Design method of self-focusing pixel with high detection efficiency based on area array SPAD

Abstract

The present invention discloses a high-detection-efficiency self-focusing pixel design method based on a planar array SPAD. The method includes the following steps: Step 1, design of a spherical doping avalanche multiplication region and an absorption region; Step 2, design of an adjustable self-focusing auxiliary electric field; Step 3, crosstalk isolation design between structural pixels. By designing the electric field distribution inside the pixel, this method converges the photo-generated carriers to the avalanche multiplication region, enabling the light absorption region to almost cover the entire pixel, significantly improving the fill factor and the photon detection efficiency. At the same time, a spherical doping is used to form a small-size avalanche amplification region in combination with a doping structure in the form of N+-P-N-P, separating the absorption region and the avalanche amplification region, effectively reducing the avalanche breakdown voltage and the dark count, and achieving the goal of improving the photon detection efficiency without deteriorating the dark count performance.

Description

Technical Field

[0001] The present invention relates to a pixel design method based on area array SPAD image sensing and 3D detection applications, and in particular to a silicon-based back-illuminated area array SPAD pixel design method with high detection efficiency based on a self-focusing auxiliary electric field. Background Art

[0002] With the development of the information industry, people have begun to use optoelectronic devices with photon detection capabilities to receive photons emitted by target objects, generate the position and shape characteristics of target objects, and thus obtain the detection technology of the target's photon image. This technology is single-photon detection imaging. At present, this technology has been widely used in high-tech fields. It has been applied in biochip detection, medical diagnosis, astronomical observation, quantum electronics and other fields, and plays an increasingly important role. However, with the further development of science and technology, ordinary optical imaging technology has become more mature. Traditional photon imaging has been subject to certain limitations in extremely weak light and high frequency, but detection technology for extremely weak light environments has a wide range of application needs. Although in recent years, with the advancement of semiconductor processing and packaging technology, the back-illuminated (BSI) CMOS image sensor obtained by 3D packaging has excellent performance indicators, it still has the disadvantages of poor weak light performance, slow response speed, and difficult noise elimination of traditional charge accumulation CMOS sensors. SPAD image sensors based on silicon-based CMOS technology have the same advantages as traditional CMOS image sensors, such as easy integration, small size, and mass production. In addition, due to their completely different working principles of avalanche signal collection and pulse digital signal output, they have extremely excellent low-light performance, extremely high response speed, and extremely low output noise. However, the existing SPAD pixel structure still has the core problems of low photon detection efficiency (PDE), high dark count rate (DCR), high after-pulse probability (AP), crosstalk between pixels (X-talk), and low fill efficiency (FF) at high pixel density, which limits the integrated application of SPAD in large-scale arrays. Therefore, it is of great significance to propose a SPAD pixel design with low avalanche breakdown voltage, miniaturization, and high detection efficiency. Summary of the invention

[0003] In order to solve the problems of reduced filling factor, low photon detection efficiency and high dark count rate of existing SPAD pixels in large-scale array applications, the present invention provides a method for designing a high detection efficiency self-focusing pixel based on array SPAD. This method collects photogenerated carriers to the avalanche multiplication region by designing the electric field distribution inside the pixel, so that the light absorption region almost covers the entire pixel, significantly improving the filling efficiency and photon detection efficiency. At the same time, spherical doping is used to form a small-sized avalanche amplification region combined with N +The -PNP doping structure separates the absorption region and the avalanche amplification region, effectively reducing the avalanche breakdown voltage and dark counts, achieving the goal of improving the photon detection efficiency without deteriorating the dark count performance.

[0004] The objective of the present invention is achieved through the following technical solutions:

[0005] A method for designing a high detection efficiency self-focusing pixel based on a planar array SPAD comprises the following steps:

[0006] Step 1: Design of spherical doping avalanche multiplication region and absorption region

[0007] The spherical doped avalanche multiplication region is composed of a high-concentration N-type sphere doped region for generating ohmic contact with the cathode and forming a PN junction, and a P-type sphere doped region for forming a PN junction; the high-concentration N-type sphere doped region is connected to the cathode working electrode, and the P-type sphere doped region is connected to the anode working electrode; the high-concentration N-type sphere doped region is an N-type sphere doped semiconductor, and the P-type sphere doped region is a P-type sphere doped semiconductor;

