CN102623476B - Self-powered CMOS (Complementary Metal-Oxide-Semiconductor Transistor) image sensor - Google Patents

Abstract

The present invention provides a self-powered CMOS image sensor, which at least includes a substrate layer, a photosensitive laminate stacked on the substrate layer, and a pixel readout unit, wherein, a column bus, an amplifier, a current source unit, an image processing unit, and The energy storage unit in the self-supply source is located in the substrate layer; the photosensitive layer includes a plurality of photosensitive layers stacked on the substrate layer for absorbing light of different wavelength bands, and each photosensitive layer includes at least two switching elements and A photosensitive element connected between the two; the pixel readout unit is located in the substrate layer or in at least one photosensitive layer in the photosensitive stack. The present invention adopts a photosensitive layer with a layered structure without a filter, which can realize the ability of a pixel to detect light in different bands and the self-powering function of the image sensor, which not only reduces the cost and complexity, but also greatly improves the conversion of the image sensor. Efficiency, image quality and integration, and enhance the adaptability of the image sensor while making full use of the energy of light.

Description

Self-Powered CMOS Image Sensor

technical field

The invention relates to a CMOS image sensor, in particular to a self-powered CMOS image sensor, belonging to the technical field of semiconductors.

Background technique

As we all know, an image sensor is a semiconductor device that converts an optical image into an electrical signal. Image sensors can be roughly classified into Charge Coupled Device (CCD) and Complementary Metal Oxide Semiconductor (CMOS) image sensors.

A CMOS image sensor is generally composed of a photosensitive element and a CMOS signal processing circuit. The current common CMOS image sensor is an active pixel image sensor (APS), which is mainly divided into reset transistor (Reset Transistor, RST), source follower transistor (Source Follower Transistor, SF) and row selection according to the number of transistors it includes. Transistor (Row Select, RS) three-tube image sensor (3T type) and four-tube image sensor including reset transistor (RST), source follower transistor (SF), row select transistor (RS) and transfer transistor (Transfer Transistor, TX) There are two types of sensors (type 4T).

On the one hand, when the existing CMOS image sensor detects visible light, it usually uses red, green, and blue color filters to block other light from passing through, and then detects the light of the corresponding color through the image sensor (usually a photodiode or CCD). . This kind of image sensor with red, green, and blue color filters, which is widely used to detect natural light, has at least the following disadvantages: First, since a pixel is composed of three or more image sensors, resulting in image The sensor area is large, which will inevitably lead to poor image quality; secondly, the performance of the filter itself will change with the change of temperature and time, which will affect the quality of imaging; finally, due to the use of filter slices, resulting in an increase in cost. Although the current prism splitting 3CCD technology can solve the above-mentioned problems, the cost increase it brings is also quite huge.

On the other hand, the existing CMOS image sensors usually detect ultraviolet light, visible light, and infrared light separately, and there is no image sensor that can simultaneously detect several different wavebands of light.

Finally, the existing CMOS image sensor cannot provide power for itself while detecting light to achieve a self-powered function.

Contents of the invention

In view of the above-mentioned shortcomings of the prior art, the object of the present invention is to provide a self-powered CMOS image sensor, which is used to solve the problem that the image sensor for detecting natural light in the prior art uses a filter and the area of the image sensor is relatively large and the filter The performance change of itself leads to the deterioration of image quality and the problem of increasing the cost of the image sensor. It is also used to solve the problem that different image sensors need to be used in the prior art to realize the detection of ultraviolet light, visible light, and infrared light. The problem with technologies that do not enable self-powered image sensors.

To achieve the above purpose and other related purposes, the present invention provides a self-powered CMOS image sensor, comprising at least:

The substrate layer at least includes a column bus, an amplifier, a current source unit for providing a constant current, an image processing unit for image processing, and an energy storage unit for self-supply power, wherein the image processing unit is at least Including clock and control circuit, row, column, layer decoding circuit, sample and hold circuit, analog-to-digital converter, image processor; the energy storage unit includes a charge storage circuit and a power conversion circuit to pass the stored charge through the power supply The conversion circuit is converted into a suitable voltage to provide power for the CMOS image sensor; the amplifier and the current source unit are connected to the column bus;

The photosensitive layer is laminated and covers the surface of the substrate layer to simultaneously absorb light of different wavelength bands, at least including multiple photosensitive layers laminated and covered on the substrate layer for absorbing the light of each of the different wavelength bands, and each of the photosensitive The layer includes at least two switching elements and a photosensitive element connected between the two switching elements;

The pixel readout unit is located in the substrate layer or in at least one photosensitive layer in the photosensitive stack, at least including a reset transistor connected to the photosensitive layer, a source follower transistor connected to the reset transistor and the photosensitive layer, and connected to The source follows a row select transistor of the transistor.

Optionally, the photosensitive layer at least includes a first photosensitive layer stacked on the substrate layer for absorbing light in the first wavelength band, and a photosensitive layer stacked on the first photosensitive layer for absorbing light in the second wavelength band. The second photosensitive layer for light.

Optionally, the photosensitive laminate further includes a third photosensitive layer laminated on the second photosensitive layer for absorbing light of a third wavelength band.

Optionally, germanium is doped in the first photosensitive layer to form a photosensitive element of germanium material for absorbing infrared light, and silicon material for absorbing visible light is formed in the second photosensitive layer by doping silicon The photosensitive element, the silicon carbide material photosensitive element used to absorb ultraviolet light is formed by doping silicon carbide in the third photosensitive layer.

Optionally, the material of each photosensitive layer is a semiconductor material whose bandgap width matches the photon energy required to be absorbed, and is selected from silicon carbide, silicon, germanium, or silicon germanium, or from doped silicon carbide , silicon, germanium, or silicon germanium.

Optionally, each photosensitive layer in the photosensitive stack is made of the same semiconductor material.

Optionally, there is an insulating layer between the substrate layer and the photosensitive stack, or/and an insulating layer between the photosensitive layers in the photosensitive stack.

Optionally, the photosensitive layer in the photosensitive stack is a composite photosensitive layer composed of multiple photosensitive sub-layers.

Optionally, there is an insulating layer between each of the photosensitive sub-layers.

Optionally, each photosensitive element of the photosensitive stack is at least one of a PN junction photodiode, a pinned photodiode, or a photogate.

Optionally, the pixel readout unit is located in a photosensitive layer of the photosensitive stack, all photosensitive elements of the photosensitive stack share the pixel readout unit, and each photosensitive element is connected to the pixel readout unit. The output units are connected through a switch element corresponding to each photosensitive element, wherein the photosensitive stack and the pixel readout unit form a pixel module.

Optionally, a plurality of the pixel readout units are respectively located in the plurality of photosensitive layers of the photosensitive stack, all the photosensitive elements of the photosensitive stack correspond to the plurality of pixel readout units, and the Each photosensitive element of the photosensitive stack is only connected to one pixel readout unit through a corresponding switch element, wherein the photosensitive stack and a plurality of pixel readout units form a pixel module.

Optionally, the pixel readout unit is located in the substrate layer, and all photosensitive elements of the photosensitive stack share the pixel readout unit, wherein the photosensitive stack is a pixel module.

