CN102842590A - Image sensor and manufacturing method thereof - Google Patents

Abstract

The invention provides an image sensor and a manufacturing method thereof, which can improve the efficiency of light absorption. The image sensor is formed in a substrate with an insulating buried layer, the substrate includes the insulating buried layer and a device layer on the surface of the insulating buried layer, the pixel circuit and the photodiode of the image sensor are formed in the device layer, and the A first covering layer is provided on the surface of the photodiode, and a second covering layer is provided on the other surface of the substrate at a position corresponding to the first covering layer, and the second covering layer is a light incident layer. The advantage of the present invention is that, through the multiple reflections of the incident light, it is absorbed multiple times in the photosensitive area, thereby improving the light absorption efficiency of the photosensitive unit based on the image sensor, and further adopting the backside illumination technology also avoids the incident light from being absorbed The shielding of the metal wiring layer improves the quantum efficiency of the image sensor.

Description

Image sensor and manufacturing method thereof

technical field

The invention relates to the field of image sensors, in particular to an image sensor and a manufacturing method thereof.

Background technique

SOI (Silicon-On-Insulator, silicon on insulating substrate) technology introduces a buried oxide layer between the top silicon and the back substrate. By forming a semiconductor thin film on an insulator, SOI material has advantages that cannot be compared with traditional bulk silicon materials: it can realize the dielectric isolation of components in integrated circuits, and completely eliminate the parasitic latch effect in bulk silicon CMOS circuits; using this Integrated circuits made of this material also have the advantages of small parasitic capacitance, high integration density, fast speed, simple process, small short channel effect, and are especially suitable for low-voltage and low-power circuits.

An image sensor is a semiconductor device that converts an optical image into an electrical signal, and is generally composed of a photosensitive pixel and a CMOS signal processing circuit. At present, the common CMOS image sensor is active pixel image sensor (APS), which is divided into three-tube image sensor (3T, including reset transistor, amplification transistor and row selection transistor) and four-tube image sensor (4T, including transfer transistor , Reset transistors, amplifier transistors and row selection transistors) two categories.

An existing CMOS image sensor pixel unit structure fabricated on an SOI substrate is shown in FIG. 1 , which adopts a fully depleted structure, including: a substrate 100 , a buried oxide layer 110 and a device layer 130 . The device layer 130 includes a photodiode 140 , a reset transistor 150 , a source follower transistor 160 and a row select transistor 170 . The photosensitive region of this pixel structure is mainly located in the PN junction depletion region of the photodiode 140 . Each transistor includes basic structures such as source, gate and drain. Please refer to the accompanying drawing 1 for the positional relationship and electrical connection relationship of the above-mentioned components.

Referring to Figure 1, the working principle of the existing pixel structure is as follows: when starting to work, the gate of the reset transistor 150 is first connected to a high level to make it conduction. When , electron-hole pairs are generated, and the integrated voltage signal is read out through the source follower transistor 160 and the row selection transistor 170 after the exposure is completed. Therefore, the value of the output voltage reflects the strength of the optical signal.

The disadvantage of the CMOS image sensor pixel circuit fabricated on the SOI substrate is that the device layer 130 of the SOI substrate is very thin, usually only tens of microns or even a dozen microns, and the optical path of the incident light in the photodiode 140 is very short. , leading to low light absorption efficiency and quantum efficiency. Especially for red and orange light with a wavelength greater than 600nm, the absorption efficiency is extremely low, and the image quality is very unsatisfactory; in addition, because the front-side illumination technology is usually used, the light must pass through a certain thickness of metal wiring layer before it hits the photosensitive area, making some oblique The incident light is blocked, reducing the quantum efficiency of this pixel structure.

Contents of the invention

The technical problem to be solved by the present invention is to provide an image sensor and its manufacturing method, which can improve the efficiency of light absorption.

In order to solve the above problems, the present invention provides an image sensor formed in a substrate with an insulating buried layer, the substrate insulating buried layer and the device layer on the surface of the insulating buried layer, the pixel circuit and the The photodiode is formed in the device layer, a first covering layer is provided on the surface of the photodiode, and a second covering layer is provided on the other surface of the substrate corresponding to the first covering layer, and the first covering layer is provided on the surface of the substrate corresponding to the first covering layer. The second covering layer is the light incident layer.

