CN102842590B - Image sensor and manufacturing method thereof - Google Patents

Abstract

The invention provides an image sensor and a manufacturing method thereof, which can be used for improving the light adsorption efficiency. The image sensor is formed in a substrate with an insulating buried layer; the substrate comprises the insulating buried layer and a device layer on the surface of the insulating buried layer. A pixel circuit and a photodiode of the image sensor are formed in the device layer. The surface of the photodiode is provided with a first cover layer. The position on the other surface of the substrate, which corresponds to the first cover layer, is provided with a second cover layer. The second cover layer is a light incidence layer. The image sensor and the manufacturing method thereof have the advantages that incident light is repeatedly reflected, so that the incident light is repeatedly adsorbed in a light sensitive region, and thus, the light adsorption efficiency on the basis of a light sensitive unit of the image sensor is improved; and furthermore, due to the adoption of a back irradiation technology, the incident light is also avoided being shielded by a metal wiring layer and the quantum efficiency of the image sensor is improved.

Description

Imageing sensor and manufacture method thereof

Technical field

The present invention relates to field of image sensors, particularly relate to a kind of imageing sensor and manufacture method thereof.

Background technology

SOI(Silicon-On-Insulator, the silicon in dielectric substrate) technology be at the bottom of top layer silicon and backing between introduce one deck and bury oxide layer.By forming semiconductive thin film on insulator, SOI material is provided with the incomparable advantage of traditional body silicon materials: the medium isolation that can realize components and parts in integrated circuit, completely eliminates the parasitic latch-up in Bulk CMOS circuit; The integrated circuit adopting this material to make also has that parasitic capacitance is little, integration density is high, speed is fast, technique is simple, short-channel effect is little and be specially adapted to the advantages such as low voltage and low power circuits.

Imageing sensor is a kind of semiconductor device optical imagery being converted to the signal of telecommunication, is generally made up of photosensitive pixel and cmos signal treatment circuit.Cmos image sensor common is at present active pixel type imageing sensor (APS), wherein be divided into again three pipe imageing sensor (3T, comprise reset transistor, amplifier transistor and row selecting transistor) and the large class of four pipe imageing sensors (4T comprises transfering transistor, reset transistor, amplifier transistor and row selecting transistor) two.

A kind ofly existingly be made in cmos image sensor pixel cell structure in SOI substrate as shown in Figure 1, employing be fully-depleted structure, comprising: substrate 100, oxygen buried layer 110 and device layer 130.Device layer 130 comprises photodiode 140, reset transistor 150, source following transistor 160 and row and selects transistor 170.The photosensitive area of this dot structure is mainly positioned at the PN junction depletion region of photodiode 140.Each transistor includes the basic structures such as source electrode, grid and drain electrode.Position relationship and the electricity annexation of each device above-mentioned please refer to accompanying drawing 1.

With reference to accompanying drawing 1, the operation principle of existing dot structure is: when starting working, first reset transistor 150 grid is added high level, make its conducting, during exposure, photodiode 140 as photoelectron collecting zone, when incident light irradiates, produce electron hole pair, after completion of the exposure and select transistor 170 to be read by integral voltage signal by source following transistor 160 and row.So the value of output voltage just reflects the power of light signal.

The shortcoming of the above-mentioned cmos image sensor image element circuit be made in SOI substrate is that the device layer 130 of SOI substrate is very thin, usually dozens of micron even tens microns are only had, the light path of incident light in photodiode 140 is very short, cause efficiency of light absorption and quantum efficiency low.Especially wavelength is greater than to the redness of 600nm, orange-colored light, absorption efficiency is extremely low, and image quality is very undesirable; In addition, owing to usually adopting front illuminated technology, light must pass over certain thickness metal wiring layer before being irradiated to photosensitive area, part oblique incident ray is blocked, reduces the quantum efficiency of this dot structure.

Summary of the invention

Technical problem to be solved by this invention is, provides a kind of imageing sensor and manufacture method thereof, can improve the efficiency of light absorption.