As mentioned above, compared with the existing CMOS image sensor, the self-powered CMOS image sensor of the present invention has the following beneficial effects:

1) Since the present invention adopts the design without optical filter, problems such as image quality deterioration caused by optical filters are avoided while reducing costs;

2) The present invention can make the required circuit structure on a single layer (substrate layer) owing to adopting the photosensitive lamination layer of laminated structure, thereby greatly improving the filling factor and further improving the quality of the image;

3) The present invention adopts a stacked structure of photosensitive layers, so that one pixel has the ability to detect light in different bands, and information such as ultraviolet light, visible light, and infrared light can be obtained at the same time in one exposure. On the one hand, a variety of Information, not only reduces the cost and complexity, but also greatly improves the conversion efficiency and image quality; on the other hand, while reading the required information, it can selectively save the remaining charges generated by different photosensitive layers, And through the power conversion circuit to provide power for the entire image sensor to realize the function of self-supply, make full use of the energy of light, and enhance the adaptability of the image sensor;

4) The self-powered CMOS image sensor provided by the present invention can be used to realize precise imaging of visible light, and the red, blue, and green photosensitive devices are prepared on different stacks located in the same area of the semiconductor substrate, which further improves the integration of the image sensor Spend.

Description of drawings

FIG. 1 is a schematic diagram of the equivalent circuit structure of the self-powered CMOS image sensor in Embodiment 1 of the present invention.

2a to 2e are three-dimensional structural schematic diagrams of different photosensitive stacks of the self-powered CMOS image sensor of the present invention.

FIG. 3 is a schematic diagram showing the relationship between the intensity of light in different wavelength bands and the depth into the material in the self-powered CMOS image sensor of the present invention.

4a to 4b are schematic diagrams of equivalent circuit structures corresponding to pixel modules of different compositions in the self-powered CMOS image sensor of the present invention.

FIG. 5 is a schematic diagram of an equivalent circuit structure diagram of the second embodiment of the self-powered CMOS image sensor of the present invention.

Component designation description

1 Substrate layer

11 Column bus

12 Amplifiers

13 Current source unit

14 Image processing unit

15 Energy storage unit

2 Photosensitive Lamination

21 Infrared photosensitive layer

22 Visible light photosensitive layer

22a, 22b, 22c Photosensitive sublayer of visible light photosensitive layer

23 UV photosensitive layer

21’ Red photosensitive layer

22’ Green photosensitive layer

23’ blue light photosensitive layer

212, 222, 232, 212’, 222’, 232’ PN junction photodiode, photosensitive element

2111, 2112, 2211, 2212, 2311, 2312 Switching elements

2111’, 2112’, 2211’, 2212’, 2311’, 2312’ switching elements

3. 3’ Pixel Readout Unit

31, 31’ reset transistor

32, 32’ Source follower transistor

33, 33’ row selection transistor

341, 342, 343 Transfer Transistor

351, 352, 353 Floating diffusion area

41, 421, 422, 422a～c, 41″, 421″, 422″ insulation layer

5. 5’ Pixel Module

6. 6’ Pixel module array

Detailed ways

Embodiments of the present invention are described below through specific examples, and 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 implementation modes, and various modifications or changes can be made to the details in this specification based on different viewpoints and applications without departing from the spirit of the present invention.

See Figures 1 through 5. It should be noted that the illustrations provided in the following specific embodiments are only schematically illustrating the basic idea of the present invention, and only the components related to the present invention are shown in the drawings rather than the number and shape of components in actual implementation. and size drawing, the type, quantity and proportion of each component can be changed arbitrarily during actual implementation, and the component layout type may also be more complex.

The existing CMOS image sensor for detecting visible light is an image sensor with three color filters of red, green and blue. Due to the use of the filter, the cost increases, the area of the image sensor is larger, and the image quality deteriorates. , and the performance change of the filter itself also affects the quality of imaging; the existing CMOS image sensors usually detect ultraviolet light, visible light, and infrared light separately, and there is no one that can simultaneously detect several different bands of light. image sensor; the existing CMOS image sensor cannot provide power for itself to achieve the function of self-supply while detecting light.

In view of this, the present invention provides a self-powered CMOS image sensor, comprising at least a substrate layer, a photosensitive stack layer stacked on the substrate layer, and a pixel readout unit, wherein the column bus, amplifier, current source unit, image processing unit , and the energy storage unit for self-power supply is located in the substrate layer; the photosensitive layer includes a plurality of photosensitive layers stacked on the substrate layer for absorbing light of different wavelengths, and each of the photosensitive layers includes at least two A switch element and a photosensitive element connected between the two; the pixel readout unit is located in the substrate layer or in at least one photosensitive layer in the photosensitive stack. The present invention adopts a photosensitive layer with a layered structure without a filter, which can realize the ability of a pixel to detect light in different bands and the self-powering function of the image sensor, which not only reduces the cost and complexity, but also greatly improves the conversion of the image sensor. Efficiency, image quality and integration, and enhance the adaptability of the image sensor while making full use of the energy of light.

Embodiment one

As shown in Figures 1 to 4, the present invention provides a self-powered CMOS image sensor, comprising at least:

The column bus 11, the amplifier 12, the current source unit 13, the image processing unit 14, and the energy storage unit 15 located in the substrate layer 1; the photosensitive layer 2 stacked on the substrate layer 1 is used to absorb light of different wavelengths A plurality of photosensitive layers, in the first embodiment, are three photosensitive layers for absorbing light in three wavelength bands, wherein, the first photosensitive layer is an infrared photosensitive layer 21, and the second photosensitive layer is a visible light photosensitive layer 22, The third photosensitive layer is an ultraviolet photosensitive layer 23; and a pixel readout unit 3 comprising at least a reset transistor 31, a source follower transistor 32, and a row selection transistor 33. In the first embodiment, the self-powered CMOS image sensor is 4T type, the pixel readout unit 3 further includes transfer transistors 341-343 and floating diffusion regions 351-353 respectively located in the photosensitive layers.

The material of the substrate layer 1 is a semiconductor material or an organic thin film material, wherein the semiconductor material is silicon, germanium, or silicon germanium, the organic thin film material is ZnO-TFT, and the structure of the substrate layer 1 is a single material A semiconductor structure or a semiconductor structure with an insulating buried layer. In the first embodiment, the substrate layer 1 is bulk silicon, but it is not limited thereto. In another embodiment, the substrate layer 1 is germanium-on-insulator ( GOI).

As shown in Figure 1, the substrate layer 1 at least includes:

The column bus 11 is connected to the amplifier 12 , the current source unit 13 and the pixel readout unit 3 .

The amplifier 12 is used to amplify the electrical signal containing pixel information read out by the pixel readout unit 3, wherein one input terminal is connected to the column bus, and the other input terminal is grounded, and the amplified signal containing The electrical signal of the pixel information is output to the image processing unit 14 .

The current source unit 13 includes a circuit (not shown) for providing a constant current, so that the gate-source current (I _ds ) of the source-following transistor 32 in the pixel readout unit 3 is kept constant, so as to ensure the source The difference between the voltage and the gate voltage is constant, so that the source voltage follows the gate voltage. One end of the current source unit 13 is connected to the column bus 11 and the amplifier 12 , and the other end is grounded.

Described image processing unit 14 is connected with the output terminal of described amplifier 12, at least comprises clock and control circuit, row, column, layer decoding circuit, sample and hold circuit, analog-to-digital converter, image processor (not shown), uses Image processing is performed with the electrical signal containing pixel information amplified by the amplifier 12 .

The energy storage unit 15 includes a charge storage circuit and a power conversion circuit (not shown), which are used to convert the stored charge into a suitable voltage through the power conversion circuit and store it. When self-supply is required, the stored charge The voltage is connected to the power supply V _dd required by the pixel readout unit 3 to provide the drain voltage V _dd for the pixel readout unit 3 and realize the self-power supply function for the CMOS image sensor.