[0008] The absorption region is composed of an N-type substrate and a bottom highly doped P+ region; the N-type substrate is an N-type doped semiconductor formed by epitaxial growth, and the bottom highly doped P+ region is a P-type doped semiconductor;

[0009] Step 2: Design of adjustable self-focusing auxiliary electric field

[0010] The adjustable self-focusing auxiliary electric field is composed of a vertical auxiliary electrode V connected to the bottom of the pixel. vertical Connect the cathode and anode to different potentials, V vertical >V anode >V cathode , where: V anode is the anode electrode potential, V cathode is the cathode electrode potential;

[0011] Step 3: Crosstalk Isolation Design between Structural Pixels

[0012] The crosstalk isolation between the structural pixels is achieved by deep buried metal grooves and deep buried oxide grooves; the deep buried metal groove runs through the entire pixel, and deep buried oxide grooves are arranged on both sides of the deep buried metal groove, and the self-focusing auxiliary electrode is connected to the N-type substrate through the deep buried metal groove.

[0013] An optimization method for the high detection efficiency self-focusing pixel based on the above-mentioned area array SPAD comprises the following steps:

[0014] Step 1: Determine the pixel depth and width of the SPAD based on the sensor size and number of pixels and the wavelength range of the detected light;

[0015] Step 2: Regulate the P-type doping concentration of the spherical doped avalanche multiplication region to optimize the avalanche breakdown voltage of the SPAD;

[0016] Step 3: Regulate the vertical auxiliary electric field to optimize the photon detection efficiency and time jitter;

[0017] Step 4: Regulate the vertical auxiliary electric field to optimize the time jitter characteristics.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] The present invention introduces a self-focusing auxiliary electric field into the structural design of the planar array pixel. At the same time, the active area doping adopts a hemispherical structure, which reduces the avalanche breakdown voltage and does not require a guard ring structure, which is beneficial to the miniaturization of the pixel. It realizes more effective aggregation of photogenerated carriers in the pixel to the avalanche multiplication region, solves the difficult problem of low photon detection efficiency when the traditional SPAD pixel is miniaturized, and at the same time separates the avalanche multiplication region and the absorption region, limiting the avalanche multiplication region to a smaller volume, thereby achieving high photon detection efficiency of the pixel while having a low dark count rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of the planar structure of the self-focusing pixel of the silicon-based area array SPAD provided by the present invention;

[0021] Figure 2 The relationship between the doping concentration of the second semiconductor in the avalanche multiplication region of the SPAD pixel provided by the present invention and the avalanche breakdown voltage;

[0022] Figure 3 The relationship between the doping concentration of the second semiconductor in the avalanche multiplication region of the SPAD pixel provided by the present invention and the longitudinal electric field distribution;

[0023] Figure 4 is the relationship between the doping concentration of the second semiconductor in the avalanche multiplication region of the SPAD pixel provided by the present invention and the maximum electron avalanche triggering probability;

[0024] Figure 5 The relationship between the SPAD pixel over-bias voltage and the maximum electron avalanche probability and the maximum electric field strength in the multiplication region provided by the present invention;

[0025] Figure 6 is the relationship between the potential of the SPAD pixel vertical auxiliary electrode and the PDE provided by the present invention;

[0026] Figure 7 The relationship between the cathode potential and PDE of the SPAD pixel provided by the present invention when the vertical auxiliary electrode potential is -8V;

[0027] Figure 8 The relationship between the doping concentration of the second semiconductor in the absorption region of the SPAD pixel provided by the present invention and the photon detection efficiency;

[0028] Fig. 9 is the relationship between the potential of the SPAD pixel vertical auxiliary electrode and the time jitter characteristic provided by the present invention;

[0029] Fig.10 It is a schematic diagram of crosstalk between pixels when the SPAD pixel provided by the present invention has no isolation structure;

[0030] Fig.11 It is a schematic diagram of crosstalk between pixels when the SPAD pixel provided by the present invention has an isolation structure;

[0031] Fig.12 Schematic diagram of potential distribution at SPAD pixel XX' provided by the present invention;

[0032] Fig.13 It is a schematic diagram of the potential distribution at the SPAD pixel Y-Y' provided by the present invention. DETAILED DESCRIPTION

[0033] The technical solution of the present invention is further described below in conjunction with the accompanying drawings, but is not limited thereto. Any modification or equivalent replacement of the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention should be included in the protection scope of the present invention.