In order to solve the problem, the invention provides a kind of imageing sensor, be formed in the substrate with insulating buried layer, the device layer on described insulated substrate buried regions and insulating buried layer surface, the image element circuit of described imageing sensor and photodiode are formed in device layer, the surface of described photodiode is provided with the first cover layer, and another position corresponding with the first cover layer, surface of described substrate is provided with the second cover layer, described second cover layer is light incident layer.

Optionally, described substrate comprises supporting layer further, and described insulating barrier is arranged between supporting layer and device layer, and described second cover layer is be arranged on described support layer surface further; The thickness of described supporting layer is less than 5 μm.

Optionally, described first tectal material is reflection enhancing material; Described second cover layer is anti-reflection film, or described second tectal material is one side transmission material, and light is greater than in dielectric layer to the transmissivity dielectric layer from dielectric layer to the transmissivity in dielectric layer; Described first cover layer and the second tectal thickness range are 1nm to 10nm.

Optionally, light focus module is provided with further on described second tectal surface.

Invention further provides a kind of manufacture method of imageing sensor, comprise the steps: to provide a substrate, described substrate comprises supporting layer, the insulating barrier of support layer surface and the device layer of surface of insulating layer; The image element circuit of making image transducer and photodiode in device layer; Thinning supporting layer; Form the first cover layer on the surface of described photodiode, and form the second cover layer on surface on the other side; Graphical first cover layer, forms the electricity through hole of photodiode; Form photodiode to be connected with the electricity between image element circuit.

Optionally, in the step of described thinning supporting layer, be that supporting layer is thinned to a thickness further; Thickness after described supporting layer is thinning is less than 5 μm.

Optionally, the step that described second tectal surface forms light focus module is included in further.

The invention has the advantages that, by the multiple reflections of incident ray, make it in photosensitive area by re-absorption effect, thus the efficiency of light absorption that improve based on imageing sensor photosensitive unit, further employing back side illuminaton technology it also avoid incident light and receives blocking of metal wiring layer, improves the quantum efficiency of imageing sensor.

Accompanying drawing explanation

Accompanying drawing 1 is a kind of existing cmos image sensor photosensitive unit structure be made in SOI substrate.

It is the implementation step schematic diagram of method described in the specific embodiment of the present invention shown in accompanying drawing 2.

It is the process schematic representation of step shown in accompanying drawing 2 shown in accompanying drawing 3A to step 3G.

It is the schematic diagram in the Fabry Perot chamber that the specific embodiment of the present invention adopts shown in accompanying drawing 4.

Embodiment

Elaborate below in conjunction with the embodiment of accompanying drawing to imageing sensor provided by the invention and manufacture method thereof.

Step S20, provides a substrate, and described substrate comprises supporting layer, the insulating barrier of support layer surface and the device layer of surface of insulating layer; Step S21, the image element circuit of making image transducer and photodiode in device layer; Step S22, thinning supporting layer to thickness; Step S23, forms the first cover layer on the surface of described photodiode, and forms the second cover layer on surface on the other side; Step S24, graphical first cover layer, forms the electricity through hole of photodiode; Step S25, forms photodiode and is connected with the electricity between image element circuit; Step S26, forms light focus module on described second tectal surface.

It is the process schematic representation of step shown in accompanying drawing 2 shown in accompanying drawing 3A to step 3G.

Shown in accompanying drawing 3A, refer step S20, provides a substrate 30, and described substrate 30 comprises supporting layer 301, the insulating barrier 302 on supporting layer 301 surface and the device layer 303 on insulating barrier 302 surface.The material of described supporting layer 301 and device layer 303 can be comprise backing material common in middle arbitrarily this area of monocrystalline silicon, and supporting layer 301 can be identical or different with the material of device layer 303.The material of described insulating barrier 302 can be any one the common insulating material comprising silica, silicon nitride and silicon oxynitride.The thickness range of device layer 303 normally 50nm between 5 μm; The thickness range of insulating barrier 302 is normally between 50nm to 300nm.