As shown in Figure 2a, the photosensitive layer 2 is laminated and covered on the substrate layer 1, and at least includes a plurality of photosensitive layers laminated and covered on the substrate layer 1 for absorbing light of different wavelength bands. In this embodiment In one, the first photosensitive layer stacked on the substrate layer for absorbing light in the first wavelength band (infrared light) is an infrared photosensitive layer 21, and is laminated on the first photosensitive layer (infrared photosensitive layer 21) The second photosensitive layer for absorbing light of the second wavelength band (visible light) is the visible light photosensitive layer 22, and the second photosensitive layer (visible light photosensitive layer 22) for absorbing the third wavelength band light (ultraviolet light) is laminated on the second photosensitive layer (visible light photosensitive layer 22). ) The third photosensitive layer is the ultraviolet light photosensitive layer 23, and the material of each photosensitive layer is a semiconductor material with a bandgap width and the photon energy to be absorbed, which are respectively selected from silicon carbide, silicon, germanium, or silicon germanium, or They are respectively selected from doped silicon carbide, silicon, germanium, or silicon germanium, wherein, in the first embodiment, the infrared photosensitive layer 21 is germanium (Ge) material, and the visible light photosensitive layer 22 is silicon ( Si) material, the ultraviolet photosensitive layer 23 is silicon carbide (SiC) material, but not limited thereto, in another embodiment, the materials of each photosensitive layer can be the same semiconductor material.

Each of the photosensitive layers includes at least two switching elements and a photosensitive element connected between the two, and each of the photosensitive elements is at least one of a PN junction photodiode, a pinned photodiode (Pinned Diode), or a photogate. One, wherein the photosensitive element is used in the image sensor of the present invention to perform photoelectric conversion during exposure, and convert the received optical signal into an electrical signal containing pixel information. In the first embodiment, each of the photosensitive layers The photosensitive elements are all PN junction photodiodes, but it is not limited thereto. In another embodiment, there are also cases where the types of photosensitive elements in each photosensitive layer are different, that is, PN junction photodiodes, pin-pricked photodiodes, A type of photodiode and a photogate (not shown); the switching elements all include an NMOS transistor, a plurality of NMOS transistors connected in parallel or in series, a PMOS transistor, or a plurality of PMOS transistors connected in parallel or in series. In the first embodiment, the switch element is an NMOS transistor.

It should be noted that the number of photosensitive layers in the photosensitive laminated layer 2 in the first embodiment is three (the first photosensitive layer is the infrared photosensitive layer 21, the second photosensitive layer is the visible light photosensitive layer 22, the third photosensitive layer layer is an ultraviolet photosensitive layer 23), but not limited thereto, in another embodiment, the number of photosensitive layers in the photosensitive laminated layer 2 can also be two for absorbing light of two different wavelength bands. A photosensitive layer, which includes a first photosensitive layer stacked on the substrate layer for absorbing light in the first wavelength band and a second photosensitive layer stacked on the first photosensitive layer for absorbing light in the second wavelength band; Of course, in other embodiments, the number of photosensitive layers in the photosensitive stack 2 may also be four photosensitive layers or more, which will not be repeated here.

It should be further explained that the photosensitive layer in the photosensitive laminate 2 can be a composite photosensitive layer composed of multiple photosensitive sub-layers, in the first embodiment, that is, the first photosensitive layer (infrared photosensitive layer 21), the second photosensitive layer (visible light photosensitive layer 22), or the third photosensitive layer (ultraviolet photosensitive layer 23) can be a composite photosensitive layer composed of multiple photosensitive sub-layers, for example, as shown in Figure 2b This is only the case where the visible light photosensitive layer 22 is a composite photosensitive layer composed of a plurality of photosensitive sublayers ( 22 a to 22 c ).

It should be pointed out that in order to avoid crosstalk between the photosensitive layers and affect the image quality, an insulating layer is used between the substrate layer 1 and the photosensitive stack 2, or/and each photosensitive layer in the photosensitive stack 2 An insulating layer is used to isolate the layers; it should be further pointed out that the insulating layer 4 is a single-layer structure or a laminated structure, and the material of each layer in the single-layer structure or the laminated structure is Any one of silicon oxide, silicon nitride, and silicon oxynitride. As shown in Figure 2c, in the first embodiment, a single-layer insulating layer 41 is used between the substrate layer 1 and the photosensitive stack 2, and an insulating layer ( Including the single-layer insulating layer 421 between the infrared light-sensitive layer 21 and the visible light-sensitive layer 22, and the single-layer insulating layer 422 between the visible light-sensitive layer 22 and the ultraviolet light-sensitive layer 23. At this time, each The photosensitive layer does not contain a photosensitive sublayer, and is a single-layer structure) for isolation; in another embodiment, when the photosensitive layer in the photosensitive laminate 2 is a composite photosensitive layer composed of a plurality of photosensitive sublayers , there is an insulating layer between each of the photosensitive sublayers. For example, only the visible light photosensitive layer 22 shown in FIG. There are single-layer insulating layers 422a-422c between the sub-layers (22a-22c); further, in other embodiments, as shown in FIG. All are single-layer structures, but the insulating layers 41 ″, 421 ″ and 422 ″ are insulating layers of a laminated structure of silicon nitride and silicon oxynitride.

Specifically, in the first embodiment, as shown in FIG. 2c, the first photosensitive layer (infrared photosensitive layer 21), the second photosensitive layer (visible light photosensitive layer 22) and the third photosensitive layer (ultraviolet photosensitive layer Layer 23) is a single-layer structure, and an insulating layer (including the insulating layer 421 between the infrared photosensitive layer 21 and the visible light photosensitive layer 22, and the visible light photosensitive layer 22 and the visible light photosensitive layer 22) is used between each photosensitive layer. The insulating layer 422) between the ultraviolet photosensitive layers 23 is isolated, and at the same time, the insulating layer 41 is used to isolate between the substrate layer 1 and the photosensitive laminated layer 2, and the insulating layers 41, 421 and 422 are all single layer structure of silicon nitride insulating layer; as shown in Figure 2c, in the present embodiment one, the photosensitive elements in each of the photosensitive layers are PN junction photodiodes; as shown in Figure 1, in the present embodiment one, The switching elements are all NMOS transistors. When a high voltage is applied to their gates, the switching elements are closed and in a conducting state, and when their gates are grounded, the switching elements are open and in an off state.