Optionally, the substrate further includes a support layer, the insulating layer is disposed between the support layer and the device layer, and the second covering layer is further disposed on the surface of the support layer; the thickness of the support layer is less than 5 μm.

Optionally, the material of the first covering layer is a reflection-enhancing material; the second covering layer is an anti-reflection film, or the material of the second covering layer is a single-sided transmission material, and the light passes from the outside of the medium layer to the medium The transmittance inside the layer is greater than the transmittance from inside the medium layer to outside the medium layer; the thickness range of the first covering layer and the second covering layer is 1nm to 10nm.

Optionally, a light focusing module is further arranged on the surface of the second covering layer.

The present invention further provides a method for manufacturing an image sensor, comprising the steps of: providing a substrate, the substrate including a supporting layer, an insulating layer on the surface of the supporting layer, and a device layer on the surface of the insulating layer; making an image in the device layer The pixel circuit and photodiode of the sensor; thinning the supporting layer; forming a first covering layer on the surface of the photodiode, and forming a second covering layer on the surface opposite to it; patterning the first covering layer to form the photodiode Electrical via; forms the electrical connection between the photodiode and the pixel circuitry.

Optionally, in the step of thinning the support layer, the support layer is further thinned to a thickness; the thickness of the support layer after thinning is less than 5 μm.

Optionally, further comprising a step of forming a light focusing module on the surface of the second covering layer.

The advantage of the present invention is that, through the multiple reflections of the incident light, it is absorbed multiple times in the photosensitive area, thereby improving the light absorption efficiency of the photosensitive unit based on the image sensor, and further adopting the backside illumination technology also avoids the incident light from being absorbed The shielding of the metal wiring layer improves the quantum efficiency of the image sensor.

Description of drawings

Accompanying drawing 1 is a kind of existing photosensitive unit structure of CMOS image sensor fabricated on SOI substrate.

Accompanying drawing 2 is a schematic diagram of implementation steps of the method described in the specific embodiment of the present invention.

Shown in accompanying drawing 3A to step 3G is the technological schematic diagram of the step shown in accompanying drawing 2.

Figure 4 is a schematic diagram of the Fabry-Perot cavity used in the specific embodiment of the present invention.

Detailed ways

Specific implementations of the image sensor and its manufacturing method provided by the present invention will be described in detail below in conjunction with the accompanying drawings.

Step S20, providing a substrate, said substrate comprising a support layer, an insulating layer on the surface of the support layer, and a device layer on the surface of the insulating layer; Step S21, fabricating a pixel circuit and a photodiode of an image sensor in the device layer; Step S22, Thinning the support layer to a certain thickness; step S23, forming a first covering layer on the surface of the photodiode, and forming a second covering layer on the surface opposite to it; step S24, patterning the first covering layer to form a photodiode electrical through hole; step S25, forming an electrical connection between the photodiode and the pixel circuit; step S26, forming a light focusing module on the surface of the second covering layer.

Shown in accompanying drawing 3A to step 3G is the technological schematic diagram of the step shown in accompanying drawing 2.

As shown in FIG. 3A , referring to step S20 , a substrate 30 is provided, and the substrate 30 includes a supporting layer 301 , an insulating layer 302 on the surface of the supporting layer 301 , and a device layer 303 on the surface of the insulating layer 302 . The materials of the support layer 301 and the device layer 303 may be any common substrate material in the field including monocrystalline silicon, and the materials of the support layer 301 and the device layer 303 may be the same or different. The material of the insulating layer 302 may be any common insulating material including silicon oxide, silicon nitride and silicon oxynitride. The thickness range of the device layer 303 is usually between 50 nm and 5 μm; the thickness range of the insulating layer 302 is usually between 50 nm and 300 nm.

As shown in accompanying drawing 3B, referring to step S21, the pixel circuit and photodiode 310 of the image sensor are manufactured in the device layer 303, and the pixel circuit specifically includes the reset transistor 150, the source follower transistor 160, the row selection transistor 170 and the electrical connection between them. connection.