Shown in accompanying drawing 3B, refer step S21, the image element circuit of making image transducer and photodiode 310 in device layer 303, image element circuit specifically comprises reset transistor 150, source following transistor 160 and row and selects transistor 170 and electricity line each other.

In this embodiment, the thickness range of device layer 303 is 50nm to 500nm, and photodiode 310 is horizontal PIN structural.Described photodiode 310 comprises P type doped region 311, and doping content is greater than 1 × 10 ¹⁸cm ^-3; Fully-depleted district 312, adopt N-type or p type impurity ion implantation, and doping content is less than 1 × 10 ¹⁵cm ^-3, or undope; N-type doped region 313, its doping content is greater than 1 × 10 ¹⁸cm ^-3.P type doped region 311, fully-depleted district 312, N-type doped region 313 are adjacent successively, doping content three orders of magnitude higher than the doping content in fully-depleted district 312 of P type doped region 311 and N-type doped region 313 or more, and ensure that fully-depleted district 312 is completely exhausted, as effective photosensitive area of imageing sensor in this embodiment.(perpendicular to the depth direction) length in fully-depleted district 312 is 1 ~ 8 μm.

The operation principle of photodiode 310 is roughly that the photohole collected in fully-depleted district 312 can shift to P type doped region 311 under the effect at internal electric field, and light induced electron also can shift to N-type doped region 313 under the effect of internal electric field.So, if by the P type doped region 311 in above-mentioned photosensitive area, just the photohole be collected in P type doped region 311 can be let out ground end; N-type doped region 313 is connected to photoelectric signal processing circuit, the light induced electron be collected in N-type doped region 313 can be read.

In this step, as embodiment, when device layer 303 thickness is greater than 2 μm, photodiode 310 can be PN junction light sensitive diode or photoelectricity door.

Shown in accompanying drawing 3C, refer step S22, thinning supporting layer 301 to thickness.The thinning support substrates 301 of grinding method that is thinning or chemical corrosion can be adopted.In this step, preferably by the reduced thickness of described supporting layer 301 to being less than 5 μm, to strengthen the transmitance of light from supporting layer 301.In other implementations, also supporting layer 301 can be removed to expose insulating barrier 302 completely, such advantage is the self-stopping technology effect that can utilize insulating barrier 302, ensure the evenness of erosional surface, in such execution mode, should ensure that insulating barrier 302 and device layer 303 have enough mechanical strengths.

Shown in accompanying drawing 3D, refer step S23, forms the first cover layer 321 on the surface of described photodiode 310, and forms the second cover layer 322 on surface on the other side.The material of described first cover layer 321 is reflection enhancing material, and reflectivity is greater than 80%, can be silverskin, the equal 1nm to 10nm of thickness range; The material of described second cover layer 322 is one side transmission material, it can be polyimide film, or there is the laminated construction of different refractivity material composition, light is greater than to the transmissivity (being usually less than 30%) substrate 30 in substrate 30 from substrate 30 to the transmissivity (being usually greater than 70%) in substrate 30, the equal 1nm to 10nm of thickness range.

As Alternate embodiments, the second cover layer 322 can also be anti-reflection film, is specifically as follows silicon nitride film, silicon oxynitride film, HfO ₂film, SiN _xo _y: H film etc.

Shown in accompanying drawing 3E, refer step S24, graphical first cover layer 321, forms the electricity through hole of photodiode 310, comprises the through hole 331 corresponding to P type doped region 311 and the through hole 333 corresponding to N-type doped region 313.Patterning process can adopt photoetching common in this area and etching process, repeats no more herein.

Shown in accompanying drawing 3F, refer step S25, forms photodiode 310 and is connected with the electricity between image element circuit, comprise the ground connection of P type doped region 311 and the electricity between N-type doped region 313 with reset transistor 150 is connected.The manufacturing process forming electricity connection is the known technology of those skilled in the art, does not repeat.