Further, in the first embodiment, as shown in FIG. 1 , the infrared photosensitive layer 21 includes switching elements 2111 and 2112 and a photosensitive element (PN junction photodiode) 212 connected therebetween, wherein, In the infrared photosensitive layer 21, a germanium PN junction photodiode photosensitive element 212 is formed by doping germanium, including a P-type region and an N-type region, the P-type region is grounded, and one end of the switching element 2111 is connected to The pixel readout unit 3 is connected, one end of the switching element 2112 is connected to the energy storage unit 15, and in the first embodiment, the other end of each switching element 2111 and 2112 is connected to the infrared light source. The floating diffusion region (Floating Diffusion, FD) 351 in the photosensitive layer 21 (in the first embodiment, the image sensor of the present invention is a 4T type, and the N-type region of the photosensitive element (PN junction photodiode) 212 passes through the transfer transistor (Transfer Transistor, TX ) 341 is connected to the floating diffusion region 351), but it is not limited thereto. In another embodiment, when the image sensor of the present invention is a 3T type, the other end of each of the switching elements 2111 and 2112 is connected to a PN junction photodiode 212 N-type region; in the first embodiment, as shown in FIG. 1 , the visible light photosensitive layer 22 includes switching elements 2211 and 2212 and a photosensitive element (PN junction photodiode) 222 connected therebetween , wherein, in the visible light photosensitive layer 22, a silicon PN junction photodiode photosensitive element 222 is formed by doping silicon, including a P-type region and an N-type region, the P-type region is grounded, and the switch element 2211 One end is connected to the pixel readout unit 3, one end of the switching element 2212 is connected to the energy storage unit 15, and in the first embodiment, the other ends of the switching elements 2211 and 2212 are connected to the The floating diffusion region (Floating Diffusion, FD) 352 in the visible light photosensitive layer 22 (in the first embodiment, the image sensor of the present invention is a 4T type, and the N-type region of the photosensitive element (PN junction photodiode) 222 passes through the transfer transistor (Transfer Transistor, TX) 342 is connected to the floating diffusion region 352), but it is not limited thereto. In another embodiment, when the image sensor of the present invention is a 3T type, the other end of each of the switching elements 2211 and 2212 is connected to the PN junction photoelectric The N-type region of diode 222; In the present embodiment one, as shown in Figure 1, comprise switch elements 2311 and 2312 and a photosensitive element (PN junction photodiode) that is connected between the two in the described ultraviolet photosensitive layer 23 ) 232, wherein, in the ultraviolet photosensitive layer 23, a silicon carbide PN junction photodiode photosensitive element (PN junction photodiode) 232 is formed by doping silicon carbide (SiC), including a P-type region and an N-type region , the P-type region is grounded, and one end of the switching element 2311 is connected to the pixel readout unit 3, one end of the switching element 2312 is connected to the energy storage unit 15, and in the first embodiment, the other end of each switching element 2311 and 2312 is connected to the Floating Diffusion (FD) 353 (In Embodiment 1, the image sensor of the present invention is a 4T type, and the N-type region of the photosensitive element (PN junction photodiode) 232 is connected to the floating diffusion through a transfer transistor (Transfer Transistor, TX) 343 Diffusion region 353), but not limited thereto, in another embodiment, when the image sensor of the present invention is a 3T type, the other end of each of the switching elements 2311 and 2312 is connected to the N-type region of the PN junction photodiode 232 .

It should be pointed out that light of different wavelengths has different response characteristics in the same semiconductor material. Only when the semiconductor energy band width E _G is smaller than the photon energy E _ph of the incident light (E _ph =hυ=hc/λ), the semiconductor Only when photons are absorbed and electron-hole pairs are generated, the strength of different semiconductor materials for different wavelengths of light absorption can be expressed by the absorption coefficient a, P(x)=P ₀ ×exp(-ax), where P ₀ is the incident light Intensity, P(x) is the light intensity at the optical path x in the semiconductor, when x=1, P(d)=P ₀ ×e, and d is defined as the photon penetration depth. The absorption coefficient a=a(λ) is a function of the wavelength, it strongly depends on the wavelength, and the absorption coefficient increases rapidly with the decrease of the wavelength, that is, to absorb light of different wavelengths, the semiconductor material needs different incident light paths and wavelengths The shorter the length, the greater the absorption coefficient of the semiconductor material for light. Completely absorbing the incident light and realizing photoelectric conversion requires the shorter the effective optical path of the incident light transmitted in the semiconductor, and vice versa. In other words, when incident lights of different wavelengths are absorbed in the same semiconductor material, the incident lights of different wavelengths enter the semiconductor material at different depths.

In order to facilitate the simultaneous absorption of light of different wavebands, so that the image sensor of the present invention has the ability to detect light of different wavebands, the photosensitive laminate 2 used to absorb light in the present invention adopts a laminated structure. The positions in layer 2 are different, so when the photosensitive layers are made of the same semiconductor material, they can also absorb light of different wavelength bands.

As shown in Figure 3, in the first embodiment, ultraviolet light, visible light and infrared light are lights with three different wavelength bands, wherein the ultraviolet light with the shortest wavelength is almost completely absorbed at the _d1 position closer to the incident surface, Visible light is almost completely absorbed at the d ₂ position relatively centered from the incident surface (d ₂ is located in the middle position relative to d ₁ and d ₃ ), while the infrared light with the relatively longest wavelength among the three bands is mainly absorbed farther away from the incident surface The d ₃ position is almost completely absorbed. In view of the above-mentioned theory, in the first embodiment, the ultraviolet photosensitive layer 23 is laminated on the visible light photosensitive layer 22, and is located in the upper part of the photosensitive laminate 2, so as to transmit ultraviolet light away from the incident surface (photosensitive laminate 2 The nearer positions of the surface) are almost all absorbed, and the distance _r3 between the bottom of the ultraviolet light-sensitive layer 23 and the incident surface (the surface of the photosensitive laminate 2) (or the ultraviolet light-sensitive layer is a composite photosensitive layer, and its sublayers The sum of the thicknesses) is between 0 and _d2 ; the visible light photosensitive layer 22 is laminated on the infrared photosensitive layer 21, and is located in the middle layer part of the photosensitive laminate 2, so as to transmit visible light at a distance from the incident surface (photosensitive laminate The relatively central position of the surface of layer 2) is almost completely absorbed, and the distance r ₂ from the bottom of the visible light photosensitive layer 22 to the incident surface (the surface of the photosensitive laminate 2) (or the visible light photosensitive layer is a composite photosensitive layer, and its sublayers The sum of the thicknesses) is between _r3 and _d3 ; the infrared photosensitive layer 21 is laminated and covered on the substrate layer 1, and is located in the lower part of the photosensitive layer 2, so as to separate the infrared light from the incident surface (photosensitive layer) The far position of the surface of the layer 2) is almost completely absorbed, and the distance _r1 between the bottom of the infrared photosensitive layer 21 and the incident surface (the surface of the photosensitive laminate 2) (or the infrared photosensitive layer is a composite photosensitive layer, each of which The sum of the thicknesses of the sublayers) is greater than r ₂ .

Similarly, in another embodiment, for visible light, since the wavelengths of blue light, green light and red light are increasing, the blue light is absorbed in the area closest to the incident surface (the surface of the photosensitive stack 2), and the red light is absorbed The absorption of green light is concentrated in the inner region farther away from the incident surface (the surface of the photosensitive laminate 2), and the absorption region of green light is between the two.

It should be further explained that, usually when the band gap of a semiconductor matches the photon energy of a certain wavelength band of light, the light of this wavelength band can be well absorbed. Therefore, the material of each photosensitive layer in the photosensitive laminated layer 2 is The forbidden band width matches the photon energy required to absorb the semiconductor material, wherein, in the first embodiment, in order to better absorb ultraviolet light, the ultraviolet photosensitive layer 23 adopts a material silicon carbide with a relatively large forbidden band width; For good absorption of visible light, the visible light photosensitive layer 22 is made of silicon with a medium band gap; in order to better absorb infrared light, the infrared photosensitive layer 21 is made of germanium with the smallest band gap.