In this specific embodiment, the device layer 303 has a thickness ranging from 50 nm to 500 nm, and the photodiode 310 is a lateral PIN structure. The photodiode 310 includes a P-type doped region 311 with a doping concentration greater than 1×10 ¹⁸ cm ⁻³ ; a fully depleted region 312 implanted with N-type or P-type impurity ions and a doping concentration less than 1×10 ¹⁵ cm ⁻³ , or not doped; the N-type doped region 313 has a doping concentration greater than 1×10 ¹⁸ cm ⁻³ . The P-type doped region 311, the fully depleted region 312, and the N-type doped region 313 are successively adjacent, and the doping concentration of the P-type doped region 311 and the N-type doped region 313 is higher than that of the fully depleted region 312. Three orders of magnitude or more, and ensure that the fully depleted region 312 is completely depleted, as the effective photosensitive region of the image sensor in this specific embodiment. The length (perpendicular to the depth direction) of the fully depleted region 312 is 1˜8 μm.

The working principle of the photodiode 310 is roughly that the photogenerated holes collected in the fully depleted region 312 will move to the P-type doped region 311 under the action of the built-in electric field, and the photogenerated electrons will also move to the P-type doped region 311 under the effect of the built-in electric field. N-type doped region 313 . Therefore, if the P-type doped region 311 in the above-mentioned photosensitive region is used, the photogenerated holes collected in the P-type doped region 311 can be leaked into the ground; the N-type doped region 313 is connected to the photoelectric signal processing circuit, The photo-generated electrons collected in the N-type doped region 313 can be read out.

In this step, as an optional embodiment, when the thickness of the device layer 303 is greater than 2 μm, the photodiode 310 may be a PN junction photodiode or a photogate structure.

As shown in FIG. 3C , referring to step S22 , the supporting layer 301 is thinned to a certain thickness. The support substrate 301 can be thinned by grinding or chemical etching. In this step, the thickness of the support layer 301 is preferably reduced to less than 5 μm, so as to enhance the transmittance of light from the support layer 301 . In other embodiments, the support layer 301 can also be completely removed to expose the insulating layer 302, which has the advantage that the self-stopping effect of the insulating layer 302 can be used to ensure the flatness of the corrosion surface. In such an embodiment, It should be ensured that the insulating layer 302 and the device layer 303 have sufficient mechanical strength.

As shown in FIG. 3D , referring to step S23 , a first covering layer 321 is formed on the surface of the photodiode 310 , and a second covering layer 322 is formed on the opposite surface. The material of the first covering layer 321 is a reflection-enhancing material with a reflectivity greater than 80%, which can be a silver film with a thickness ranging from 1 nm to 10 nm; the material of the second covering layer 322 is a single-sided transmission material, which can be poly Imide film, or a laminated structure composed of materials with different refractive indices, the transmittance of light from the outside of the substrate 30 to the inside of the substrate 30 (usually greater than 70%) is greater than the transmission from the inside of the substrate 30 to the outside of the substrate 30 rate (usually less than 30%), and the thickness ranges from 1nm to 10nm.

As an optional implementation manner, the second covering layer 322 may also be an anti-reflection film, specifically a silicon nitride film, a silicon oxynitride film, a HfO ₂ film, a SiN _x O _y :H film, or the like.

As shown in FIG. 3E , referring to step S24, the first covering layer 321 is patterned to form the electrical through holes of the photodiode 310, including the through holes 331 corresponding to the P-type doped regions 311 and the through holes corresponding to the N-type doped regions 313. through hole 333 . The patterning process can adopt common photolithography and etching processes in the art, which will not be repeated here.

As shown in FIG. 3F , referring to step S25 , the electrical connection between the photodiode 310 and the pixel circuit is formed, including the grounding of the P-type doped region 311 and the electrical connection between the N-type doped region 313 and the reset transistor 150 . The manufacturing methods for forming the electrical connection are well-known techniques for those skilled in the art, and will not be repeated here.

