysilicon gate region, 17. Sixth polysilicon gate region, 18. First shallow trench isolation region, 19. Second shallow trench isolation region, 20. Third shallow trench isolation region, 21. Fourth shallow trench isolation region. Detailed Implementation

[0028] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0029] The purpose of this invention is to provide a biomimetic visual sensor with pulse coding capability and its fabrication method. This device can convert light quanta into spike pulses, simulating human visual perception organs, converting stimuli into spike pulse frequencies, and then transmitting the spike-coded information to the central nervous system for processing. The fabrication method is simple, utilizing existing standard microelectronic processes without requiring additional process modifications, and it is easily integrated into large-scale arrays.

[0030] To make the above-mentioned objects, features, and advantages of the present invention clearer, the present invention will be briefly described below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the present invention, and other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of the present invention.

[0031] like Figure 1 and Figure 3As shown, the present invention provides a biomimetic vision sensor with pulse coding capability; an annular N-type buried layer region (102) is provided on the P-type substrate (101); an annular N-type deep well region (103) is provided on the N-type buried layer region (102). The N-type deep well region (103) is provided with the following regions from left to right: an annular first shallow trench isolation region (201), an annular first N+ heavily doped region (105), an annular second shallow trench isolation region (202), an annular first polysilicon gate region (205), an annular first P+ heavily doped region (106), an annular second polysilicon gate region (206), an annular second P+ heavily doped region (107), an annular third polysilicon gate region (207), a circular third P+ heavily doped region (108), a circular N-well region (104), an annular fourth polysilicon gate region (208), an annular fourth P+ heavily doped region (109), an annular fifth polysilicon gate region (209), an annular fifth P+ heavily doped region (110), an annular sixth polysilicon gate region (210), an annular third shallow trench isolation region (203), an annular second N+ heavily doped region (111), and an annular fourth shallow trench isolation region (204).

[0032] like Figure 2 As shown, the N-well region (104) and the third P+ heavily doped region (108) constitute a deep PN junction, the first P+ heavily doped region (106), the second P+ heavily doped region (107), the fourth P+ heavily doped region (109), the fifth P+ heavily doped region (110), and the N-type deep well region (103) constitute a middle PN junction, and the first to sixth polysilicon gate regions (205, 206, 207, 208, 209, 210) and the N-type deep well region (103) constitute a shallow PN junction; the three-layer PN junction realizes photon detection of different wavelengths by controlling the bias voltage of the anode and cathode of the control device. Figure 4 The diagram shows the pulse excitation frequency waveforms generated by the device of the present invention under four similar wavelengths and the same illumination intensity. The excitation pulse frequencies responding to green, yellow, orange, and red light are 3.51 × 10⁻⁶, respectively. 5 2.66×10 5 1.87×10 5 and 1.28×10 5 The difference in excitation pulse frequency between adjacent colors exceeds 59,000 Hz, indicating that the device has good selectivity in distinguishing different colors of light. By integrating this device into a bionic vision system, it can simulate the human visual perception organ, converting light stimulation into peak pulse frequencies, and then transmitting the information encoded by the peaks to the central nervous system for processing.

[0033] The P-type substrate, N-type buried layer region, N-type deep well region, N-well region, polysilicon gate region, N+ heavily doped region, P+ heavily doped region, and shallow trench isolation region in this invention can be realized using existing standard microelectronic manufacturing processes without the need for additional masks or process modifications.

[0034] The present invention proposes a method for fabricating a biomimetic vision sensor with pulse coding capability, comprising the following steps:

[0035] Step 1: A thin silicon oxide (SiO2) layer is grown on a P-type substrate using dry oxygen oxidation to protect the silicon wafer surface and control the doping depth. The temperature is approximately 900℃.

[0036] Step 2: The pattern on the mask of the N-type buried layer region is transferred to the wafer using micro-development technology to form the photoresist pattern of the N-type buried layer region, while the photoresist is retained in the non-N-type buried layer region; high-energy phosphorus ions are implanted to form the N-type buried layer region, and the photoresist is removed.