The pixel readout unit 3 (not shown) is located in the substrate layer 1 or in at least one photosensitive layer in the photosensitive stack 2, and is used to convert the electrical signal containing pixel information obtained by the photosensitive stack 2 into (Each photosensitive element performs photoelectric conversion during exposure to convert the received light signal into an electrical signal containing pixel information), and is read out to the image processing unit 14 through the column bus 11 and the amplifier 12 under the influence of the current source unit 13, at least including A reset transistor 31 (ResetTransistor, RST) connected to the photosensitive layer, a source follower transistor 32 (Source FollowerTransistor, SF) connected to the reset transistor 31 and the photosensitive layer, and a row selection transistor 33 (SF) connected to the source follower transistor 32 Row Select, RS), wherein, the reset transistor 31, the source follower transistor 32, and the row selection transistor 33 all include an NMOS transistor, a plurality of parallel or series NMOS transistors, a PMOS transistor, or a plurality of parallel or series A type of PMOS transistor.

It should be pointed out that the self-powered CMOS image sensor is a 3T-type image sensor or a 4T-type image sensor, wherein the pixel readout unit 3 (not shown) of the 4T-type image sensor also includes The transfer transistors (Transfer Transistor, TX) 341-343 and floating diffusion regions (Floating Diffusion, FD) 351-353 corresponding to the respective photosensitive elements, wherein the transfer transistors include one NMOS transistor, a plurality of NMOS transistors connected in parallel or in series , a PMOS transistor, or one of a plurality of PMOS transistors connected in parallel or in series.

In the first embodiment, the self-powered CMOS image sensor is a 4T image sensor, and the reset transistor 31 , the source follower transistor 32 , the row selection transistor 33 , and the transfer transistor 34 are all an NMOS transistor. But not limited thereto, in another embodiment, the transistors in the pixel readout unit 3 can be the above-mentioned series/parallel connection or PMOS.

It should be pointed out that the pixel readout unit 3 is located in the substrate layer 1 or in at least one photosensitive layer in the photosensitive stack 2, as shown in Figures 1, 4a and 4b, there are the following situations:

As shown in Figure 1, the pixel readout unit 3 is located in one photosensitive layer of the photosensitive stack 2, all the photosensitive elements of the photosensitive stack 2 share the pixel readout unit 3, and each photosensitive The element is connected to the pixel readout unit 3 through a switch element corresponding to each photosensitive element, wherein the photosensitive stack 2 and the pixel readout unit 3 form a pixel module 5;

As shown in Figure 4a, the pixel readout unit 3 is located in the substrate layer 1, and all the photosensitive elements of the photosensitive stack 2 share the pixel readout unit 3, wherein the photosensitive stack 2 is a pixel module 5;

As shown in Figure 4b, a plurality of the pixel readout units 3 are respectively located in the plurality of photosensitive layers of the photosensitive stack 2, and the electrical signals containing pixel information of all the photosensitive elements of the photosensitive stack 2 pass through the The multiple pixel readout units 3 mentioned above are read out, and each photosensitive element of the photosensitive stack 2 is only connected to one of the pixel readout units 3 through its respective corresponding switch element, wherein the The photosensitive stack 2 and the plurality of pixel readout units 3 form a pixel module 5 . In other words, in the plurality of pixel readout units 3 respectively located in the plurality of photosensitive layers of the photosensitive stack 2, in one embodiment, as shown in FIG. 4b, each of the pixel readout units 3 Corresponding to one photosensitive element, that is, each of the photosensitive elements is connected to the corresponding pixel readout unit 3 through its respective corresponding switching element; in another embodiment, there is a pixel unit 3 corresponding to a plurality of photosensitive elements In this case, each of the photosensitive elements that share one pixel readout unit 3 is connected to the corresponding shared pixel readout unit 3 through a corresponding switch element.

It should be further pointed out that a plurality of the pixel modules 5 form a pixel module array 6 (not shown), and the electrical signal containing pixel information of the pixel module array 6 passes through the pixel readout unit 3, column The bus 11 and the amplifier 12 are transmitted to the image processing unit 14 to perform image processing on the electrical signal including pixel information.

Specifically, in the first embodiment, as shown in FIG. 1, the pixel readout unit 3 is located in a photosensitive layer of the photosensitive stack 2, wherein the photosensitive layer is a visible light photosensitive layer 22 (not in Semiconductor structure Figure 2c shows the specific structure of the pixel readout unit 3), all photosensitive elements (PN junction photodiodes 212-232) of the photosensitive stack 2 share the pixel readout unit 3, and the self-powered CMOS The image sensor is a 4T-type image sensor, and the floating diffusion regions 351-353 corresponding to each of the photosensitive elements (PN junction photodiodes 212, 222, and 232) are connected to the pixel readout unit 3 through the corresponding photosensitive elements. A switching element 2111, 2211 and 2311 are connected, that is, the floating diffusion region 351 corresponding to the PN junction photodiode 212 is connected to the pixel readout unit 3 through the switching element 2111, and the PN junction photodiode 222 corresponds to The floating diffusion region 352 is connected to the pixel readout unit 3 through the switch element 2211 , and the floating diffusion region 353 corresponding to the PN junction photodiode 232 is connected to the pixel readout unit 3 through the switch element 2311 . The photosensitive stack 2 and the pixel readout unit 3 form a pixel module 5, and further, a plurality of the pixel modules 5 form a pixel module array 6 (not shown), and the pixel module array 6 includes pixels The electrical signal of information is transmitted to the image processing unit 14 through the pixel readout unit 3 , the column bus 11 and the amplifier 12 to perform image processing on the electrical signal containing pixel information.

It should be specially noted that in the first embodiment, the self-powered CMOS image sensor is a 4T image sensor, and the photosensitive elements (PN junction photodiodes 212, 222 and 232) of each photosensitive layer share the pixel readout. Output unit 3, as shown in Figure 1, the connection mode of each of the transistors in the pixel readout unit 3 is as follows:

The reset transistor 31 is used for the image sensor of the present invention to reset the photosensitive elements (PN junction photodiodes 212, 222 and 232) before exposure. Photodiodes 212, 222 and 232) are respectively connected to the floating diffusion regions 351-353 corresponding to the reset transistor 31 through the switching elements 2111, 2211 and 2311 respectively, and the reset function is controlled by the reset signal Reset signal (in FIG. 1 RST endpoint) for control. In FIG. 1, the reset transistor 31 is an NMOS transistor. Since the self-powered CMOS image sensor is a 4T image sensor, the source of the reset transistor 31 is in phase with one end of each of the switching elements 2111, 2211 and 2311. connected, and the other end of each of the switching elements 2111, 2211 and 2311 is respectively located in each of the photosensitive layers corresponding to the respective photosensitive elements (PN junction photodiodes 212, 222 and 232) floating diffusion regions (FD) 351, 352 and 353, that is, each of the switching elements 2111, 2211, and 2311 respectively connects the source of the reset transistor 31 to the floating diffusion of the respective photosensitive elements (PN junction photodiodes 212, 222, and 232) in each of the photosensitive layers. Fields (FD) 351 , 352 and 353 are connected; the drain of the reset transistor 31 is connected to the power supply V _dd , and the power supply V _dd is a positive power supply.

The source follower transistor 32 keeps its gate-source current (I _ds ) constant by means of the constant current provided by the current source unit 13, so as to ensure that the difference between the source voltage and the gate voltage is constant, so as to achieve the source voltage following the gate The purpose of the electrode voltage is to read out the electrical signal including pixel information generated by the photosensitive elements (PN junction photodiodes 212 , 222 and 232 ) through photoelectric conversion. In FIG. 1 , the source follower transistor 32 is an NMOS transistor, and since the self-powered CMOS image sensor is a 4T image sensor, the gate of the source follower transistor 32 is connected to each of the photosensitive layers described above. The corresponding floating diffusion regions 351, 352 and 353 of the photosensitive elements (PN junction photodiodes 212, 222 and 232) are respectively connected through the switching elements 2111, 2211 and 2311. At this time, the source follows the gate of the transistor 32 The sources of the reset transistors 31 are all connected to one end of each of the switch elements 2111 , 2211 and 2311 . The drain of the source-following transistor 32 is connected to the power supply V _dd , and the source of the source-following transistor 32 is an output terminal for outputting an electrical signal containing pixel information.