Shown in accompanying drawing 3G, refer step S26, forms light focus module on described second tectal surface.Light focus module 390 is formed on the surface of described second cover layer 322.The lenticule 392 that light focus module 390 comprises colour filter 391 and is positioned on colour filter 391.Colour filter 391 and lenticule 392 are the general component part of cmos image sensor.Its function and manufacturing process are the known technology of those skilled in the art, do not repeat.It should be noted that red, green, blue three kinds of colour filters by wave-length coverage, three maximum emission wavelengths, i.e. λ=743nm deriving in above-mentioned derivation must be covered, corresponding red color filter; λ=437nm, corresponding green color filter; λ=400nm, corresponding blue color filter.

The imageing sensor described in this embodiment is shown in accompanying drawing 3G, be formed in the substrate 300 with insulating buried layer 302 and device layer 303, photodiode 310, transistor 170 etc. is selected to be formed in device layer 303 with the reset transistor 150 of image element circuit, source following transistor 160 and row, the surface coverage of photodiode 310 has the first cover layer 321, and surface coverage on the other side has the second cover layer 322.If all removed by supporting layer 301, then imageing sensor also can only include insulating barrier 302 and device layer 303.Described photodiode 310 comprises P type doped region 311, and doping content is greater than 1 × 10 ¹⁸cm ^-3; Fully-depleted district 312, adopt N-type or p type impurity ion implantation, and doping content is less than 1 × 10 ¹⁵cm ^-3, or undope; N-type doped region 313, its doping content is greater than 1 × 10 ¹⁸cm ^-3, described second cover layer 322 is light incident layer.

First cover layer 321, supporting layer 301, insulating barrier 302, device layer 303 and the second cover layer 322 together constitute a Fabry Perot chamber.It is the schematic diagram in Fabry Perot chamber shown in accompanying drawing 4.After the light of a certain specific wavelength injects Fabry Perot chamber (photosensitive area), have part light and repeatedly reflect in cavity, another part light can appear chamber from the film of housing surface.For a Fabry Perot chamber, the ratio of the photosynthetic transmitted light of total reflection depends on the thickness of housing surface material character and cavity.By suitable selection material and cavity thickness, most of incident light can be allowed to become the reverberation in cavity, thus by photosensitive area re-absorption effect.

This embodiment make use of the feature in Fabry Perot chamber, by the multiple reflections of incident ray, makes it by re-absorption effect in photosensitive area, thus improves the efficiency of light absorption based on imageing sensor.

Concrete, if cavity length is L, lambda1-wavelength is λ, and incidence angle is θ, and the reflectivity (supposing that the film reflectivity of both sides is identical) of reflectance coating is R, and cavity is homogeneous refraction materials and refractive index is n, then Fabry Perot chamber is for the emissivity R of this incident light _e(all energy of reflection light in cavity account for the ratio of incident light energy) is $R_E=\frac{F{\sin }^2\left(\frac{\mathrm{\δ}}{2}\right)}{1+F{\sin }^2\left(\frac{\mathrm{\δ}}{2}\right)}.$ Wherein $F=\frac{4R}{{(1-R)}^2},$ Be called fineness; $\mathrm{\δ}=\frac{2\mathrm{\π}}{\mathrm{\λ}}\·2\mathrm{nL}\cos \mathrm{\θ}$ For the difference of adjacent two bundle reverberation.

To make the total reflectivity R of cavity _eget maximum, can solve: (k gets natural number).And in image sensor application, incidence angle θ is generally 0, so namely, when incident light gets this wavelength, there is maximum reflectivity in Fabry Perot chamber.