[0034] The present invention designs a silicon-based back-illuminated area array SPAD self-focusing pixel structure, such as Figure 1As shown, region 1 is an N-type spherical doped semiconductor in a spherical doped avalanche multiplication region with a doping radius of R1; region 2 is a P-type spherical doped semiconductor in a spherical doped avalanche multiplication region with a doping radius of R2; region 3 is an N-type doped semiconductor in an absorption region, formed by epitaxial growth, with a thickness of H1; region 4 is a P-type doped semiconductor in an absorption region, with a doping depth of H2; region 5 is a deeply buried oxide trench between SPAD pixels, with a depth of H1 and a width of W1; region 6 is a deeply buried metal trench, with a depth of H1+H2 and a width of W2, running through the entire pixel; region 7 is a cathode working electrode, connected to region 1; region 8 is an anode working electrode, connected to region 2; region 9 is a self-focusing auxiliary electrode, connected to region 4 through region 6; W3 is the width of the P region, and X-X' and Y-Y' are the two ends of the horizontal section and the vertical section, respectively. The present invention utilizes the potential difference formed between the self-focusing auxiliary electrode and the anode in region 9 to form an electric field distribution from the avalanche multiplication region to the edge of the pixel inside the pixel, and gathers the photogenerated electrons to the avalanche multiplication region, so that the light absorption region almost covers the entire pixel, significantly improving the filling efficiency and photon detection efficiency. At the same time, spherical doping is used to form the active region, avoiding the restrictions of the guard ring structure on the SPAD performance and size miniaturization while reducing the avalanche breakdown voltage. The pixel as a whole adopts an N+-PNP doping structure to separate the absorption region and the avalanche amplification region, effectively reducing the dark count rate, and achieving the goal of improving the photon detection efficiency without deteriorating the dark count performance. The specific design steps are as follows:

[0035] Step 1: Design of spherical doping avalanche multiplication region and absorption region

[0036] The spherical doped avalanche multiplication region is composed of a high concentration (concentration greater than or equal to 6×10 18 cm -3 ) N-type ball doping area, P-type ball doping area for forming PN junction (concentration is less than or equal to 1×10 17 cm -3 ) composition, spherical doping is more evenly distributed than planar doping, no guard ring structure is required, and the avalanche breakdown voltage is lower under the same PN junction doping concentration conditions.

[0037] The absorption region is composed of an N-type substrate and a highly doped P+ region at the bottom. In order to better absorb the light in the short-wave frequency band, the P+ doping concentration in the absorption region should be greater than or equal to 1×10 17 cm -3 When the vertical auxiliary electrode V vertical When connected to a negative potential, the space charge region formed by the N-type substrate and the P+ region widens, and the photogenerated carriers are efficiently detected in the entire region. This is equivalent to the working principle of traditional PIN detection. The photogenerated carriers generated in the entire pixel will drift in the electric field to the avalanche multiplication region, triggering the avalanche effect and being amplified and read out.

[0038] Step 2: Design of adjustable self-focusing auxiliary electric field

[0039] The adjustable self-focusing auxiliary electric field is composed of a vertical auxiliary electrode V connected to the bottom of the pixel. vertical The cathode and anode are connected to different potentials. When working, V vertical Connect to a lower potential, the anode is connected to zero potential or a lower potential, and the cathode is connected to a higher potential, that is, V vertical >V anode >V cathode , where: V anode is the anode electrode potential, V cathode This electric field distribution can gather the photogenerated electrons generated in the entire pixel volume to the avalanche multiplication region near the cathode, significantly improving the photon detection efficiency of the SPAD pixel. At the same time, in this structure, the avalanche multiplication region and the absorption region are separated. The volume of the avalanche multiplication region is very small and its electric field is mainly regulated by the external bias voltage applied by the cathode and anode, and is not affected by the auxiliary electric field. Therefore, the dark count rate can be effectively suppressed while ensuring the large detection efficiency of the pixel.