The row selection transistor 33 is used to output the electrical signal including pixel information output by the source of the source follower transistor 32 . In FIG. 1, the row selection transistor M3 is an NMOS transistor, the gate of the row selection transistor 33 is connected to the row selection signal RS, and the drain of the row selection transistor 33 is connected to the source of the source follower transistor 32. , the source of the row selection transistor 33 is the output terminal.

The transfer transistors 341, 342 and 343 are used to input the electrical signals including pixel information generated by the PN junction photodiodes 212, 222 and 232 to the The source follower transistor 32 is connected to the source of the reset transistor 31 . In the first embodiment, as shown in FIG. 1, the transfer transistors 341, 342 and 343 are NMOS transistors, the gates of the transfer transistors 341, 342 and 343 are connected to the transfer signal TX, and the transfer transistors 341, 342 The sources (not shown) of the transistors 341 and 343 are connected to the N-type regions of the PN junction photodiodes 212, 222 and 232, and the drains of the transfer transistors 341, 342 and 343 are connected to the sources through the floating diffusion regions 351, 352 and 353. Follow the connection between the gate of the transistor 32 and the source of the reset transistor 31 .

Before explaining the working principle, it should be noted that since each of the switching elements is an NMOS transistor, when the gate of the switching element is at a high level, the switching element is in a conducting state (that is, the switch is closed). When grounded, the switching element is in the off state (that is, the switch is open). Switching elements 2111, 2211, and 2311 select different photosensitive layers to read out electrical signals containing pixel information, and switching elements 2112, 2212, and 2312 select and export the photoelectrically converted charges of different photosensitive layers to the energy storage unit 15 , converted to a suitable voltage by the power conversion circuit and stored. When self-supply is required, the stored voltage is connected to the power supply V _dd required by the pixel readout unit 3 to provide drain for the pixel readout unit 3. The pole voltage V _dd realizes the self-power supply function for the CMOS image sensor.

The self-powered CMOS image sensor will be described by taking only the electrical signal containing pixel information of the ultraviolet photosensitive layer 23 to be read, and the charge photoelectrically converted by the visible light photosensitive layer 22 and the infrared photosensitive layer 21 through their respective photosensitive elements to be stored. How it works:

Select natural light as the light source, and before exposure, ground the gate of the switching element 2312 to make it in an off state (that is, the switch is turned on), then the corresponding photosensitive element (PN junction photodiode) in the ultraviolet photosensitive layer 23 The floating diffusion (FD) 353 of 232) is disconnected from the energy storage unit 15, and does not store the photoelectrically converted charge of the ultraviolet photosensitive layer 23, and at the same time, provides a high voltage for the gate of the switching element 2311 Make it in the conduction state (that is, the switch is closed), then the floating diffusion region (FD) 353 corresponding to the photosensitive element (PN junction photodiode 232) in the ultraviolet photosensitive layer 23 is connected to the source of the reset transistor 31 When the reset signal Reset of the reset transistor 31 is high level, the reset transistor 31 is turned on, and the PN junction photodiode 232 passes through the corresponding transfer transistor 343 (its gate is connected to the transfer signal TX, its The source is connected to the N-type region of the PN junction photodiode 232 (not shown in the figure) and the floating diffusion region 353 is connected to the drain power supply V _dd of the reset transistor 31. Under the action of the power supply V _dd , the PN The junction photodiode 232 is reverse-biased and will clear all accumulated charges of the PN junction photodiode 232 to realize reset.

The drain of the source-follower transistor 32 is connected to the power supply _Vdd . When the self-powered CMOS image sensor is reset, natural light is selected as the light source for exposure. The photosensitive element (PN junction photodiode 232 ) performs photoelectric conversion, and converts the received optical signal into an electrical signal containing pixel information. Since the switching element 2312 is in an off state (the gate of the switching element is grounded) and the switching element 2311 is in a conducting state (the gate of the switching element is connected to high Level), then through the transfer transistor 343 corresponding to the PN junction photodiode 232 (its gate is connected to the transfer signal TX, and its source is connected to the N-type region of the PN junction photodiode 232, not shown) and the floating diffusion region 353 is connected to the gate of the source-following transistor 32, and under the action of the power supply _Vdd and the current source unit 13, the source of the source-following transistor 32 outputs an electrical signal including pixel information to the row selection transistor 33 DRAIN.

The gate of the row selection transistor 33 is connected to the row selection signal RS, and its drain is an input terminal connected to the source of the source follower transistor 32 , wherein the source is an output terminal. The row selection transistor 33 is used to output the electrical signal including pixel information output by the source of the source follower transistor 32 to the column bus 11 .

At the same time, when natural light is selected as the light source for exposure, the gates of the switching elements 2211 and 2111 are grounded to make them in an off state (that is, the switch is turned on), and a high level is provided for the gates of the switching elements 2212 and 2112 to make them on. state (that is, the switch is closed), the PN junction photodiodes 222 and 212 corresponding to the visible light photosensitive layer 22 and the infrared light photosensitive layer 21 are photoelectrically converted to obtain charges, which are exported to the energy storage unit 15 through the switch elements 2212 and 2112 .

Each photosensitive layer of the photosensitive stack 2 and the shared pixel readout unit 3 form a pixel module 5 , and a plurality of the pixel modules 5 form a pixel module array 6 (not shown). The electrical signals containing pixel information of each ultraviolet photosensitive layer 23 of the pixel module array 6 are transmitted to the image processing unit 14 through the pixel readout unit 3, the column bus 11 and the amplifier 12, so as to Image processing is carried out on electrical signals containing pixel information; the charge obtained by each visible light photosensitive layer 22 and infrared light photosensitive layer 21 of the pixel module array 6 is exported to the energy storage unit 15 through the switching elements 2212 and 2112, and passed through The power conversion circuit wherein is converted to a suitable voltage and stored. When self-supply is required, the stored voltage is connected to the power supply V _dd required by the pixel readout unit 3 to provide a drain for the pixel readout unit 3. The pole voltage V _dd realizes the self-power supply function for the CMOS image sensor.

Similarly, in the first embodiment, if it is only necessary to read the electrical signal containing the pixel information of the visible light sensitive layer 22 or the infrared light sensitive layer 21, and simultaneously store the infrared light sensitive layer 21 and the ultraviolet light sensitive layer 23, or store The charge photoelectrically converted by the corresponding photosensitive elements in the visible light photosensitive layer 22 and the ultraviolet light photosensitive layer 23 only needs to adjust the switching elements 2211, 2112 and 2312 to be in the conduction state (that is, the gate of the switching element is applied with a high voltage) and at the same time the switching element 2212 , 2111 and 2311 are in the off state (that is, the gate of the switching element is grounded), or the switching elements 2111, 2212, and 2312 are adjusted to be in the on state (that is, the gate of the switching element is applied with a high voltage) and at the same time the switching elements 2112, 2211 and 2311 are The off state (that is, the gate of the switching element is grounded) is enough.