Notice, the cavity in the Fabry Perot chamber in this embodiment is actual three layers of (supporting layer 301, insulating barrier 302, device layer 303, if supporting layer 301 is entirely removed in other embodiments, then should have two-layer), but not homogeneous refraction materials.Now ${\mathrm{\δ}}^{\′}=\frac{2\mathrm{\π}}{\mathrm{\λ}}\·(2n_1{L}_1\cos {\mathrm{\θ}}_1+2n_2{L}_2\cos {\mathrm{\θ}}_2+2n_3{L}_3\cos {\mathrm{\θ}}_3).$ Wherein n _sfor supporting layer 301 and device layer 303(are reduced to both for same material) refractive index, n _ofor the refractive index of insulating barrier 302.In fact, when incidence angle is normal incidence, namely during θ=0, ${\mathrm{\δ}}_0=\frac{2\mathrm{\π}}{\mathrm{\λ}}\·(2n_S{L}_1+2n_O{L}_2+2n_S{L}_3)=\frac{2\mathrm{\π}}{\mathrm{\λ}}\·2n_{\mathrm{eff}}L.$ Due in actual process, device layer 303 thickness L ₁with the thickness L of supporting layer 301 ₃sum is much larger than insulating barrier 302 thickness L ₂, therefore n _eff≈ n _s=3.42, namely the Fabry Perot cavity of this embodiment structure can be equivalent to homogeneous device layer 303 medium cavity material.

Suitably get the value of cavity length L, under k gets several different natural situation, Fabry Perot cavity configuration in the present invention can have maximum reflectivity to the visible ray of ruddiness, green glow and blue light three wave bands simultaneously, thus makes incident light by multiple reflections in photosensitive area, re-absorption effect.In the present embodiment, cavity medium is silicon, gets L=380nm, as k=3, and λ=743nm, corresponding ruddiness; As k=5, λ=437nm, corresponding green glow; As k=6, λ=400nm, corresponding blue light.

The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (9)

1. an imageing sensor, be formed in the substrate with insulating buried layer, the device layer on described insulated substrate buried regions and insulating buried layer surface, the image element circuit of described imageing sensor and photodiode are formed in device layer, it is characterized in that, the surface of described photodiode is provided with the first cover layer, and another position corresponding with the first cover layer, surface of described substrate is provided with the second cover layer, described second cover layer is light incident layer, described first tectal material is reflection enhancing material, described second tectal material is one side transmission material, light is greater than in dielectric layer to the transmissivity dielectric layer from dielectric layer to the transmissivity in dielectric layer, described first cover layer and the second cover layer together constitute a Fabry Perot chamber.

2. imageing sensor according to claim 1, is characterized in that, described substrate comprises supporting layer further, and described insulating barrier is arranged between supporting layer and device layer, and described second cover layer is be arranged on described support layer surface further.

3. imageing sensor according to claim 2, is characterized in that, the thickness of described supporting layer is less than 5 μm.

4. imageing sensor according to claim 1, is characterized in that, described first cover layer and the second tectal thickness range are 1nm to 10nm.

5. imageing sensor according to claim 1, is characterized in that, is provided with light focus module further on described second tectal surface.

6. a manufacture method for imageing sensor, is characterized in that, comprises the steps:

There is provided a substrate, described substrate comprises supporting layer, the insulating barrier of support layer surface and the device layer of surface of insulating layer;

The image element circuit of making image transducer and photodiode in device layer;

Thinning supporting layer;

The first cover layer is formed on the surface of described photodiode, and form the second cover layer on surface on the other side, described first tectal material is reflection enhancing material, described second tectal material is one side transmission material, light is greater than to the transmissivity dielectric layer in dielectric layer from dielectric layer to the transmissivity in dielectric layer, and described first cover layer and the second cover layer together constitute a Fabry Perot chamber;

Graphical first cover layer, forms the electricity through hole of photodiode;

Form photodiode to be connected with the electricity between image element circuit.

7. the manufacture method of imageing sensor according to claim 6, is characterized in that, in the step of described thinning supporting layer, is that supporting layer is thinned to a thickness further.

8. the manufacture method of imageing sensor according to claim 7, is characterized in that, the thickness after described supporting layer is thinning is less than 5 μm.

9. method for making image sensor according to claim 6, is characterized in that, is included in the step that described second tectal surface forms light focus module further.