[0040] Step 3: Crosstalk Isolation Design between Structural Pixels

[0041] The crosstalk isolation between the structural pixels is achieved through deep buried metal and oxide grooves. The deep buried metal full groove isolation structure not only provides a vertical auxiliary electric field by connecting to the back P+ area, but also runs through the entire pixel and carries a negative potential, which will drive electrons with the same charge to converge to the middle of the pixel, which helps to increase the efficiency of photon detection. At the same time, the metal has a large reflection coefficient for light. The light that is obliquely incident into the pixel will be reflected again through the metal isolation layer into the photosensitive area of ​​the pixel to be detected. This feature can increase the detection efficiency of the pixel on the one hand, and on the other hand, it can also well isolate the optical crosstalk between pixels. The role of the oxide isolation layer is to isolate the metal isolation layer and the photosensitive area of ​​the pixel to avoid the uncontrollable electric field distribution in the photosensitive area. Degraded performance, at the same time, the structure can well isolate the electronic crosstalk between pixels.

[0042] An optimization method for the high detection efficiency self-focusing pixel based on the above-mentioned area array SPAD comprises the following steps:

[0043] Step 1: Determine the pixel depth and width of the SPAD based on the sensor size and number of pixels and the wavelength range of the detected light;

[0044] Step 2: Regulate the P-type doping concentration in the avalanche multiplication region to optimize the avalanche breakdown voltage of the SPAD;

[0045] Step 3: Regulate the vertical auxiliary electric field to optimize the photon detection efficiency and time jitter. The photon detection efficiency can be calculated by the quantum efficiency η and the avalanche multiplication probability P. ava And the filling factor FF multiplied together, the expression is as follows:

[0046]

[0047]

[0048]

[0049] Where: V b is the bias voltage, V br is the avalanche breakdown voltage.

[0050] In TCAD software, the cathode current I Cathode and the light source current I s The ratio of the external quantum efficiency is η = I Cathode / I S By adding the Geiger model to the TCAD software, the avalanche multiplication probability P in the avalanche zone can be extracted. ava ;

[0051] Step 4: Regulate the vertical auxiliary electric field to optimize the time jitter characteristics. The time jitter characteristics are obtained by transient response simulation in TCAD software.

[0052] Traditional SPADs all use a planar doping process. The rectangular structure brought about by planar doping will cause premature breakdown due to the tip effect of the edge, which will affect the uniformity of the electric field distribution in the avalanche multiplication zone of the SPAD, and thus make the multiplication zone of the SPAD unable to work in an avalanche state, affecting the performance of the SPAD. In order to avoid premature edge breakdown, SPADs made using standard CMOS processes generally use a guard ring structure. The existing guard ring structure either limits the miniaturization of the SPAD pixel size or limits the performance of the SPAD. Setting the highly doped active area of ​​the SPAD into a spherical structure can, on the one hand, eliminate the instability factors caused by edge pre-breakdown. On the other hand, unlike typical SPADs, the multiplication field is not only formed by the strong built-in electric field brought about by the high doping in the PN junction, but also due to the geometric shape of the active area. The density of the electric field lines is the highest near the surface of the spherical structure, so the effect of field line crowding forms a strong electric field on the surface of the spherical active area, inducing an avalanche effect. Therefore, the spherical active area has a lower avalanche breakdown voltage than the traditional planar active area under the same conditions. The theoretical formula is as follows:

[0053] Electric field distribution in spherical active area:

[0054]

[0055]

[0056] Electric field distribution in the planar active area:

[0057]

[0058]

[0059] Where Ψ is the potential function, ρ is the charge density, ε is the dielectric constant, ε0 is the vacuum dielectric constant, e is the unit electron charge, N D is the impurity doping concentration, R D is the boundary radius of the spherical doping depletion region, r j is the radius of the PN junction interface, x D is the boundary position of the depletion region in a planar doped semiconductor. The ratio of the maximum electric field strength between the two is:

[0060]

[0061] It can be concluded from the above formula that, under the same conditions, the avalanche breakdown voltage of the spherical doping structure is lower than that of the planar doping structure.

[0062] Example:

[0063] This embodiment is mainly aimed at the application of imaging area array image sensors. To achieve more than three million pixels in an area of ​​one inch, the pixel size must be less than 6.4μm, so the pixel size (including the isolation structure) in this embodiment is 5.5μm, which can achieve about four million pixels in an area of ​​one inch. In terms of longitudinal size, because the wavelength of visible light after passing through the Bayer filter is mainly concentrated in 400-800nm, according to the absorption curve of light in silicon, the absorption depth of light with a wavelength of 800nm ​​in silicon is about 10μm, so the longitudinal size of the pixel in this embodiment is 10μm. SPAD pixels of other sizes can also be optimized by adopting the model of the present invention.