As shown in FIG. 3G , referring to step S26 , a light focusing module is formed on the surface of the second covering layer. A light focusing module 390 is formed on the surface of the second covering layer 322 . The light focusing module 390 includes a color filter 391 and a microlens 392 on the color filter 391 . Both color filters 391 and microlenses 392 are common components of CMOS image sensors. Its functions and manufacturing methods are well known to those skilled in the art, and will not be described in detail. It should be noted that the wavelength ranges that the red, green and blue color filters can pass must cover the three maximum emission wavelengths derived in the above derivation, that is, λ=743nm, corresponding to the red color filter; λ=437nm , corresponding to the green color filter; λ=400nm, corresponding to the blue color filter.

Figure 3G shows the image sensor described in this specific embodiment, which is formed in a substrate 300 with an insulating buried layer 302 and a device layer 303, a photodiode 310, a reset transistor 150 of a pixel circuit, and a source follower The transistor 160 and the row selection transistor 170 are formed in the device layer 303 , the surface of the photodiode 310 is covered with a first covering layer 321 , and the opposite surface is covered with a second covering layer 322 . If the supporting layer 301 is completely removed, the image sensor may also only include the insulating layer 302 and the device layer 303 . The photodiode 310 includes a P-type doped region 311 with a doping concentration greater than 1×10 ¹⁸ cm ⁻³ ; a fully depleted region 312 implanted with N-type or P-type impurity ions and a doping concentration less than 1×10 ¹⁵ cm ⁻³ , or not doped; the N-type doped region 313 has a doping concentration greater than 1×10 ¹⁸ cm ⁻³ , and the second covering layer 322 is a light incident layer.

The first covering layer 321 , the supporting layer 301 , the insulating layer 302 , the device layer 303 and the second covering layer 322 together form a Fabry-Perot cavity. Shown in accompanying drawing 4 is the schematic diagram of Fabry-Perot cavity. When light of a certain wavelength is injected into the Fabry-Perot cavity (photosensitive area), part of the light will be refracted repeatedly in the cavity, and the other part of the light will pass out of the cavity through the film on the surface of the cavity. For a Fabry-Perot cavity, the ratio of total reflected light to transmitted light depends on the material properties of the cavity surface and the thickness of the cavity. By properly selecting the material and the thickness of the cavity, most of the incident light can be reflected in the cavity, and thus absorbed by the photosensitive area multiple times.

This specific embodiment utilizes the characteristics of the Fabry-Perot cavity, and through multiple reflections of the incident light, it is absorbed multiple times in the photosensitive area, thereby improving the light absorption efficiency based on the image sensor.

Specifically, if the length of the cavity is L, the wavelength of the incident light is λ, the angle of incidence is θ, the reflectivity of the reflective film (assuming the reflectivity of the films on both sides is the same) is R, the cavity is a material with a uniform refractive index and the refractive index is n, then the emissivity _RE of the Fabry-Perot cavity for the incident light (the ratio of all reflected light energy in the cavity to the incident light energy) is $R_{\mathrm{E.}}=\frac{f{\sin }^2\left(\frac{\mathrm{\δ}}{2}\right)}{1+f{\sin }^2\left(\frac{\mathrm{\δ}}{2}\right)}.$ in $f=\frac{4R}{{(1-R)}^2},$ called fineness; $\mathrm{\δ}=\frac{2\mathrm{\π}}{\mathrm{\λ}}\&Center\; Dot;2\mathrm{nL}\cos \mathrm{\θ}$ is the phase difference between two adjacent beams of reflected light.

（k取自然数）。而在图像传感器应用中，入射角θ一般为0，于是

即入射光取该波长时，法布里珀罗腔有最大的反射率。To maximize the total reflectivity R _E of the cavity, it can be solved as follows:

(k is a natural number). In image sensor applications, the incident angle θ is generally 0, so

That is, when the incident light takes this wavelength, the Fabry-Perot cavity has the maximum reflectivity.