[0037] Step 3: The pattern on the N-type deep well region mask is transferred to the wafer using micro-development technology to form the photoresist pattern of the N-type deep well region, while the photoresist is retained in the non-N-type deep well regions; high-energy phosphorus ions are implanted to form the N-type deep well region, and the photoresist is removed.

[0038] Step 4: Apply photoresist onto the silicon wafer to form a uniform photoresist layer. Define the first to fourth shallow trench isolation regions, and expose the photoresist using a mask and ultraviolet light to form a pattern. Use dry etching to remove the silicon oxide layer and some silicon, forming a shallow trench structure with an etching depth of approximately 0.4 μm and an etching angle of approximately 10° with the sidewall of the shallow trench. Use low-pressure chemical vapor deposition (LPCVD) to fill the trench with approximately 1.0 μm thick silicon oxide. Remove excess silicon oxide material using chemical mechanical polishing (CMP) to smooth the silicon wafer surface.

[0039] Step 5: Transfer the pattern on the N-well mask to the wafer using micro-development technology to form the N-well photoresist pattern, while retaining the photoresist in non-N-well areas; implant high-energy phosphorus ions to form local N-wells, and remove the photoresist.

[0040] Step 6: Anneal the ion implantation region. Use rapid thermal annealing (RTA) technology to repair ion implantation damage and activate dopants, effectively reducing dopant diffusion and thus precisely controlling the doping concentration and doping distribution in the shallow region. The temperature is controlled at 1000℃-1300℃ for about 10-60 seconds.

[0041] Step 7: Growth of a thick gate oxide layer. A thick silicon dioxide layer is grown using dry oxidation at a temperature of approximately 900°C. Polysilicon with a thickness of approximately 200 nm is deposited using LPCVD. The gate is formed through photolithography and etching, and the photoresist is removed.

[0042] Step 8: Coat the silicon wafer with photoresist, defining the first and second N+ heavily doped regions. Irradiate the photoresist with a mask and ultraviolet light to form a pattern. Implant heavily doped arsenic ions to form N+ heavily doped regions, then remove the photoresist.

[0043] Step Nine: Coat the silicon wafer with photoresist, defining the first to fifth P+ heavily doped regions. Irradiate the photoresist with a mask and ultraviolet light to form a pattern. Implant heavily doped boron ions to form P+ heavily doped regions, then remove the photoresist. Annealing is then performed to repair crystal damage on the silicon surface caused by ion implantation and to eliminate impurity migration in the heavily doped regions.

[0044] Any aspects of this invention not described in detail are well-known to those skilled in the art.

[0045] The embodiments of the present invention have been described in detail above, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions, and variations can be made to these embodiments based on the concept of the present invention and on the basis of the prior art, and these variations still fall within the protection scope of the present invention.

Claims

1. A bionic visual sensor with pulse coding capability, characterized in that: a P-type substrate (101) is provided with an N-type buried layer region (102); the N-type buried layer region (102) is provided with an N-type deep well region (103); the N-type deep well region (103) is sequentially provided with a first shallow trench isolation region (201), a first N+ heavily doped region (105), a second shallow trench isolation region (202), a first polysilicon gate region (205), a first P+ heavily doped region (106), a second polysilicon gate region (206), a second P+ heavily doped region (107), a third polysilicon gate region (207), a third P+ heavily doped region (108), an N-well region (104), a fourth polysilicon gate region (208), a fourth P+ heavily doped region (109), a fifth polysilicon gate region (209), a fifth P+ heavily doped region (110), a sixth polysilicon gate region (210), a third shallow trench isolation region (203), a second N+ heavily doped region (111), and a fourth shallow trench isolation region (204) from left to right; the first N+ heavily doped region (105) and the second N+ heavily doped region (111) are connected to a cathode through a metal lead, and the first to sixth polysilicon gate regions (205, 206, 207, 208, 209, 210) and the first to fifth P+ heavily doped regions (106, 107, 108, 109, 110) are connected to an anode through a metal lead.