[0064] The self-focusing pixel structure based on silicon-based area array SPAD involved in this embodiment is as follows: Figure 1 As shown in the figure, 1 is the N-type doped semiconductor in the avalanche multiplication region. In order to reduce the avalanche breakdown voltage and increase the electric field strength in the avalanche multiplication region, the doping concentration in region 1 should be set to 1×10 17 / cm 3 or higher, in this embodiment 1×10 17 / cm 3 2 is a P-type doped semiconductor in the avalanche multiplication region. The relationship between its doping concentration and avalanche breakdown voltage is as follows: Figure 2 As shown in the figure, the relationship between the doping concentration and the longitudinal electric field distribution is as follows: Figure 3As shown in the figure, the relationship between the doping concentration and the electron avalanche triggering probability in the avalanche multiplication region is as follows: Figure 4 3 is an N-type doped semiconductor in the absorption region, which is similar to the intrinsic doping of silicon. The doping concentration in this embodiment is 1×10 14 / cm 3 ; 4 is the P-type doped semiconductor in the absorption region. The doping concentration in this region should be 1×10 17 / cm 3 or higher, in this embodiment, the initial value is 1×10 19 / cm 3 ; 5 is a deep buried oxide trench between SPAD pixels, SiO2 in this embodiment, 6 is a deep buried metal trench, Cu in this embodiment; 7 is a cathode working electrode; 8 is an anode working electrode; 9 is a self-focusing auxiliary electrode. It can be changed according to different pixel sizes. In this embodiment, R1 = 0.5 μm, R2 = 1.2 μm, H1 = 9.5 μm, H2 = 0.5 μm, W1 = 0.2 μm, and W2 = 0.3 μm.

[0065] pass Figure 3 It can be observed that as the doping concentration of the second-type semiconductor in the avalanche region increases, the high electric field distribution range becomes smaller and the undepleted portion becomes larger. During the photoelectric detection process, electrons will pass through this undepleted region. On the one hand, this region is connected to the zero potential through the anode, and the movement speed of electrons in this region will slow down; on the other hand, the recombination probability of electrons and the majority carrier holes will increase, thereby affecting the photon detection efficiency. In addition, as the doping concentration in this region gradually increases, the probability of electron avalanche triggering does not increase monotonically, but increases at a rate of 5×10 16 / cm 3 The peak value is reached near the P region, and then the avalanche triggering probability decreases first and then increases with the increase of the doping concentration in the P region. The specific relationship is as follows: Figure 4 As shown in the figure, this is because the probability of electron avalanche breakdown is related to many factors such as semiconductor doping concentration, electric field strength, and structural size. When the doping concentration increases, the depletion region shifts toward N+, and the avalanche region becomes smaller. Although the electric field strength is very high at this time, the avalanche effect requires a certain distance to accumulate. Therefore, the probability of electron avalanche breakdown does not increase monotonically with the increase of doping concentration or electric field strength. From the simulation results, it can be seen that in this embodiment, when the doping concentration is 5×10 16 / cm 3 When , it is most conducive for electrons to produce avalanche effect.

[0066] Figure 5is the relationship between the over-bias voltage and the maximum electron avalanche triggering probability and the maximum electric field strength in the multiplication region. As the over-bias voltage increases, the electric field strength and the electron avalanche triggering probability in the avalanche multiplication region of the SPAD pixel gradually increase. As can be seen from the figure, when the over-bias voltage reaches 3.5V, the electric field strength is very close to 7×10 5 V / cm. According to the BTBT theory, when the electric field strength is greater than 7×10 5 V / cm, the strong electric field will introduce noise caused by band-to-band tunneling, increase the dark count rate of SPAD, and affect the working performance. Therefore, in this embodiment, the overbias voltage V ex =3.5V, the probability of electron avalanche at this time is 86%.