The present invention provides a self-powered CMOS image sensor, which adopts the design without optical filter, reduces the cost and avoids the problems of poor image quality caused by the optical filter; The required circuit structure can be made on a separate layer (substrate layer), thereby greatly improving the fill factor and further improving the quality of the image. At the same time, due to the use of a laminated photosensitive layer, a pixel has the ability to detect light in different bands , a single exposure can simultaneously obtain information such as ultraviolet light, visible light and infrared light. On the one hand, a variety of information can be obtained through one chip, which not only reduces the cost and complexity, but also greatly improves the conversion efficiency and image quality. On the other hand, , while reading the required information, you can selectively save the remaining charges generated by different photosensitive layers, and provide power for the entire image sensor through the power conversion circuit, so as to realize the function of self-supply and make full use of the energy of light , and enhanced the adaptability of the image sensor.

Embodiment two

Embodiment 2 adopts basically the same technical solution as Embodiment 1, that is, the components included in the self-powered CMOS image sensor provided by Embodiment 2 and Embodiment 1, the related connection methods between the components and their working principles are basically the same, and the difference is The self-powered CMOS image sensor in the second embodiment is 3T type, and the pixel readout unit 3' is located in the substrate layer, and the photosensitive layer 2' is used to absorb three photosensitive layers of light in three wavelength bands, the first photosensitive layer It is a red photosensitive layer 21', the second photosensitive layer is a green photosensitive layer 22', and the third photosensitive layer is a blue photosensitive layer 23', and each photosensitive layer is made of silicon (Si) material.

It should be pointed out that light of different wavelengths has different response characteristics in the same semiconductor material. To absorb light of different wavelengths, a semiconductor material needs a different incident light path. The shorter the effective optical path of the incident light in the semiconductor, the longer it is vice versa. In order to facilitate the simultaneous absorption of light of different wavelength bands, so that the image sensor of the present invention has the ability to detect light of different wavelength bands, the photosensitive laminated layer used to absorb light in the present invention adopts a laminated structure. The positions of the layers are different, so when the photosensitive layers are made of the same semiconductor material, they can also absorb light of different wavelength bands. For visible light, since the wavelengths of blue light, green light and red light are increasing, the blue light is absorbed in the area closest to the incident surface (the surface of the photosensitive stack), and the absorption of red light is concentrated in the area far from the incident surface (photosensitive stack). The surface of the layer) the far inner region, the absorption region of green light is between the two.

Therefore, in the second embodiment, the first photosensitive layer (red light photosensitive layer 21 ′) for absorbing light in the first wavelength band (red light) is laminated and covered on the substrate layer 1, and is located at the bottom of the photosensitive stacked layer 2 ′. The lowermost layer; the second photosensitive layer (green light photosensitive layer 22') for absorbing light of the second wavelength band (green light) is stacked on the first photosensitive layer (red light photosensitive layer 21'); for absorbing the third The third photosensitive layer (blue light photosensitive layer 23 ′) of wavelength band light (blue light) is stacked on the second photosensitive layer (green light photosensitive layer 22 ′), and is located on the uppermost layer of the photosensitive stack 2 ′.

The equivalent circuit diagram FIG. 5 in the second embodiment is basically the same as the equivalent circuit diagram FIG. 1 in the first embodiment. Compared with the 4T self-powered CMOS image sensor in the first embodiment, the pixel readout unit 3' of the 3T self-powered CMOS image sensor in the second embodiment lacks the transfer transistor (TX) and the floating diffusion region in the first embodiment (FD); Compared with the pixel readout unit 3 in the first embodiment is located in the visible light photosensitive layer 22 , the pixel readout unit 3 ′ in the second embodiment is located in the substrate layer 1 . The specific differences are as follows:

Please refer to FIG. 5. In the second embodiment, the self-powered CMOS image sensor provided by the present invention is a 3T type, and the pixel readout unit 3' includes a reset transistor 31', a source follower transistor 32', and a row selection transistor 33' , while in the second embodiment, the pixel readout unit is located in the substrate layer 1, and all photosensitive elements (PN junction photodiodes 212', 222' and 232') of the photosensitive stack 2' share the pixel readout unit. Unit 3', wherein the photosensitive stack 2' is a pixel module 5'.

As shown in Figure 5, in the second embodiment, the switch elements 2111' and 2112' located in the red photosensitive layer 21' are connected to the N-type region of the PN junction photodiode 212'; Each of the switching elements 2211' and 2212' of the photosensitive layer 22' is connected to the N-type region of the PN junction photodiode 222'; each of the switching elements 2311' and 2312' located in the blue photosensitive layer 23' is connected to the PN junction photodiode N-type region of diode 232'.

Further, as shown in FIG. 5 , in the second embodiment, the source of the reset transistor 31 ′ of the pixel readout unit 3 ′ is connected to one end of each of the switching elements 2111 ′, 2211 ′, and 2311 ′, and The other end of each of the switching elements 2111', 2211' and 2311' is connected to the N of each photosensitive element (PN junction photodiode 212', 222' and 232') in each of the photosensitive layers 21', 22', 23' respectively. type region; the source of the pixel readout unit 3' follows the gate of the transistor 32' and each of the photosensitive elements (PN junction photodiodes 212', 222' and 232') in each of the photosensitive layers are respectively The N-type regions of the N-type region are respectively connected to each of the switching elements 2111', 2211' and 2311'. At this time, the gate of the source follower transistor 32' and the source of the reset transistor 31' are connected to each of the switching elements 2111' , 2211' and one end of 2311' are connected.

In order to further clarify the working principle of the present invention in the second embodiment, please refer to FIG. 5 to only read the electrical signal containing pixel information of the blue photosensitive layer 23', store the green photosensitive layer 22' and the red photosensitive layer 21 'Taking the photoelectrically converted charges of their respective photosensitive elements as an example, the working principle of the self-powered CMOS image sensor in the second embodiment is described:

Selecting natural light as the light source, before exposure, when the switch element 2312' is in the off state (the gate of the switch element is grounded) and the switch element 2311' is in the conduction state (the gate of the switch element is connected to high level), the reset The reset signal Reset of the transistor 31' provides a high level, then the reset transistor 31' is turned on, and the N-type region of the PN junction photodiode 232' is connected to its drain power supply V _dd through the reset transistor 31'. Under the action of the power supply V _dd , the PN junction photodiode 232 ′ is reverse-biased and all accumulated charges of the PN junction photodiode 232 ′ are cleared to realize reset.

The drain of the source-following transistor 32' is connected to the power supply _Vdd . When the self-powered CMOS image sensor is reset, natural light is selected as the light source for exposure. The photosensitive element (PN junction photodiode) located in the blue light-sensitive layer 23'232') to perform photoelectric conversion, and convert the received optical signal into an electrical signal containing pixel information. Since the switching element 2312' is in an off state (the gate of the switching element is grounded) and the switching element 2311' is in an on state (the switching element Gate connected to high level), then the N-type region of the PN junction photodiode 232' is connected to the gate of the source follower transistor 32', under the action of the power supply V _dd and the current source unit 13, the The source of the source follower transistor 32' outputs an electrical signal including pixel information to the drain of the row selection transistor 33'.

The gate of the row selection transistor 33' is connected to the row selection signal RS, and its drain is an input terminal connected to the source of the source follower transistor 32', wherein the source is an output terminal. The row selection transistor 33' is used to output the electrical signal including pixel information output by the source of the source follower transistor 32' to the column bus 11.

At the same time, when natural light is selected as the light source for exposure, the gates of the switching elements 2211' and 2111' are grounded to make them in an off state (that is, the switch is turned on), and a high level is provided for the gates of the switching elements 2212' and 2112' to enable It is in the conduction state (that is, the switch is closed), so that the PN junction photodiodes 222' and 212' corresponding to the green photosensitive layer 22' and the red photosensitive layer 21' undergo photoelectric conversion to obtain charges, and pass through the switching elements 2212' and 2112' Exported to the energy storage unit 15.