Note that the cavity of the Fabry-Perot cavity in this specific embodiment actually has three layers (support layer 301, insulating layer 302, device layer 303, if the support layer 301 is completely removed in other specific embodiments, then it should There are two layers), not a uniform index material. at this time ${\mathrm{\δ}}^{\′}=\frac{2\mathrm{\π}}{\mathrm{\λ}}\·(2{\mathrm{no}}_1{L}_1\cos {\mathrm{\θ}}_1+2{\mathrm{no}}_2{L}_2\cos {\mathrm{\θ}}_2+2{\mathrm{no}}_3{L}_3\cos {\mathrm{\θ}}_3).$ Where n _S is the refractive index of the support layer 301 and the device layer 303 (simplified as both are the same material), and n _O is the refractive index of the insulating layer 302 . In fact, when the angle of incidence is normal incidence, that is, θ=0, ${\mathrm{\δ}}_0=\frac{2\mathrm{\π}}{\mathrm{\λ}}\&Center\; Dot;(2{\mathrm{no}}_S{L}_1+2{\mathrm{no}}_o{L}_2+2{\mathrm{no}}_S{L}_3)=\frac{2\mathrm{\π}}{\mathrm{\λ}}\&Center\; Dot;2{\mathrm{no}}_{\mathrm{eff}}L.$ Since in the actual process, the sum of the thickness L ₁ of the device layer 303 and the thickness L ₃ of the supporting layer 301 is much greater than the thickness L ₂ of the insulating layer 302, so n _eff ≈ n _S =3.42, that is, the Fabry of the structure of this specific embodiment The Perot cavity can be equivalent to the uniform material of the device layer 303 dielectric cavity.

Appropriately take the value of the cavity length L, under the condition that k takes several different natural numbers, the Fabry-Perot cavity structure in the present invention can simultaneously have the maximum Reflectivity, so that the incident light is reflected and absorbed multiple times in the photosensitive area. In this embodiment, the cavity medium is silicon, and L=380nm, when k=3, λ=743nm, corresponding to red light; when k=5, λ=437nm, corresponding to green light; when k=6, λ=400nm, corresponding to blue light.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications should also be considered Be the protection scope of the present invention.

Claims (10)

1. An image sensor is formed in a substrate with an insulating buried layer, the substrate insulating buried layer and the device layer on the surface of the insulating buried layer, and the pixel circuit and photodiode of the image sensor are formed in the device layer , characterized in that a first covering layer is provided on the surface of the photodiode, and a second covering layer is provided on the other surface of the substrate corresponding to the first covering layer, and the second covering layer is the light incident layer. 2. The image sensor according to claim 1, wherein the substrate further comprises a supporting layer, the insulating layer is disposed between the supporting layer and the device layer, and the second covering layer is further disposed on the the surface of the support layer. 3. The image sensor according to claim 2, wherein the supporting layer has a thickness less than 5 μm. 4. The image sensor according to claim 1, wherein the material of the first covering layer is a reflection enhancing material; the second covering layer is an anti-reflection film, or the material of the second covering layer is For single-sided transmissive materials, the transmittance of light from the outside of the medium layer to the inside of the medium layer is greater than the transmittance from the inside of the medium layer to the outside of the medium layer. 5. The image sensor according to claim 1 or 4, characterized in that, the thickness of the first covering layer and the second covering layer both range from 1 nm to 10 nm. 6 . The image sensor according to claim 1 , wherein a light focusing module is further arranged on the surface of the second covering layer. 7. A method for manufacturing an image sensor, comprising the steps of: A substrate is provided, and the substrate includes a support layer, an insulating layer on the surface of the support layer, and a device layer on the surface of the insulating layer; Make the pixel circuit and photodiode of the image sensor in the device layer; Thinning of the support layer; forming a first covering layer on the surface of the photodiode, and forming a second covering layer on the surface opposite thereto; patterning the first covering layer to form electrical vias for photodiodes; An electrical connection is formed between the photodiode and the pixel circuit. 8 . The manufacturing method of the image sensor according to claim 7 , wherein in the step of thinning the supporting layer, the supporting layer is further thinned to a certain thickness. 9 . The method for manufacturing an image sensor according to claim 8 , wherein the thickness of the support layer after thinning is less than 5 μm. 10 . The method for manufacturing an image sensor according to claim 7 , further comprising a step of forming a light focusing module on the surface of the second covering layer. 11 .

Images