[0067] First, the Atlas software was used to simulate the device and read the designed device structure. A light source with a wavelength of 300nm to 1000nm was added to the back of the SPAD pixel through the optical module. The cathode and anode were set to zero potential at the same time. Potentials of 0V, -2V, -4V, -6V, and -8V were applied to the vertical auxiliary electrode respectively. The source photocurrent and cathode current were extracted. The corresponding PDE under each potential condition was calculated using formula (1), as follows: Figure 6 As shown. After adding a vertical auxiliary electric field to the pixel structure, the PDE is significantly improved, and it increases with the decrease of the auxiliary electrode potential (the photogenerated carriers are electrons). In order to be closest to the actual working situation, the photon detection efficiency when the P-type spherical well is completely depleted is simulated. Next, the potential of the vertical auxiliary electrode is kept at -8V, and 0V, 2V, 4V, 6V, 8V, and 10V bias voltages are applied to the cathode respectively. The photon detection efficiency under different cathode bias voltages is simulated and calculated as shown below. Figure 7 As shown, when the cathode bias is less than 10V, for example 8V, the depletion layer is fully widened and the electric field strength does not reach the avalanche effect threshold, indicating that this is closest to the actual working state of SPAD, and the photon detection efficiency at this time is 75.6%.

[0068] Figure 8 is the relationship between the doping concentration of the second-type semiconductor in the absorption region and the photon detection efficiency. It can be seen from the figure that as the doping concentration of the P+ region decreases, the detection efficiency of the SPAD pixel for photons in the short-wave range is significantly improved, from 1×10 19 / cm 3 The 40% of the time is increased to 1×10 17 / cm 3 About 85% of the time.

[0069] Fig. 9This is the relationship between the vertical auxiliary electrode potential and the time jitter characteristic of the SPAD pixel in this embodiment. As the vertical auxiliary electrode potential decreases, the time jitter decreases. When the potential is -24V, the time jitter parameter is reduced to 262ps. However, at this time, because the electric field strength in the pixel absorption area is very large, the dark current value also increases. Under the premise of not increasing noise interference, when the vertical auxiliary potential is -20V, the time jitter is 524ps. At this time, the time jitter mentioned is the maximum value. In general research, the SPAD time jitter parameter is expressed by FWHM (full width at half maximum). At this time, the SPAD pixel time jitter parameter is converted into FWHM to represent about 262ps.

[0070] Fig.10 Schematic diagram of crosstalk between pixels when there is no isolation structure. Fig.11 The schematic diagram of crosstalk between pixels when there is an isolation structure. The same working conditions are applied to three connected pixels, and a light source window is set within the middle structure. The wavelength range of the light source is 300nm to 1300nm, and the light intensity is 0.1W / cm 2 , rename the cathodes of the left and right pixels to Cathode_left and Cathode_right, detect the currents of the three electrodes respectively, and obtain the current distribution curves of the three electrodes under different wavelength conditions. Fig.11 It can be seen that for the pixel structure optimized without isolation structure, under the illumination of light sources with different wavelengths, both left and right electrodes detected a certain amount of photocurrent, and the order of magnitude is not much different from that of the middle pixel electrode with direct light, indicating that when there is no isolation structure, the crosstalk between pixels is very obvious, with a crosstalk rate of up to 92%, which will have a great negative impact on the work of SPAD, especially when large-scale array SPAD is used in image sensors, the crosstalk between pixels will seriously affect the imaging quality such as color accuracy. Fig.11 It can be clearly seen that after adding the isolation structure optimization, the left and right electrodes of the pixel can hardly detect any photocurrent signal, and the crosstalk between pixels is as low as 0.2%. Moreover, due to the reflection of the metal isolation layer and the self-focusing auxiliary electric field applied by it as a vertical auxiliary electrode, the photocurrent detected by the electrode of the middle pixel is much larger than the photocurrent detected by the electrode of the middle pixel without the isolation structure. Therefore, the simulation results well illustrate that the optimization of the crosstalk between SPAD pixels by the oxide plus metal isolation structure is very obvious, and the overall crosstalk is reduced by 99.8% compared with the pixels without the isolation structure.

[0071] Fig.12 for Figure 1 Schematic diagram of the potential distribution along the tangent line X-X' in the middle. It can be seen from the figure that the potential at both ends of the SPAD pixel is low and high in the middle (here taking electrons as signal charges as an example), which is conducive to the convergence of photogenerated electrons to the middle of the pixel. Fig.13 for Figure 1Schematic diagram of the potential distribution along the Y-Y' tangent line in the middle. It can be seen from the figure that the potential near the cathode in the SPAD pixel is high, and the potential on the back is low, which is conducive to the photogenerated electrons converging to the avalanche multiplication zone near the cathode and triggering the avalanche effect to be detected.