2. The bionic vision sensor with pulse coding capability according to claim 1, characterized in that: The N-well region (104) and the third P+ heavily doped region (108) form a deep PN junction, the first P+ heavily doped region (106), the second P+ heavily doped region (107), the fourth P+ heavily doped region (109), the fifth P+ heavily doped region (110), and the N-type deep well region (103) form a middle PN junction, and the first to sixth polysilicon gate regions (205, 206, 207, 208, 209, 210) and the N-type deep well region (103) form a shallow PN junction; the three-layer PN junctions realize detection of different wave bands by controlling the bias voltage of the anode and the cathode of the device.

3. The bionic vision sensor with pulse coding capability according to claim 2, characterized in that: When a light quantum is absorbed by the three-layer PN junctions, a macroscopic current pulse is output from the anode under the action of the avalanche multiplication region formed by the N-well region (104) and the third P+ heavily doped region (108); under the same time and light intensity, the frequency of the current pulse caused by different wavelengths of photons is greatly different, and the device can realize color-selective perception. 4.A method for manufacturing the bionic visual sensor with pulse coding capability according to any one of claims 1-3, comprising the following steps: Step one: grow a thin silicon oxide (SiO2) layer on the P-type substrate by oxygen dry oxidation to protect the surface of the silicon wafer and control the depth of doping, at a temperature of 900℃; Step two: transfer the pattern on the N-type buried layer region mask to the wafer by micro-lithography technology to form a photoresist pattern of the N-type buried layer region, and the non-N-type buried layer region retains the photoresist; inject high-energy phosphorus ions to form the N-type buried layer region, and remove the photoresist; Step three: transfer the pattern on the N-type deep well mask to the wafer by micro- lithography technology, form the photoresist pattern of N-type deep well, and the non-N-type deep well area retains the photoresist; Inject high-energy phosphorus ions to form N-type deep well area, and remove the photoresist; Step four: coat photoresist on the silicon wafer to form a uniform photoresist layer, define the first to fourth shallow trench isolation regions, use a mask and ultraviolet light to expose the photoresist to form a pattern, use dry etching to remove the silicon oxide layer and part of the silicon to form a shallow trench structure, the etching depth is 0.4 μm, the etching angle is 10° with the shallow trench sidewall, use low pressure chemical vapor deposition (LPCVD) technology to fill 1.0 μm thick silicon oxide in the trench, and remove the excess silicon oxide material by chemical mechanical polishing (CMP) technology to make the silicon wafer surface flat; Step five: transfer the pattern on the N-well mask to the wafer by micro- lithography technology, form the photoresist pattern of N-well, and the non-N-well area retains the photoresist; Inject high-energy phosphorus ions to form a local N-well, and remove the photoresist; Step six: anneal the ion implantation area, use rapid thermal annealing technology (RTA) to repair ion implantation damage and activate dopants, effectively reduce dopant diffusion, and thus accurately control the doping concentration and the doping distribution of the shallow layer area, the temperature is controlled at 1000-1300℃, and the duration is 10-60 seconds; Step seven: grow a thick gate oxide layer, grow a layer of thick silicon dioxide by dry oxidation, the temperature is 900℃, deposit 200nm thick polysilicon by LPCVD, form the gate by photoetching and etching, and remove the photoresist; Step eight: coat photoresist on the silicon wafer to define the first and second N+ heavily doped regions, use a mask and ultraviolet light to expose the photoresist to form a pattern, inject heavily doped arsenic ions to form N+ heavily doped regions, and remove the photoresist; Step nine: coat photoresist on the silicon wafer to define the first to fifth P+ heavily doped regions, use a mask and ultraviolet light to expose the photoresist to form a pattern, inject heavily doped boron ions to form P+ heavily doped regions, remove the photoresist, and anneal to repair the silicon surface crystal damage caused by ion implantation and eliminate impurity migration in the heavily doped region.

Images