Each photosensitive layer of the photosensitive stack 2' and its shared pixel readout unit 3' located on the substrate layer 1 form a pixel module 5', and a plurality of the pixel modules 5' form a pixel module array 6' (not shown icon). The electrical signals containing pixel information of each blue photosensitive layer 23' of the pixel module array 6' are transmitted to the image processing unit 14 through the pixel readout unit 3', the column bus 11 and the amplifier 12, so as to The electric signal containing pixel information is subjected to image processing; the charge obtained by each green photosensitive layer 22' and red photosensitive layer 21' of the pixel module array 6' is exported to the The energy storage unit 15 is converted into a suitable voltage by the power conversion circuit therein and stored. When self-supply is required, the stored voltage is connected to the power supply V _dd required by the pixel readout unit 3', which is The pixel readout unit 3' provides a drain voltage V _dd to realize a self-power supply function for the CMOS image sensor.

Similarly, in the second embodiment, if it is only necessary to read the electrical signal containing pixel information of the green photosensitive layer 22' or the red photosensitive layer 21', and simultaneously store the red photosensitive layer 21' and the blue photosensitive layer 23 ', or store the charge photoelectrically converted by the corresponding photosensitive elements in the green photosensitive layer 22' and the blue photosensitive layer 23', it is only necessary to adjust the switch elements 2211', 2112' and 2312' to be in the conduction state (that is, the gate of the switch element Apply a high voltage) and at the same time switch elements 2212', 2111' and 2311' are in the off state (that is, the gate of the switch element is grounded), or adjust the switch elements 2111', 2212' and 2312' to be in the conduction state (that is, the gate of the switch element A high voltage is applied) and at the same time the switching elements 2112 ′, 2211 ′ and 2311 ′ are turned off (that is, the gates of the switching elements are grounded).

To sum up, the present invention provides a self-powered CMOS image sensor, which adopts a design without a filter, reduces costs and avoids problems such as image quality deterioration caused by filters; Lamination, so that the required circuit structure can be made on a separate layer (substrate layer), thereby greatly improving the fill factor and further improving the quality of the image. With the capability of different bands of light, information such as blue light, green light and red light can be obtained at the same time in one exposure. On the one hand, multiple information can be obtained through one chip, which not only reduces the cost and complexity, but also greatly improves the conversion efficiency and image quality. Quality, on the other hand, while reading out the required information, it can selectively save the residual charge generated by different photosensitive layers, and provide power for the entire image sensor through the power conversion circuit to realize the function of self-supply, Make full use of the energy of light and enhance the adaptability of the image sensor. The self-powered CMOS image sensor provided by the present invention can be used to realize precise imaging of visible light, and the red, blue, and green photosensitive devices are prepared on different stacked layers located in the same area of the semiconductor substrate, which further improves the integration degree of the image sensor. Therefore, the present invention effectively overcomes various shortcomings in the prior art and has high industrial application value.

The above-mentioned embodiments only illustrate the principles and effects of the present invention, but are not intended to limit the present invention. Anyone skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those skilled in the art without departing from the spirit and technical ideas disclosed in the present invention should still be covered by the claims of the present invention.

Claims (12)

1. A self-powered CMOS image sensor, characterized in that it at least includes: The substrate layer at least includes a column bus, an amplifier, a current source unit for providing a constant current, an image processing unit for image processing, and an energy storage unit for self-supply power, wherein the image processing unit is at least Including clock and control circuit, row, column, layer decoding circuit, sample and hold circuit, analog-to-digital converter, image processor; the energy storage unit includes a charge storage circuit and a power conversion circuit to pass the stored charge through the power supply The conversion circuit is converted into a suitable voltage to provide power for the CMOS image sensor; the amplifier and the current source unit are connected to the column bus; The photosensitive layer is laminated and covers the surface of the substrate layer to simultaneously absorb light of different wavelength bands, at least including multiple photosensitive layers laminated and covered on the substrate layer for absorbing the light of each of the different wavelength bands, and each of the photosensitive The layer includes at least two switching elements and a photosensitive element connected between the two switching elements; The pixel readout unit is located in the substrate layer or in at least one photosensitive layer in the photosensitive stack, at least including a reset transistor connected to the photosensitive layer, a source follower transistor connected to the reset transistor and the photosensitive layer, and connected to The row selection transistor of the source follower transistor; wherein, when a plurality of the pixel readout units are respectively located in the plurality of photosensitive layers of the photosensitive stack, all the photosensitive elements of the photosensitive stack correspond to the A plurality of pixel readout units, and each photosensitive element of the photosensitive stack is only connected to one of the pixel readout units through a corresponding switch element, wherein the photosensitive stack is connected to a plurality of the pixel readout units The pixel readout unit forms a pixel module. 2. The self-powered CMOS image sensor according to claim 1, wherein the photosensitive layer at least includes a first photosensitive layer layered on the substrate layer for absorbing light in the first wavelength band, and a layered layer A second photosensitive layer on the first photosensitive layer for absorbing light of the second wavelength band. 3 . The self-powered CMOS image sensor according to claim 2 , wherein the photosensitive layer further comprises a third photosensitive layer laminated on the second photosensitive layer for absorbing light of a third wavelength band. 4 . 4. The self-powered CMOS image sensor according to claim 3, characterized in that: germanium material photosensitive elements for absorbing infrared light are formed in the first photosensitive layer by doping germanium, and the second photosensitive layer In the third photosensitive layer, a silicon carbide photosensitive element for absorbing ultraviolet light is formed by doping silicon carbide. 5. The self-powered CMOS image sensor according to claim 1, characterized in that: the material of each photosensitive layer is a semiconductor material with a band gap matching the photon energy required to be absorbed, and is selected from silicon carbide, silicon, germanium, respectively , or silicon germanium, or respectively selected from doped silicon carbide, silicon, germanium, or silicon germanium. 6 . The self-powered CMOS image sensor according to claim 1 , wherein the photosensitive layers in the photosensitive stack are made of the same semiconductor material. 7. The self-powered CMOS image sensor according to claim 1, characterized in that: there is an insulating layer between the substrate layer and the photosensitive stack, or/and there is an insulating layer between each photosensitive layer in the photosensitive stack layer. 8. The self-powered CMOS image sensor according to claim 1, wherein the photosensitive layer in the photosensitive stack is a composite photosensitive layer composed of multiple photosensitive sub-layers. 9. The self-powered CMOS image sensor according to claim 8, wherein an insulating layer is provided between each of the photosensitive sub-layers. 10 . The self-powered CMOS image sensor according to claim 1 , wherein each photosensitive element of the photosensitive stack is at least one of a PN junction photodiode, a pinned photodiode, or a photogate. 11 . 11. The self-powered CMOS image sensor according to claim 1, wherein the pixel readout unit is located in a photosensitive layer of the photosensitive stack, and all photosensitive elements of the photosensitive stack share the A pixel readout unit, and each photosensitive element is connected to the pixel readout unit through a switch element corresponding to each photosensitive element, wherein the photosensitive stack and the pixel readout unit form a pixel module. 12. The self-powered CMOS image sensor according to claim 1, wherein the pixel readout unit is located in the substrate layer, and all photosensitive elements of the photosensitive stack share the pixel readout unit, wherein, The photosensitive stack is a pixel module.