[0072] This embodiment proposes a new SPAD pixel structure based on CMOS technology, which can maintain a high photon detection efficiency, a low dark count rate and an operating voltage when the pixel is miniaturized, and at the same time well limits the time jitter and crosstalk between pixels. At the same time, a self-focusing auxiliary electric field model that can improve the photon detection efficiency of the array SPAD pixel is constructed. It can effectively improve the photon detection efficiency and time resolution capability of the SPAD pixel by adjusting the electric field distribution in the absorption zone without affecting the electric field in the avalanche multiplication zone.

Claims

1. A high detection efficiency self-focusing pixel design method based on area array SPAD, characterized in that The method comprises the following steps: Step 1: Design of spherical doping avalanche multiplication region and absorption region The spherical doped avalanche multiplication region is composed of a high-concentration N-type sphere doped region for generating ohmic contact with the cathode and forming a PN junction, and a P-type sphere doped region for forming a PN junction; the high-concentration N-type sphere doped region is connected to the cathode working electrode, and the P-type sphere doped region is connected to the anode working electrode; the high-concentration N-type sphere doped region is an N-type sphere doped semiconductor, and the P-type sphere doped region is a P-type sphere doped semiconductor; The absorption region is composed of an N-type substrate and a bottom highly doped P+ region; the N-type substrate is an N-type doped semiconductor formed by epitaxial growth, and the bottom highly doped P+ region is a P-type doped semiconductor; Step 2: Design of adjustable self-focusing auxiliary electric field The adjustable self-focusing auxiliary electric field is composed of a vertical auxiliary electrode V connected to the bottom of the pixel. vertical Connect the cathode and anode to different potentials, V vertical >V anode >V cathode , where: V anode is the anode electrode potential, V cathode is the cathode electrode potential; Step 3: Crosstalk Isolation Design between Structural Pixels The crosstalk isolation between the structural pixels is achieved by deep buried metal grooves and deep buried oxide grooves; the deep buried metal groove runs through the entire pixel, and deep buried oxide grooves are arranged on both sides of the deep buried metal groove, and the self-focusing auxiliary electrode is connected to the N-type substrate through the deep buried metal groove.

2. The high detection efficiency self-focusing pixel design method based on area array SPAD according to claim 1 is characterized in that In the N-type ball doping region, the doping concentration is greater than or equal to 6×10 18 cm -3 ; In the P-type ball doping area, the doping concentration is less than or equal to 1×10 17 cm -3 .

3. The high detection efficiency self-focusing pixel design method based on area array SPAD according to claim 1 is characterized in that In the bottom highly doped P+ region, the P+ doping concentration should be greater than or equal to 1×10 17 cm -3 .

4. An optimization method for high detection efficiency self-focusing pixel based on area array SPAD designed by the method described in any one of claims 1 to 3, characterized in that The method comprises the following steps: Step 1: Determine the pixel depth and width of the SPAD based on the sensor size and number of pixels and the wavelength range of the detected light; Step 2: Regulate the P-type doping concentration of the spherical doped avalanche multiplication region to optimize the avalanche breakdown voltage of the SPAD; Step 3: Regulate the vertical auxiliary electric field to optimize the photon detection efficiency and time jitter; Step 4: Regulate the vertical auxiliary electric field to optimize the time jitter characteristics.

5. The optimization method of high detection efficiency self-focusing pixel based on area array SPAD according to claim 4 is characterized in that The photon detection efficiency is expressed as the quantum efficiency η and the avalanche multiplication probability P ava And the filling factor FF multiplied together, the expression is as follows: PDE=η×P ava ×FF。 6. The optimization method of high detection efficiency self-focusing pixel based on area array SPAD according to claim 5 is characterized in that The expression of the quantum efficiency η is as follows: Where: V b is the bias voltage, V br is the avalanche breakdown voltage.

7. The optimization method of high detection efficiency self-focusing pixel based on area array SPAD according to claim 5 is characterized in that The quantum efficiency η is calculated in TCAD software by the cathode current I Cathode and the light source current I s The ratio is obtained, that is, η=I Cathode / I S ; By adding the Geiger model in TCAD software, the avalanche multiplication probability P in the avalanche zone is extracted ava .

Images